{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 16155,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00018570102135561745,
      "grad_norm": 12.281526850417075,
      "learning_rate": 0.0,
      "loss": 0.9325,
      "mean_token_accuracy": 0.7545486688613892,
      "num_tokens": 38904.0,
      "step": 1
    },
    {
      "epoch": 0.0003714020427112349,
      "grad_norm": 15.251340171602223,
      "learning_rate": 6.188118811881188e-10,
      "loss": 0.9306,
      "mean_token_accuracy": 0.7646465301513672,
      "num_tokens": 69345.0,
      "step": 2
    },
    {
      "epoch": 0.0005571030640668524,
      "grad_norm": 13.04541349135724,
      "learning_rate": 1.2376237623762375e-09,
      "loss": 1.0061,
      "mean_token_accuracy": 0.7449073791503906,
      "num_tokens": 105744.0,
      "step": 3
    },
    {
      "epoch": 0.0007428040854224698,
      "grad_norm": 12.364130570125262,
      "learning_rate": 1.8564356435643563e-09,
      "loss": 0.9813,
      "mean_token_accuracy": 0.7486875653266907,
      "num_tokens": 144063.0,
      "step": 4
    },
    {
      "epoch": 0.0009285051067780873,
      "grad_norm": 16.132561456473617,
      "learning_rate": 2.475247524752475e-09,
      "loss": 0.9711,
      "mean_token_accuracy": 0.7575291395187378,
      "num_tokens": 172781.0,
      "step": 5
    },
    {
      "epoch": 0.0011142061281337048,
      "grad_norm": 14.92428448145872,
      "learning_rate": 3.0940594059405942e-09,
      "loss": 0.982,
      "mean_token_accuracy": 0.7524327635765076,
      "num_tokens": 204706.0,
      "step": 6
    },
    {
      "epoch": 0.0012999071494893223,
      "grad_norm": 15.348579560430322,
      "learning_rate": 3.7128712871287126e-09,
      "loss": 0.9513,
      "mean_token_accuracy": 0.7632513046264648,
      "num_tokens": 235707.0,
      "step": 7
    },
    {
      "epoch": 0.0014856081708449396,
      "grad_norm": 14.013230345371205,
      "learning_rate": 4.331683168316832e-09,
      "loss": 0.9412,
      "mean_token_accuracy": 0.7626265287399292,
      "num_tokens": 269057.0,
      "step": 8
    },
    {
      "epoch": 0.001671309192200557,
      "grad_norm": 15.760408457706534,
      "learning_rate": 4.95049504950495e-09,
      "loss": 0.9441,
      "mean_token_accuracy": 0.7618371248245239,
      "num_tokens": 298630.0,
      "step": 9
    },
    {
      "epoch": 0.0018570102135561746,
      "grad_norm": 15.111743466195906,
      "learning_rate": 5.569306930693069e-09,
      "loss": 0.9548,
      "mean_token_accuracy": 0.7563327550888062,
      "num_tokens": 329471.0,
      "step": 10
    },
    {
      "epoch": 0.002042711234911792,
      "grad_norm": 12.076420602602257,
      "learning_rate": 6.1881188118811884e-09,
      "loss": 0.905,
      "mean_token_accuracy": 0.7640678882598877,
      "num_tokens": 368856.0,
      "step": 11
    },
    {
      "epoch": 0.0022284122562674096,
      "grad_norm": 16.02807233112152,
      "learning_rate": 6.806930693069306e-09,
      "loss": 0.9759,
      "mean_token_accuracy": 0.7575510144233704,
      "num_tokens": 397706.0,
      "step": 12
    },
    {
      "epoch": 0.002414113277623027,
      "grad_norm": 13.916897718386572,
      "learning_rate": 7.425742574257425e-09,
      "loss": 0.9238,
      "mean_token_accuracy": 0.7606444358825684,
      "num_tokens": 430993.0,
      "step": 13
    },
    {
      "epoch": 0.0025998142989786446,
      "grad_norm": 13.28039173605592,
      "learning_rate": 8.044554455445545e-09,
      "loss": 0.9801,
      "mean_token_accuracy": 0.750443160533905,
      "num_tokens": 466582.0,
      "step": 14
    },
    {
      "epoch": 0.002785515320334262,
      "grad_norm": 14.560732212778332,
      "learning_rate": 8.663366336633664e-09,
      "loss": 0.958,
      "mean_token_accuracy": 0.7593448162078857,
      "num_tokens": 499316.0,
      "step": 15
    },
    {
      "epoch": 0.002971216341689879,
      "grad_norm": 14.23506714102803,
      "learning_rate": 9.28217821782178e-09,
      "loss": 0.9475,
      "mean_token_accuracy": 0.7577542662620544,
      "num_tokens": 532813.0,
      "step": 16
    },
    {
      "epoch": 0.003156917363045497,
      "grad_norm": 14.924098584735386,
      "learning_rate": 9.9009900990099e-09,
      "loss": 0.9966,
      "mean_token_accuracy": 0.746097981929779,
      "num_tokens": 563979.0,
      "step": 17
    },
    {
      "epoch": 0.003342618384401114,
      "grad_norm": 12.141482429002664,
      "learning_rate": 1.051980198019802e-08,
      "loss": 0.9268,
      "mean_token_accuracy": 0.7580799460411072,
      "num_tokens": 603702.0,
      "step": 18
    },
    {
      "epoch": 0.003528319405756732,
      "grad_norm": 17.562126362757322,
      "learning_rate": 1.1138613861386138e-08,
      "loss": 1.0232,
      "mean_token_accuracy": 0.7445839643478394,
      "num_tokens": 630803.0,
      "step": 19
    },
    {
      "epoch": 0.003714020427112349,
      "grad_norm": 13.688906428315182,
      "learning_rate": 1.1757425742574257e-08,
      "loss": 0.9996,
      "mean_token_accuracy": 0.7468609809875488,
      "num_tokens": 665104.0,
      "step": 20
    },
    {
      "epoch": 0.0038997214484679664,
      "grad_norm": 13.379946166022373,
      "learning_rate": 1.2376237623762377e-08,
      "loss": 0.9947,
      "mean_token_accuracy": 0.7488545179367065,
      "num_tokens": 701122.0,
      "step": 21
    },
    {
      "epoch": 0.004085422469823584,
      "grad_norm": 11.105612438965306,
      "learning_rate": 1.2995049504950493e-08,
      "loss": 0.8731,
      "mean_token_accuracy": 0.7719202637672424,
      "num_tokens": 744217.0,
      "step": 22
    },
    {
      "epoch": 0.004271123491179201,
      "grad_norm": 15.599882675615945,
      "learning_rate": 1.3613861386138613e-08,
      "loss": 1.0059,
      "mean_token_accuracy": 0.7472463846206665,
      "num_tokens": 773950.0,
      "step": 23
    },
    {
      "epoch": 0.004456824512534819,
      "grad_norm": 12.947729388300733,
      "learning_rate": 1.4232673267326732e-08,
      "loss": 0.9167,
      "mean_token_accuracy": 0.7639309167861938,
      "num_tokens": 810069.0,
      "step": 24
    },
    {
      "epoch": 0.004642525533890436,
      "grad_norm": 13.178349079560427,
      "learning_rate": 1.485148514851485e-08,
      "loss": 0.9995,
      "mean_token_accuracy": 0.7471704483032227,
      "num_tokens": 845720.0,
      "step": 25
    },
    {
      "epoch": 0.004828226555246054,
      "grad_norm": 14.347105205872719,
      "learning_rate": 1.5470297029702968e-08,
      "loss": 1.0473,
      "mean_token_accuracy": 0.7324591279029846,
      "num_tokens": 878336.0,
      "step": 26
    },
    {
      "epoch": 0.005013927576601671,
      "grad_norm": 13.123106527275354,
      "learning_rate": 1.608910891089109e-08,
      "loss": 0.9279,
      "mean_token_accuracy": 0.7625272870063782,
      "num_tokens": 914698.0,
      "step": 27
    },
    {
      "epoch": 0.005199628597957289,
      "grad_norm": 14.417503388444395,
      "learning_rate": 1.6707920792079207e-08,
      "loss": 0.9548,
      "mean_token_accuracy": 0.757725715637207,
      "num_tokens": 947850.0,
      "step": 28
    },
    {
      "epoch": 0.005385329619312906,
      "grad_norm": 14.658229527103591,
      "learning_rate": 1.732673267326733e-08,
      "loss": 0.9582,
      "mean_token_accuracy": 0.7585372924804688,
      "num_tokens": 979825.0,
      "step": 29
    },
    {
      "epoch": 0.005571030640668524,
      "grad_norm": 13.178341800885647,
      "learning_rate": 1.7945544554455443e-08,
      "loss": 0.9878,
      "mean_token_accuracy": 0.744724452495575,
      "num_tokens": 1015113.0,
      "step": 30
    },
    {
      "epoch": 0.005756731662024141,
      "grad_norm": 13.294922950883441,
      "learning_rate": 1.856435643564356e-08,
      "loss": 0.9289,
      "mean_token_accuracy": 0.764358401298523,
      "num_tokens": 1050245.0,
      "step": 31
    },
    {
      "epoch": 0.005942432683379758,
      "grad_norm": 15.717257818979252,
      "learning_rate": 1.9183168316831682e-08,
      "loss": 1.0001,
      "mean_token_accuracy": 0.7461848258972168,
      "num_tokens": 1079371.0,
      "step": 32
    },
    {
      "epoch": 0.006128133704735376,
      "grad_norm": 11.893446327334946,
      "learning_rate": 1.98019801980198e-08,
      "loss": 0.956,
      "mean_token_accuracy": 0.7577181458473206,
      "num_tokens": 1120219.0,
      "step": 33
    },
    {
      "epoch": 0.006313834726090994,
      "grad_norm": 12.784041399063018,
      "learning_rate": 2.042079207920792e-08,
      "loss": 0.9604,
      "mean_token_accuracy": 0.7591623067855835,
      "num_tokens": 1156661.0,
      "step": 34
    },
    {
      "epoch": 0.0064995357474466105,
      "grad_norm": 12.939928709709134,
      "learning_rate": 2.103960396039604e-08,
      "loss": 0.999,
      "mean_token_accuracy": 0.7444314956665039,
      "num_tokens": 1192349.0,
      "step": 35
    },
    {
      "epoch": 0.006685236768802228,
      "grad_norm": 12.277856448251448,
      "learning_rate": 2.1658415841584157e-08,
      "loss": 0.9364,
      "mean_token_accuracy": 0.7628231048583984,
      "num_tokens": 1230371.0,
      "step": 36
    },
    {
      "epoch": 0.006870937790157846,
      "grad_norm": 14.517229382940373,
      "learning_rate": 2.2277227722772275e-08,
      "loss": 0.93,
      "mean_token_accuracy": 0.7673546075820923,
      "num_tokens": 1261575.0,
      "step": 37
    },
    {
      "epoch": 0.007056638811513464,
      "grad_norm": 13.354578882226331,
      "learning_rate": 2.2896039603960393e-08,
      "loss": 0.9586,
      "mean_token_accuracy": 0.7547529935836792,
      "num_tokens": 1296218.0,
      "step": 38
    },
    {
      "epoch": 0.0072423398328690805,
      "grad_norm": 14.741110920550586,
      "learning_rate": 2.3514851485148515e-08,
      "loss": 1.0195,
      "mean_token_accuracy": 0.7458503842353821,
      "num_tokens": 1327288.0,
      "step": 39
    },
    {
      "epoch": 0.007428040854224698,
      "grad_norm": 14.131879342940096,
      "learning_rate": 2.4133663366336632e-08,
      "loss": 0.9729,
      "mean_token_accuracy": 0.756169855594635,
      "num_tokens": 1360024.0,
      "step": 40
    },
    {
      "epoch": 0.007613741875580316,
      "grad_norm": 15.235813179894754,
      "learning_rate": 2.4752475247524754e-08,
      "loss": 0.9642,
      "mean_token_accuracy": 0.757699191570282,
      "num_tokens": 1390116.0,
      "step": 41
    },
    {
      "epoch": 0.007799442896935933,
      "grad_norm": 14.638883791183087,
      "learning_rate": 2.537128712871287e-08,
      "loss": 0.9768,
      "mean_token_accuracy": 0.7508838176727295,
      "num_tokens": 1420760.0,
      "step": 42
    },
    {
      "epoch": 0.00798514391829155,
      "grad_norm": 11.855410152090798,
      "learning_rate": 2.5990099009900986e-08,
      "loss": 0.9494,
      "mean_token_accuracy": 0.7575046420097351,
      "num_tokens": 1458170.0,
      "step": 43
    },
    {
      "epoch": 0.008170844939647167,
      "grad_norm": 12.738764946674928,
      "learning_rate": 2.6608910891089107e-08,
      "loss": 0.9107,
      "mean_token_accuracy": 0.7680773735046387,
      "num_tokens": 1494368.0,
      "step": 44
    },
    {
      "epoch": 0.008356545961002786,
      "grad_norm": 11.600917689835264,
      "learning_rate": 2.7227722772277225e-08,
      "loss": 0.9539,
      "mean_token_accuracy": 0.7522016763687134,
      "num_tokens": 1533234.0,
      "step": 45
    },
    {
      "epoch": 0.008542246982358403,
      "grad_norm": 16.186840366508804,
      "learning_rate": 2.7846534653465347e-08,
      "loss": 1.0352,
      "mean_token_accuracy": 0.7428628206253052,
      "num_tokens": 1559774.0,
      "step": 46
    },
    {
      "epoch": 0.00872794800371402,
      "grad_norm": 12.636400840289744,
      "learning_rate": 2.8465346534653465e-08,
      "loss": 0.9889,
      "mean_token_accuracy": 0.7463810443878174,
      "num_tokens": 1595582.0,
      "step": 47
    },
    {
      "epoch": 0.008913649025069638,
      "grad_norm": 11.342340819273529,
      "learning_rate": 2.9084158415841586e-08,
      "loss": 0.9268,
      "mean_token_accuracy": 0.7623275518417358,
      "num_tokens": 1635030.0,
      "step": 48
    },
    {
      "epoch": 0.009099350046425255,
      "grad_norm": 13.822156033990087,
      "learning_rate": 2.97029702970297e-08,
      "loss": 0.937,
      "mean_token_accuracy": 0.761754035949707,
      "num_tokens": 1666965.0,
      "step": 49
    },
    {
      "epoch": 0.009285051067780872,
      "grad_norm": 10.989830208731032,
      "learning_rate": 3.032178217821782e-08,
      "loss": 0.8772,
      "mean_token_accuracy": 0.7765008807182312,
      "num_tokens": 1707071.0,
      "step": 50
    },
    {
      "epoch": 0.00947075208913649,
      "grad_norm": 13.049264323212519,
      "learning_rate": 3.0940594059405936e-08,
      "loss": 0.9293,
      "mean_token_accuracy": 0.7608314156532288,
      "num_tokens": 1740394.0,
      "step": 51
    },
    {
      "epoch": 0.009656453110492107,
      "grad_norm": 10.866121754472816,
      "learning_rate": 3.155940594059406e-08,
      "loss": 0.8912,
      "mean_token_accuracy": 0.7686104774475098,
      "num_tokens": 1780628.0,
      "step": 52
    },
    {
      "epoch": 0.009842154131847726,
      "grad_norm": 11.344005974305139,
      "learning_rate": 3.217821782178218e-08,
      "loss": 0.9797,
      "mean_token_accuracy": 0.7478541731834412,
      "num_tokens": 1820586.0,
      "step": 53
    },
    {
      "epoch": 0.010027855153203343,
      "grad_norm": 12.550324852034942,
      "learning_rate": 3.27970297029703e-08,
      "loss": 0.9702,
      "mean_token_accuracy": 0.753760576248169,
      "num_tokens": 1854850.0,
      "step": 54
    },
    {
      "epoch": 0.01021355617455896,
      "grad_norm": 12.139800534579882,
      "learning_rate": 3.3415841584158415e-08,
      "loss": 0.9939,
      "mean_token_accuracy": 0.7439626455307007,
      "num_tokens": 1891313.0,
      "step": 55
    },
    {
      "epoch": 0.010399257195914578,
      "grad_norm": 14.112122453327757,
      "learning_rate": 3.403465346534653e-08,
      "loss": 0.9882,
      "mean_token_accuracy": 0.7482730150222778,
      "num_tokens": 1922239.0,
      "step": 56
    },
    {
      "epoch": 0.010584958217270195,
      "grad_norm": 13.155649588045778,
      "learning_rate": 3.465346534653466e-08,
      "loss": 1.0406,
      "mean_token_accuracy": 0.7345909476280212,
      "num_tokens": 1955260.0,
      "step": 57
    },
    {
      "epoch": 0.010770659238625812,
      "grad_norm": 10.039111876722092,
      "learning_rate": 3.527227722772277e-08,
      "loss": 0.9645,
      "mean_token_accuracy": 0.7492612600326538,
      "num_tokens": 1993927.0,
      "step": 58
    },
    {
      "epoch": 0.01095636025998143,
      "grad_norm": 9.380232195326512,
      "learning_rate": 3.5891089108910886e-08,
      "loss": 0.9629,
      "mean_token_accuracy": 0.7480544447898865,
      "num_tokens": 2033997.0,
      "step": 59
    },
    {
      "epoch": 0.011142061281337047,
      "grad_norm": 9.407358884401134,
      "learning_rate": 3.6509900990099004e-08,
      "loss": 0.9318,
      "mean_token_accuracy": 0.7551981806755066,
      "num_tokens": 2073290.0,
      "step": 60
    },
    {
      "epoch": 0.011327762302692664,
      "grad_norm": 10.328079372712539,
      "learning_rate": 3.712871287128712e-08,
      "loss": 0.915,
      "mean_token_accuracy": 0.7627955675125122,
      "num_tokens": 2108115.0,
      "step": 61
    },
    {
      "epoch": 0.011513463324048283,
      "grad_norm": 10.284154297136496,
      "learning_rate": 3.774752475247525e-08,
      "loss": 0.9261,
      "mean_token_accuracy": 0.7617260217666626,
      "num_tokens": 2143910.0,
      "step": 62
    },
    {
      "epoch": 0.0116991643454039,
      "grad_norm": 10.241508498041933,
      "learning_rate": 3.8366336633663365e-08,
      "loss": 0.9352,
      "mean_token_accuracy": 0.7566842436790466,
      "num_tokens": 2179285.0,
      "step": 63
    },
    {
      "epoch": 0.011884865366759517,
      "grad_norm": 9.407593339923595,
      "learning_rate": 3.898514851485148e-08,
      "loss": 0.9344,
      "mean_token_accuracy": 0.7582117319107056,
      "num_tokens": 2218253.0,
      "step": 64
    },
    {
      "epoch": 0.012070566388115135,
      "grad_norm": 9.786654725408988,
      "learning_rate": 3.96039603960396e-08,
      "loss": 0.8626,
      "mean_token_accuracy": 0.7779631614685059,
      "num_tokens": 2253926.0,
      "step": 65
    },
    {
      "epoch": 0.012256267409470752,
      "grad_norm": 9.088393513716817,
      "learning_rate": 4.0222772277227725e-08,
      "loss": 0.9649,
      "mean_token_accuracy": 0.7524577975273132,
      "num_tokens": 2294271.0,
      "step": 66
    },
    {
      "epoch": 0.012441968430826369,
      "grad_norm": 11.471854962327837,
      "learning_rate": 4.084158415841584e-08,
      "loss": 1.016,
      "mean_token_accuracy": 0.7446283102035522,
      "num_tokens": 2325434.0,
      "step": 67
    },
    {
      "epoch": 0.012627669452181987,
      "grad_norm": 9.191005604081251,
      "learning_rate": 4.146039603960396e-08,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7745319604873657,
      "num_tokens": 2364117.0,
      "step": 68
    },
    {
      "epoch": 0.012813370473537604,
      "grad_norm": 10.123242294422033,
      "learning_rate": 4.207920792079208e-08,
      "loss": 0.9367,
      "mean_token_accuracy": 0.7569961547851562,
      "num_tokens": 2400386.0,
      "step": 69
    },
    {
      "epoch": 0.012999071494893221,
      "grad_norm": 11.006168101113746,
      "learning_rate": 4.26980198019802e-08,
      "loss": 0.9382,
      "mean_token_accuracy": 0.7575139999389648,
      "num_tokens": 2431739.0,
      "step": 70
    },
    {
      "epoch": 0.01318477251624884,
      "grad_norm": 13.019507506629346,
      "learning_rate": 4.3316831683168315e-08,
      "loss": 1.0788,
      "mean_token_accuracy": 0.719434380531311,
      "num_tokens": 2457480.0,
      "step": 71
    },
    {
      "epoch": 0.013370473537604457,
      "grad_norm": 8.79324106046586,
      "learning_rate": 4.393564356435643e-08,
      "loss": 0.9144,
      "mean_token_accuracy": 0.7607080340385437,
      "num_tokens": 2497800.0,
      "step": 72
    },
    {
      "epoch": 0.013556174558960075,
      "grad_norm": 9.579099754119628,
      "learning_rate": 4.455445544554455e-08,
      "loss": 0.9025,
      "mean_token_accuracy": 0.765103816986084,
      "num_tokens": 2534518.0,
      "step": 73
    },
    {
      "epoch": 0.013741875580315692,
      "grad_norm": 11.169235319101793,
      "learning_rate": 4.517326732673267e-08,
      "loss": 1.0073,
      "mean_token_accuracy": 0.7413194179534912,
      "num_tokens": 2565670.0,
      "step": 74
    },
    {
      "epoch": 0.013927576601671309,
      "grad_norm": 10.200970074852036,
      "learning_rate": 4.5792079207920787e-08,
      "loss": 0.9051,
      "mean_token_accuracy": 0.7656915187835693,
      "num_tokens": 2600879.0,
      "step": 75
    },
    {
      "epoch": 0.014113277623026927,
      "grad_norm": 10.387644275221003,
      "learning_rate": 4.641089108910891e-08,
      "loss": 0.8967,
      "mean_token_accuracy": 0.7674727439880371,
      "num_tokens": 2636001.0,
      "step": 76
    },
    {
      "epoch": 0.014298978644382544,
      "grad_norm": 10.362539656510378,
      "learning_rate": 4.702970297029703e-08,
      "loss": 0.8738,
      "mean_token_accuracy": 0.7735746502876282,
      "num_tokens": 2673387.0,
      "step": 77
    },
    {
      "epoch": 0.014484679665738161,
      "grad_norm": 10.251987597140023,
      "learning_rate": 4.764851485148515e-08,
      "loss": 0.9267,
      "mean_token_accuracy": 0.7578580975532532,
      "num_tokens": 2709586.0,
      "step": 78
    },
    {
      "epoch": 0.01467038068709378,
      "grad_norm": 9.933103495924902,
      "learning_rate": 4.8267326732673265e-08,
      "loss": 0.8726,
      "mean_token_accuracy": 0.7699190974235535,
      "num_tokens": 2748440.0,
      "step": 79
    },
    {
      "epoch": 0.014856081708449397,
      "grad_norm": 9.14983241712133,
      "learning_rate": 4.888613861386138e-08,
      "loss": 0.9265,
      "mean_token_accuracy": 0.753955066204071,
      "num_tokens": 2786813.0,
      "step": 80
    },
    {
      "epoch": 0.015041782729805013,
      "grad_norm": 7.9962638247308915,
      "learning_rate": 4.950495049504951e-08,
      "loss": 0.8518,
      "mean_token_accuracy": 0.77618408203125,
      "num_tokens": 2825615.0,
      "step": 81
    },
    {
      "epoch": 0.015227483751160632,
      "grad_norm": 8.95603682802859,
      "learning_rate": 5.0123762376237625e-08,
      "loss": 0.9596,
      "mean_token_accuracy": 0.7497905492782593,
      "num_tokens": 2857793.0,
      "step": 82
    },
    {
      "epoch": 0.015413184772516249,
      "grad_norm": 7.759185161707874,
      "learning_rate": 5.074257425742574e-08,
      "loss": 0.8847,
      "mean_token_accuracy": 0.7624572515487671,
      "num_tokens": 2892665.0,
      "step": 83
    },
    {
      "epoch": 0.015598885793871866,
      "grad_norm": 7.993588673804027,
      "learning_rate": 5.1361386138613855e-08,
      "loss": 0.8869,
      "mean_token_accuracy": 0.7667703628540039,
      "num_tokens": 2923545.0,
      "step": 84
    },
    {
      "epoch": 0.015784586815227482,
      "grad_norm": 7.239491482611295,
      "learning_rate": 5.198019801980197e-08,
      "loss": 0.9259,
      "mean_token_accuracy": 0.7535645365715027,
      "num_tokens": 2959632.0,
      "step": 85
    },
    {
      "epoch": 0.0159702878365831,
      "grad_norm": 6.550799052948793,
      "learning_rate": 5.25990099009901e-08,
      "loss": 0.9761,
      "mean_token_accuracy": 0.745664119720459,
      "num_tokens": 3001509.0,
      "step": 86
    },
    {
      "epoch": 0.01615598885793872,
      "grad_norm": 8.096517348266879,
      "learning_rate": 5.3217821782178215e-08,
      "loss": 0.9324,
      "mean_token_accuracy": 0.7500585913658142,
      "num_tokens": 3031714.0,
      "step": 87
    },
    {
      "epoch": 0.016341689879294335,
      "grad_norm": 6.553533278152012,
      "learning_rate": 5.383663366336633e-08,
      "loss": 0.912,
      "mean_token_accuracy": 0.7599869966506958,
      "num_tokens": 3070511.0,
      "step": 88
    },
    {
      "epoch": 0.016527390900649953,
      "grad_norm": 6.722619525762078,
      "learning_rate": 5.445544554455445e-08,
      "loss": 0.9131,
      "mean_token_accuracy": 0.7561958432197571,
      "num_tokens": 3107702.0,
      "step": 89
    },
    {
      "epoch": 0.016713091922005572,
      "grad_norm": 6.5333969634311915,
      "learning_rate": 5.5074257425742575e-08,
      "loss": 0.8862,
      "mean_token_accuracy": 0.7667458057403564,
      "num_tokens": 3145678.0,
      "step": 90
    },
    {
      "epoch": 0.016898792943361187,
      "grad_norm": 7.082968539907543,
      "learning_rate": 5.569306930693069e-08,
      "loss": 0.956,
      "mean_token_accuracy": 0.7479490041732788,
      "num_tokens": 3178634.0,
      "step": 91
    },
    {
      "epoch": 0.017084493964716806,
      "grad_norm": 6.243346407160964,
      "learning_rate": 5.631188118811881e-08,
      "loss": 0.8981,
      "mean_token_accuracy": 0.7625658512115479,
      "num_tokens": 3219719.0,
      "step": 92
    },
    {
      "epoch": 0.017270194986072424,
      "grad_norm": 7.785916363539197,
      "learning_rate": 5.693069306930693e-08,
      "loss": 0.9541,
      "mean_token_accuracy": 0.7477093935012817,
      "num_tokens": 3249018.0,
      "step": 93
    },
    {
      "epoch": 0.01745589600742804,
      "grad_norm": 6.962078941397661,
      "learning_rate": 5.754950495049505e-08,
      "loss": 0.9368,
      "mean_token_accuracy": 0.7529942989349365,
      "num_tokens": 3283393.0,
      "step": 94
    },
    {
      "epoch": 0.017641597028783658,
      "grad_norm": 6.094846567965009,
      "learning_rate": 5.816831683168317e-08,
      "loss": 0.9178,
      "mean_token_accuracy": 0.753459095954895,
      "num_tokens": 3322406.0,
      "step": 95
    },
    {
      "epoch": 0.017827298050139277,
      "grad_norm": 6.938740032881691,
      "learning_rate": 5.878712871287128e-08,
      "loss": 0.9668,
      "mean_token_accuracy": 0.7471771240234375,
      "num_tokens": 3354738.0,
      "step": 96
    },
    {
      "epoch": 0.01801299907149489,
      "grad_norm": 7.08071404348821,
      "learning_rate": 5.94059405940594e-08,
      "loss": 0.8979,
      "mean_token_accuracy": 0.7683308124542236,
      "num_tokens": 3384517.0,
      "step": 97
    },
    {
      "epoch": 0.01819870009285051,
      "grad_norm": 6.6493963184490985,
      "learning_rate": 6.002475247524753e-08,
      "loss": 0.9516,
      "mean_token_accuracy": 0.7507437467575073,
      "num_tokens": 3417660.0,
      "step": 98
    },
    {
      "epoch": 0.01838440111420613,
      "grad_norm": 6.082321361036866,
      "learning_rate": 6.064356435643564e-08,
      "loss": 0.8507,
      "mean_token_accuracy": 0.775923490524292,
      "num_tokens": 3454003.0,
      "step": 99
    },
    {
      "epoch": 0.018570102135561744,
      "grad_norm": 5.2856967061177125,
      "learning_rate": 6.126237623762376e-08,
      "loss": 0.9046,
      "mean_token_accuracy": 0.7614622116088867,
      "num_tokens": 3497588.0,
      "step": 100
    },
    {
      "epoch": 0.018755803156917362,
      "grad_norm": 5.880226425412394,
      "learning_rate": 6.188118811881187e-08,
      "loss": 0.8303,
      "mean_token_accuracy": 0.7787228226661682,
      "num_tokens": 3534090.0,
      "step": 101
    },
    {
      "epoch": 0.01894150417827298,
      "grad_norm": 6.463488963008283,
      "learning_rate": 6.25e-08,
      "loss": 0.843,
      "mean_token_accuracy": 0.7747796773910522,
      "num_tokens": 3566704.0,
      "step": 102
    },
    {
      "epoch": 0.0191272051996286,
      "grad_norm": 6.127875461892286,
      "learning_rate": 6.311881188118812e-08,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7680296897888184,
      "num_tokens": 3600273.0,
      "step": 103
    },
    {
      "epoch": 0.019312906220984215,
      "grad_norm": 5.683438339864242,
      "learning_rate": 6.373762376237623e-08,
      "loss": 0.8757,
      "mean_token_accuracy": 0.7674393057823181,
      "num_tokens": 3637319.0,
      "step": 104
    },
    {
      "epoch": 0.019498607242339833,
      "grad_norm": 5.9500057718178185,
      "learning_rate": 6.435643564356436e-08,
      "loss": 0.9097,
      "mean_token_accuracy": 0.756157636642456,
      "num_tokens": 3672382.0,
      "step": 105
    },
    {
      "epoch": 0.019684308263695452,
      "grad_norm": 5.567308510819414,
      "learning_rate": 6.497524752475247e-08,
      "loss": 0.8296,
      "mean_token_accuracy": 0.7772420048713684,
      "num_tokens": 3709764.0,
      "step": 106
    },
    {
      "epoch": 0.019870009285051067,
      "grad_norm": 5.727744403971997,
      "learning_rate": 6.55940594059406e-08,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7819373607635498,
      "num_tokens": 3745037.0,
      "step": 107
    },
    {
      "epoch": 0.020055710306406686,
      "grad_norm": 5.878595120180238,
      "learning_rate": 6.621287128712872e-08,
      "loss": 0.8464,
      "mean_token_accuracy": 0.7728503942489624,
      "num_tokens": 3778162.0,
      "step": 108
    },
    {
      "epoch": 0.020241411327762304,
      "grad_norm": 5.860084973185647,
      "learning_rate": 6.683168316831683e-08,
      "loss": 0.9455,
      "mean_token_accuracy": 0.7461979389190674,
      "num_tokens": 3812990.0,
      "step": 109
    },
    {
      "epoch": 0.02042711234911792,
      "grad_norm": 5.154064468403025,
      "learning_rate": 6.745049504950495e-08,
      "loss": 0.8233,
      "mean_token_accuracy": 0.7768290638923645,
      "num_tokens": 3852326.0,
      "step": 110
    },
    {
      "epoch": 0.020612813370473538,
      "grad_norm": 5.193106590044432,
      "learning_rate": 6.806930693069307e-08,
      "loss": 0.7861,
      "mean_token_accuracy": 0.7880715131759644,
      "num_tokens": 3889422.0,
      "step": 111
    },
    {
      "epoch": 0.020798514391829157,
      "grad_norm": 5.246481923181212,
      "learning_rate": 6.868811881188119e-08,
      "loss": 0.8788,
      "mean_token_accuracy": 0.7650119066238403,
      "num_tokens": 3927364.0,
      "step": 112
    },
    {
      "epoch": 0.02098421541318477,
      "grad_norm": 5.94539480889029,
      "learning_rate": 6.930693069306931e-08,
      "loss": 0.8699,
      "mean_token_accuracy": 0.769237756729126,
      "num_tokens": 3959173.0,
      "step": 113
    },
    {
      "epoch": 0.02116991643454039,
      "grad_norm": 6.56949079847115,
      "learning_rate": 6.992574257425743e-08,
      "loss": 0.9007,
      "mean_token_accuracy": 0.7528756260871887,
      "num_tokens": 3986570.0,
      "step": 114
    },
    {
      "epoch": 0.02135561745589601,
      "grad_norm": 6.256155760890374,
      "learning_rate": 7.054455445544554e-08,
      "loss": 0.9031,
      "mean_token_accuracy": 0.7615695595741272,
      "num_tokens": 4016796.0,
      "step": 115
    },
    {
      "epoch": 0.021541318477251624,
      "grad_norm": 5.076897995294835,
      "learning_rate": 7.116336633663365e-08,
      "loss": 0.7986,
      "mean_token_accuracy": 0.7819298505783081,
      "num_tokens": 4055253.0,
      "step": 116
    },
    {
      "epoch": 0.021727019498607242,
      "grad_norm": 6.461593776483173,
      "learning_rate": 7.178217821782177e-08,
      "loss": 0.8603,
      "mean_token_accuracy": 0.7688264846801758,
      "num_tokens": 4083980.0,
      "step": 117
    },
    {
      "epoch": 0.02191272051996286,
      "grad_norm": 4.954238352610565,
      "learning_rate": 7.24009900990099e-08,
      "loss": 0.8739,
      "mean_token_accuracy": 0.758368968963623,
      "num_tokens": 4124162.0,
      "step": 118
    },
    {
      "epoch": 0.022098421541318476,
      "grad_norm": 5.601155724537467,
      "learning_rate": 7.301980198019801e-08,
      "loss": 0.8334,
      "mean_token_accuracy": 0.7729095220565796,
      "num_tokens": 4158554.0,
      "step": 119
    },
    {
      "epoch": 0.022284122562674095,
      "grad_norm": 5.518230681391863,
      "learning_rate": 7.363861386138613e-08,
      "loss": 0.8468,
      "mean_token_accuracy": 0.770266592502594,
      "num_tokens": 4192916.0,
      "step": 120
    },
    {
      "epoch": 0.022469823584029713,
      "grad_norm": 5.070712340799337,
      "learning_rate": 7.425742574257424e-08,
      "loss": 0.8157,
      "mean_token_accuracy": 0.7744628190994263,
      "num_tokens": 4230560.0,
      "step": 121
    },
    {
      "epoch": 0.02265552460538533,
      "grad_norm": 5.528576976122172,
      "learning_rate": 7.487623762376237e-08,
      "loss": 0.7989,
      "mean_token_accuracy": 0.7836339473724365,
      "num_tokens": 4263870.0,
      "step": 122
    },
    {
      "epoch": 0.022841225626740947,
      "grad_norm": 6.23956694379196,
      "learning_rate": 7.54950495049505e-08,
      "loss": 0.8605,
      "mean_token_accuracy": 0.7644728422164917,
      "num_tokens": 4293146.0,
      "step": 123
    },
    {
      "epoch": 0.023026926648096566,
      "grad_norm": 6.673765878788607,
      "learning_rate": 7.61138613861386e-08,
      "loss": 0.8442,
      "mean_token_accuracy": 0.7713407278060913,
      "num_tokens": 4320246.0,
      "step": 124
    },
    {
      "epoch": 0.02321262766945218,
      "grad_norm": 5.432532672235324,
      "learning_rate": 7.673267326732673e-08,
      "loss": 0.7834,
      "mean_token_accuracy": 0.7868940830230713,
      "num_tokens": 4353301.0,
      "step": 125
    },
    {
      "epoch": 0.0233983286908078,
      "grad_norm": 5.9053252846149755,
      "learning_rate": 7.735148514851484e-08,
      "loss": 0.8539,
      "mean_token_accuracy": 0.763981819152832,
      "num_tokens": 4385583.0,
      "step": 126
    },
    {
      "epoch": 0.023584029712163418,
      "grad_norm": 4.9629095139208586,
      "learning_rate": 7.797029702970297e-08,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7826667428016663,
      "num_tokens": 4424998.0,
      "step": 127
    },
    {
      "epoch": 0.023769730733519033,
      "grad_norm": 5.406564862416191,
      "learning_rate": 7.858910891089109e-08,
      "loss": 0.771,
      "mean_token_accuracy": 0.7840637564659119,
      "num_tokens": 4458855.0,
      "step": 128
    },
    {
      "epoch": 0.02395543175487465,
      "grad_norm": 5.500969814678387,
      "learning_rate": 7.92079207920792e-08,
      "loss": 0.82,
      "mean_token_accuracy": 0.7723861336708069,
      "num_tokens": 4492524.0,
      "step": 129
    },
    {
      "epoch": 0.02414113277623027,
      "grad_norm": 5.408321435097889,
      "learning_rate": 7.982673267326733e-08,
      "loss": 0.8062,
      "mean_token_accuracy": 0.7759542465209961,
      "num_tokens": 4527717.0,
      "step": 130
    },
    {
      "epoch": 0.024326833797585885,
      "grad_norm": 5.353280584141635,
      "learning_rate": 8.044554455445545e-08,
      "loss": 0.8332,
      "mean_token_accuracy": 0.7696431875228882,
      "num_tokens": 4565657.0,
      "step": 131
    },
    {
      "epoch": 0.024512534818941504,
      "grad_norm": 6.6246142856608214,
      "learning_rate": 8.106435643564356e-08,
      "loss": 0.8551,
      "mean_token_accuracy": 0.7675604224205017,
      "num_tokens": 4594066.0,
      "step": 132
    },
    {
      "epoch": 0.024698235840297122,
      "grad_norm": 5.982479576867681,
      "learning_rate": 8.168316831683169e-08,
      "loss": 0.7784,
      "mean_token_accuracy": 0.7833530902862549,
      "num_tokens": 4627963.0,
      "step": 133
    },
    {
      "epoch": 0.024883936861652738,
      "grad_norm": 6.48794394262648,
      "learning_rate": 8.23019801980198e-08,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7949097156524658,
      "num_tokens": 4659272.0,
      "step": 134
    },
    {
      "epoch": 0.025069637883008356,
      "grad_norm": 5.515635416178758,
      "learning_rate": 8.292079207920792e-08,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7839304208755493,
      "num_tokens": 4698830.0,
      "step": 135
    },
    {
      "epoch": 0.025255338904363975,
      "grad_norm": 5.722892106404616,
      "learning_rate": 8.353960396039605e-08,
      "loss": 0.7596,
      "mean_token_accuracy": 0.7882273197174072,
      "num_tokens": 4735010.0,
      "step": 136
    },
    {
      "epoch": 0.02544103992571959,
      "grad_norm": 5.271075803950178,
      "learning_rate": 8.415841584158416e-08,
      "loss": 0.7798,
      "mean_token_accuracy": 0.7808873653411865,
      "num_tokens": 4775578.0,
      "step": 137
    },
    {
      "epoch": 0.02562674094707521,
      "grad_norm": 5.805326405348337,
      "learning_rate": 8.477722772277228e-08,
      "loss": 0.7585,
      "mean_token_accuracy": 0.7888742685317993,
      "num_tokens": 4811209.0,
      "step": 138
    },
    {
      "epoch": 0.025812441968430827,
      "grad_norm": 5.184542739315937,
      "learning_rate": 8.53960396039604e-08,
      "loss": 0.7477,
      "mean_token_accuracy": 0.7879731059074402,
      "num_tokens": 4850726.0,
      "step": 139
    },
    {
      "epoch": 0.025998142989786442,
      "grad_norm": 4.984785255031237,
      "learning_rate": 8.60148514851485e-08,
      "loss": 0.8547,
      "mean_token_accuracy": 0.7633746266365051,
      "num_tokens": 4890560.0,
      "step": 140
    },
    {
      "epoch": 0.02618384401114206,
      "grad_norm": 5.615666164354212,
      "learning_rate": 8.663366336633663e-08,
      "loss": 0.7997,
      "mean_token_accuracy": 0.777504026889801,
      "num_tokens": 4923240.0,
      "step": 141
    },
    {
      "epoch": 0.02636954503249768,
      "grad_norm": 5.516274967105238,
      "learning_rate": 8.725247524752474e-08,
      "loss": 0.8384,
      "mean_token_accuracy": 0.7631633281707764,
      "num_tokens": 4955726.0,
      "step": 142
    },
    {
      "epoch": 0.026555246053853298,
      "grad_norm": 4.442864033954862,
      "learning_rate": 8.787128712871287e-08,
      "loss": 0.7437,
      "mean_token_accuracy": 0.7916553616523743,
      "num_tokens": 4998864.0,
      "step": 143
    },
    {
      "epoch": 0.026740947075208913,
      "grad_norm": 5.63457673713007,
      "learning_rate": 8.849009900990098e-08,
      "loss": 0.7533,
      "mean_token_accuracy": 0.7873605489730835,
      "num_tokens": 5032480.0,
      "step": 144
    },
    {
      "epoch": 0.02692664809656453,
      "grad_norm": 4.520414770537282,
      "learning_rate": 8.91089108910891e-08,
      "loss": 0.813,
      "mean_token_accuracy": 0.7656753659248352,
      "num_tokens": 5074953.0,
      "step": 145
    },
    {
      "epoch": 0.02711234911792015,
      "grad_norm": 4.58815715817732,
      "learning_rate": 8.972772277227723e-08,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7762000560760498,
      "num_tokens": 5110997.0,
      "step": 146
    },
    {
      "epoch": 0.027298050139275765,
      "grad_norm": 3.9493517314749464,
      "learning_rate": 9.034653465346534e-08,
      "loss": 0.7193,
      "mean_token_accuracy": 0.7955378293991089,
      "num_tokens": 5147696.0,
      "step": 147
    },
    {
      "epoch": 0.027483751160631384,
      "grad_norm": 4.648242327991254,
      "learning_rate": 9.096534653465346e-08,
      "loss": 0.7874,
      "mean_token_accuracy": 0.7812566757202148,
      "num_tokens": 5178479.0,
      "step": 148
    },
    {
      "epoch": 0.027669452181987002,
      "grad_norm": 4.2108080012318005,
      "learning_rate": 9.158415841584157e-08,
      "loss": 0.7899,
      "mean_token_accuracy": 0.7743276357650757,
      "num_tokens": 5213205.0,
      "step": 149
    },
    {
      "epoch": 0.027855153203342618,
      "grad_norm": 4.541396898465088,
      "learning_rate": 9.22029702970297e-08,
      "loss": 0.7553,
      "mean_token_accuracy": 0.7832079529762268,
      "num_tokens": 5247898.0,
      "step": 150
    },
    {
      "epoch": 0.028040854224698236,
      "grad_norm": 4.0278950472737565,
      "learning_rate": 9.282178217821782e-08,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7862650752067566,
      "num_tokens": 5279428.0,
      "step": 151
    },
    {
      "epoch": 0.028226555246053855,
      "grad_norm": 3.79475664737156,
      "learning_rate": 9.344059405940593e-08,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7809543013572693,
      "num_tokens": 5313776.0,
      "step": 152
    },
    {
      "epoch": 0.02841225626740947,
      "grad_norm": 3.564629178279863,
      "learning_rate": 9.405940594059406e-08,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7917947173118591,
      "num_tokens": 5348651.0,
      "step": 153
    },
    {
      "epoch": 0.02859795728876509,
      "grad_norm": 3.8627349811535985,
      "learning_rate": 9.467821782178217e-08,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7849693298339844,
      "num_tokens": 5379716.0,
      "step": 154
    },
    {
      "epoch": 0.028783658310120707,
      "grad_norm": 3.632139015221107,
      "learning_rate": 9.52970297029703e-08,
      "loss": 0.722,
      "mean_token_accuracy": 0.800014853477478,
      "num_tokens": 5412184.0,
      "step": 155
    },
    {
      "epoch": 0.028969359331476322,
      "grad_norm": 3.1118383065268307,
      "learning_rate": 9.591584158415842e-08,
      "loss": 0.6821,
      "mean_token_accuracy": 0.8042863011360168,
      "num_tokens": 5450998.0,
      "step": 156
    },
    {
      "epoch": 0.02915506035283194,
      "grad_norm": 3.132643763140073,
      "learning_rate": 9.653465346534653e-08,
      "loss": 0.6175,
      "mean_token_accuracy": 0.823994517326355,
      "num_tokens": 5488094.0,
      "step": 157
    },
    {
      "epoch": 0.02934076137418756,
      "grad_norm": 3.612650540928513,
      "learning_rate": 9.715346534653465e-08,
      "loss": 0.724,
      "mean_token_accuracy": 0.7975143194198608,
      "num_tokens": 5517363.0,
      "step": 158
    },
    {
      "epoch": 0.029526462395543174,
      "grad_norm": 3.2706889616251464,
      "learning_rate": 9.777227722772277e-08,
      "loss": 0.744,
      "mean_token_accuracy": 0.7862457036972046,
      "num_tokens": 5552162.0,
      "step": 159
    },
    {
      "epoch": 0.029712163416898793,
      "grad_norm": 3.3220157955396523,
      "learning_rate": 9.839108910891089e-08,
      "loss": 0.7657,
      "mean_token_accuracy": 0.780240535736084,
      "num_tokens": 5591493.0,
      "step": 160
    },
    {
      "epoch": 0.02989786443825441,
      "grad_norm": 3.268604151389279,
      "learning_rate": 9.900990099009901e-08,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7854490876197815,
      "num_tokens": 5625160.0,
      "step": 161
    },
    {
      "epoch": 0.030083565459610027,
      "grad_norm": 3.0488304726310402,
      "learning_rate": 9.962871287128713e-08,
      "loss": 0.72,
      "mean_token_accuracy": 0.7900708913803101,
      "num_tokens": 5666291.0,
      "step": 162
    },
    {
      "epoch": 0.030269266480965645,
      "grad_norm": 3.439134409203031,
      "learning_rate": 1.0024752475247525e-07,
      "loss": 0.7582,
      "mean_token_accuracy": 0.7866404056549072,
      "num_tokens": 5697361.0,
      "step": 163
    },
    {
      "epoch": 0.030454967502321264,
      "grad_norm": 2.9443524354428847,
      "learning_rate": 1.0086633663366336e-07,
      "loss": 0.6621,
      "mean_token_accuracy": 0.810836136341095,
      "num_tokens": 5730822.0,
      "step": 164
    },
    {
      "epoch": 0.03064066852367688,
      "grad_norm": 3.2918060670470957,
      "learning_rate": 1.0148514851485149e-07,
      "loss": 0.731,
      "mean_token_accuracy": 0.7916313409805298,
      "num_tokens": 5760834.0,
      "step": 165
    },
    {
      "epoch": 0.030826369545032498,
      "grad_norm": 3.266731284923963,
      "learning_rate": 1.021039603960396e-07,
      "loss": 0.707,
      "mean_token_accuracy": 0.7984671592712402,
      "num_tokens": 5790912.0,
      "step": 166
    },
    {
      "epoch": 0.031012070566388116,
      "grad_norm": 3.1965643826846675,
      "learning_rate": 1.0272277227722771e-07,
      "loss": 0.7367,
      "mean_token_accuracy": 0.783413290977478,
      "num_tokens": 5823525.0,
      "step": 167
    },
    {
      "epoch": 0.03119777158774373,
      "grad_norm": 3.0550024857252227,
      "learning_rate": 1.0334158415841583e-07,
      "loss": 0.7627,
      "mean_token_accuracy": 0.7790672779083252,
      "num_tokens": 5858800.0,
      "step": 168
    },
    {
      "epoch": 0.03138347260909935,
      "grad_norm": 2.9403156632238128,
      "learning_rate": 1.0396039603960394e-07,
      "loss": 0.6511,
      "mean_token_accuracy": 0.8085392713546753,
      "num_tokens": 5895732.0,
      "step": 169
    },
    {
      "epoch": 0.031569173630454965,
      "grad_norm": 3.3446258168596756,
      "learning_rate": 1.0457920792079207e-07,
      "loss": 0.701,
      "mean_token_accuracy": 0.7925446033477783,
      "num_tokens": 5926054.0,
      "step": 170
    },
    {
      "epoch": 0.03175487465181059,
      "grad_norm": 3.216356521868566,
      "learning_rate": 1.051980198019802e-07,
      "loss": 0.655,
      "mean_token_accuracy": 0.8097130060195923,
      "num_tokens": 5954471.0,
      "step": 171
    },
    {
      "epoch": 0.0319405756731662,
      "grad_norm": 2.9552516015720642,
      "learning_rate": 1.058168316831683e-07,
      "loss": 0.6806,
      "mean_token_accuracy": 0.8016306757926941,
      "num_tokens": 5990144.0,
      "step": 172
    },
    {
      "epoch": 0.03212627669452182,
      "grad_norm": 2.990020861917187,
      "learning_rate": 1.0643564356435643e-07,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7881941199302673,
      "num_tokens": 6023364.0,
      "step": 173
    },
    {
      "epoch": 0.03231197771587744,
      "grad_norm": 2.550293309855238,
      "learning_rate": 1.0705445544554454e-07,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7999500632286072,
      "num_tokens": 6063363.0,
      "step": 174
    },
    {
      "epoch": 0.032497678737233054,
      "grad_norm": 2.7677825509869614,
      "learning_rate": 1.0767326732673267e-07,
      "loss": 0.7341,
      "mean_token_accuracy": 0.7844390273094177,
      "num_tokens": 6097720.0,
      "step": 175
    },
    {
      "epoch": 0.03268337975858867,
      "grad_norm": 3.055123118340999,
      "learning_rate": 1.0829207920792079e-07,
      "loss": 0.6497,
      "mean_token_accuracy": 0.8088997602462769,
      "num_tokens": 6126380.0,
      "step": 176
    },
    {
      "epoch": 0.03286908077994429,
      "grad_norm": 2.6453965748719925,
      "learning_rate": 1.089108910891089e-07,
      "loss": 0.6826,
      "mean_token_accuracy": 0.8067673444747925,
      "num_tokens": 6160802.0,
      "step": 177
    },
    {
      "epoch": 0.03305478180129991,
      "grad_norm": 2.78182153234213,
      "learning_rate": 1.0952970297029703e-07,
      "loss": 0.6734,
      "mean_token_accuracy": 0.8008120656013489,
      "num_tokens": 6196787.0,
      "step": 178
    },
    {
      "epoch": 0.03324048282265552,
      "grad_norm": 2.977810367937884,
      "learning_rate": 1.1014851485148515e-07,
      "loss": 0.6588,
      "mean_token_accuracy": 0.8067640066146851,
      "num_tokens": 6235639.0,
      "step": 179
    },
    {
      "epoch": 0.033426183844011144,
      "grad_norm": 3.2058623875089194,
      "learning_rate": 1.1076732673267326e-07,
      "loss": 0.6769,
      "mean_token_accuracy": 0.8047792315483093,
      "num_tokens": 6272681.0,
      "step": 180
    },
    {
      "epoch": 0.03361188486536676,
      "grad_norm": 2.609746676108621,
      "learning_rate": 1.1138613861386139e-07,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8120006322860718,
      "num_tokens": 6312919.0,
      "step": 181
    },
    {
      "epoch": 0.033797585886722374,
      "grad_norm": 2.724257129965755,
      "learning_rate": 1.120049504950495e-07,
      "loss": 0.638,
      "mean_token_accuracy": 0.8077677488327026,
      "num_tokens": 6349072.0,
      "step": 182
    },
    {
      "epoch": 0.033983286908077996,
      "grad_norm": 2.7735145660639864,
      "learning_rate": 1.1262376237623762e-07,
      "loss": 0.7045,
      "mean_token_accuracy": 0.794433057308197,
      "num_tokens": 6386609.0,
      "step": 183
    },
    {
      "epoch": 0.03416898792943361,
      "grad_norm": 3.2780973391691997,
      "learning_rate": 1.1324257425742575e-07,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7933633923530579,
      "num_tokens": 6423190.0,
      "step": 184
    },
    {
      "epoch": 0.034354688950789226,
      "grad_norm": 3.044557606394739,
      "learning_rate": 1.1386138613861386e-07,
      "loss": 0.6622,
      "mean_token_accuracy": 0.8040051460266113,
      "num_tokens": 6454702.0,
      "step": 185
    },
    {
      "epoch": 0.03454038997214485,
      "grad_norm": 3.2460819670118664,
      "learning_rate": 1.1448019801980198e-07,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7910071611404419,
      "num_tokens": 6485314.0,
      "step": 186
    },
    {
      "epoch": 0.034726090993500464,
      "grad_norm": 2.8468123997356014,
      "learning_rate": 1.150990099009901e-07,
      "loss": 0.6638,
      "mean_token_accuracy": 0.8015689253807068,
      "num_tokens": 6515733.0,
      "step": 187
    },
    {
      "epoch": 0.03491179201485608,
      "grad_norm": 2.8239512662076054,
      "learning_rate": 1.1571782178217822e-07,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8103127479553223,
      "num_tokens": 6551639.0,
      "step": 188
    },
    {
      "epoch": 0.0350974930362117,
      "grad_norm": 3.056774301386138,
      "learning_rate": 1.1633663366336634e-07,
      "loss": 0.7202,
      "mean_token_accuracy": 0.784325122833252,
      "num_tokens": 6584095.0,
      "step": 189
    },
    {
      "epoch": 0.035283194057567316,
      "grad_norm": 2.8096757584910526,
      "learning_rate": 1.1695544554455445e-07,
      "loss": 0.7186,
      "mean_token_accuracy": 0.7904573082923889,
      "num_tokens": 6621273.0,
      "step": 190
    },
    {
      "epoch": 0.03546889507892293,
      "grad_norm": 2.9401648976360244,
      "learning_rate": 1.1757425742574257e-07,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8042293787002563,
      "num_tokens": 6657366.0,
      "step": 191
    },
    {
      "epoch": 0.03565459610027855,
      "grad_norm": 2.723243882537313,
      "learning_rate": 1.1819306930693068e-07,
      "loss": 0.6739,
      "mean_token_accuracy": 0.8023632168769836,
      "num_tokens": 6693038.0,
      "step": 192
    },
    {
      "epoch": 0.03584029712163417,
      "grad_norm": 3.2171784905181426,
      "learning_rate": 1.188118811881188e-07,
      "loss": 0.59,
      "mean_token_accuracy": 0.8216841220855713,
      "num_tokens": 6722413.0,
      "step": 193
    },
    {
      "epoch": 0.03602599814298978,
      "grad_norm": 2.986711990732557,
      "learning_rate": 1.194306930693069e-07,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7936371564865112,
      "num_tokens": 6754621.0,
      "step": 194
    },
    {
      "epoch": 0.036211699164345405,
      "grad_norm": 3.234656532594235,
      "learning_rate": 1.2004950495049505e-07,
      "loss": 0.6631,
      "mean_token_accuracy": 0.8035776019096375,
      "num_tokens": 6785023.0,
      "step": 195
    },
    {
      "epoch": 0.03639740018570102,
      "grad_norm": 2.979540023174373,
      "learning_rate": 1.2066831683168316e-07,
      "loss": 0.6176,
      "mean_token_accuracy": 0.815335214138031,
      "num_tokens": 6826890.0,
      "step": 196
    },
    {
      "epoch": 0.036583101207056636,
      "grad_norm": 2.8230060141577664,
      "learning_rate": 1.2128712871287127e-07,
      "loss": 0.639,
      "mean_token_accuracy": 0.8108487725257874,
      "num_tokens": 6862307.0,
      "step": 197
    },
    {
      "epoch": 0.03676880222841226,
      "grad_norm": 3.3388923913610897,
      "learning_rate": 1.2190594059405938e-07,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7933251261711121,
      "num_tokens": 6901526.0,
      "step": 198
    },
    {
      "epoch": 0.03695450324976787,
      "grad_norm": 3.538662746251179,
      "learning_rate": 1.2252475247524752e-07,
      "loss": 0.6609,
      "mean_token_accuracy": 0.8073746562004089,
      "num_tokens": 6933443.0,
      "step": 199
    },
    {
      "epoch": 0.03714020427112349,
      "grad_norm": 3.1037020097528587,
      "learning_rate": 1.2314356435643563e-07,
      "loss": 0.6491,
      "mean_token_accuracy": 0.805438756942749,
      "num_tokens": 6966360.0,
      "step": 200
    },
    {
      "epoch": 0.03732590529247911,
      "grad_norm": 3.0239211062377076,
      "learning_rate": 1.2376237623762375e-07,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8041214942932129,
      "num_tokens": 7003222.0,
      "step": 201
    },
    {
      "epoch": 0.037511606313834725,
      "grad_norm": 3.1355165384928023,
      "learning_rate": 1.2438118811881188e-07,
      "loss": 0.6625,
      "mean_token_accuracy": 0.804086446762085,
      "num_tokens": 7032736.0,
      "step": 202
    },
    {
      "epoch": 0.03769730733519034,
      "grad_norm": 2.872135930065527,
      "learning_rate": 1.25e-07,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8038128614425659,
      "num_tokens": 7066384.0,
      "step": 203
    },
    {
      "epoch": 0.03788300835654596,
      "grad_norm": 2.8921999380250627,
      "learning_rate": 1.256188118811881e-07,
      "loss": 0.6778,
      "mean_token_accuracy": 0.8008349537849426,
      "num_tokens": 7097817.0,
      "step": 204
    },
    {
      "epoch": 0.03806870937790158,
      "grad_norm": 3.137599044922498,
      "learning_rate": 1.2623762376237624e-07,
      "loss": 0.6462,
      "mean_token_accuracy": 0.8060303926467896,
      "num_tokens": 7134254.0,
      "step": 205
    },
    {
      "epoch": 0.0382544103992572,
      "grad_norm": 2.7260144587616515,
      "learning_rate": 1.2685643564356435e-07,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8103823661804199,
      "num_tokens": 7167738.0,
      "step": 206
    },
    {
      "epoch": 0.038440111420612814,
      "grad_norm": 2.9199069768032753,
      "learning_rate": 1.2747524752475247e-07,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8040699362754822,
      "num_tokens": 7200642.0,
      "step": 207
    },
    {
      "epoch": 0.03862581244196843,
      "grad_norm": 3.012702452353903,
      "learning_rate": 1.280940594059406e-07,
      "loss": 0.678,
      "mean_token_accuracy": 0.8015775680541992,
      "num_tokens": 7240926.0,
      "step": 208
    },
    {
      "epoch": 0.03881151346332405,
      "grad_norm": 3.505868976936711,
      "learning_rate": 1.2871287128712872e-07,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8109216690063477,
      "num_tokens": 7276916.0,
      "step": 209
    },
    {
      "epoch": 0.03899721448467967,
      "grad_norm": 3.2887696801161574,
      "learning_rate": 1.2933168316831683e-07,
      "loss": 0.669,
      "mean_token_accuracy": 0.8005737066268921,
      "num_tokens": 7313285.0,
      "step": 210
    },
    {
      "epoch": 0.03918291550603528,
      "grad_norm": 3.3013337049900455,
      "learning_rate": 1.2995049504950494e-07,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8077247142791748,
      "num_tokens": 7352430.0,
      "step": 211
    },
    {
      "epoch": 0.039368616527390904,
      "grad_norm": 2.523221639272339,
      "learning_rate": 1.3056930693069308e-07,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8132665753364563,
      "num_tokens": 7390230.0,
      "step": 212
    },
    {
      "epoch": 0.03955431754874652,
      "grad_norm": 3.769874167075641,
      "learning_rate": 1.311881188118812e-07,
      "loss": 0.6363,
      "mean_token_accuracy": 0.8123612403869629,
      "num_tokens": 7420955.0,
      "step": 213
    },
    {
      "epoch": 0.039740018570102134,
      "grad_norm": 3.8484075782804203,
      "learning_rate": 1.318069306930693e-07,
      "loss": 0.6838,
      "mean_token_accuracy": 0.7999054193496704,
      "num_tokens": 7454657.0,
      "step": 214
    },
    {
      "epoch": 0.039925719591457756,
      "grad_norm": 4.314117357985662,
      "learning_rate": 1.3242574257425744e-07,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8096458911895752,
      "num_tokens": 7493183.0,
      "step": 215
    },
    {
      "epoch": 0.04011142061281337,
      "grad_norm": 3.273750151878999,
      "learning_rate": 1.3304455445544555e-07,
      "loss": 0.6061,
      "mean_token_accuracy": 0.8164314031600952,
      "num_tokens": 7524139.0,
      "step": 216
    },
    {
      "epoch": 0.040297121634168986,
      "grad_norm": 2.597286042730483,
      "learning_rate": 1.3366336633663366e-07,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8045079112052917,
      "num_tokens": 7564348.0,
      "step": 217
    },
    {
      "epoch": 0.04048282265552461,
      "grad_norm": 3.0655900871835717,
      "learning_rate": 1.342821782178218e-07,
      "loss": 0.612,
      "mean_token_accuracy": 0.8140318989753723,
      "num_tokens": 7599340.0,
      "step": 218
    },
    {
      "epoch": 0.040668523676880224,
      "grad_norm": 3.646756337355025,
      "learning_rate": 1.349009900990099e-07,
      "loss": 0.7029,
      "mean_token_accuracy": 0.790885865688324,
      "num_tokens": 7634309.0,
      "step": 219
    },
    {
      "epoch": 0.04085422469823584,
      "grad_norm": 3.053409408648923,
      "learning_rate": 1.3551980198019802e-07,
      "loss": 0.6647,
      "mean_token_accuracy": 0.8053398728370667,
      "num_tokens": 7664530.0,
      "step": 220
    },
    {
      "epoch": 0.04103992571959146,
      "grad_norm": 3.069130618887164,
      "learning_rate": 1.3613861386138613e-07,
      "loss": 0.6536,
      "mean_token_accuracy": 0.7993614673614502,
      "num_tokens": 7697623.0,
      "step": 221
    },
    {
      "epoch": 0.041225626740947076,
      "grad_norm": 3.2596009442862828,
      "learning_rate": 1.3675742574257427e-07,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8061733245849609,
      "num_tokens": 7729848.0,
      "step": 222
    },
    {
      "epoch": 0.04141132776230269,
      "grad_norm": 4.108837209226566,
      "learning_rate": 1.3737623762376238e-07,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7946263551712036,
      "num_tokens": 7763723.0,
      "step": 223
    },
    {
      "epoch": 0.04159702878365831,
      "grad_norm": 3.023887729902987,
      "learning_rate": 1.379950495049505e-07,
      "loss": 0.683,
      "mean_token_accuracy": 0.8006438612937927,
      "num_tokens": 7797014.0,
      "step": 224
    },
    {
      "epoch": 0.04178272980501393,
      "grad_norm": 3.478409331724564,
      "learning_rate": 1.3861386138613863e-07,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8108468055725098,
      "num_tokens": 7832092.0,
      "step": 225
    },
    {
      "epoch": 0.04196843082636954,
      "grad_norm": 2.9044800687578722,
      "learning_rate": 1.3923267326732674e-07,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8175848722457886,
      "num_tokens": 7866744.0,
      "step": 226
    },
    {
      "epoch": 0.042154131847725165,
      "grad_norm": 3.2452634251732984,
      "learning_rate": 1.3985148514851485e-07,
      "loss": 0.656,
      "mean_token_accuracy": 0.8082076907157898,
      "num_tokens": 7894406.0,
      "step": 227
    },
    {
      "epoch": 0.04233983286908078,
      "grad_norm": 3.3598227599776638,
      "learning_rate": 1.40470297029703e-07,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8188878893852234,
      "num_tokens": 7928193.0,
      "step": 228
    },
    {
      "epoch": 0.042525533890436396,
      "grad_norm": 4.20440560354158,
      "learning_rate": 1.4108910891089107e-07,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8160256743431091,
      "num_tokens": 7962638.0,
      "step": 229
    },
    {
      "epoch": 0.04271123491179202,
      "grad_norm": 2.7285219973880332,
      "learning_rate": 1.4170792079207919e-07,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8200306296348572,
      "num_tokens": 7994316.0,
      "step": 230
    },
    {
      "epoch": 0.04289693593314763,
      "grad_norm": 2.8855415255514476,
      "learning_rate": 1.423267326732673e-07,
      "loss": 0.6616,
      "mean_token_accuracy": 0.8007534742355347,
      "num_tokens": 8030937.0,
      "step": 231
    },
    {
      "epoch": 0.04308263695450325,
      "grad_norm": 3.13274659469829,
      "learning_rate": 1.4294554455445543e-07,
      "loss": 0.6596,
      "mean_token_accuracy": 0.8035494089126587,
      "num_tokens": 8064380.0,
      "step": 232
    },
    {
      "epoch": 0.04326833797585887,
      "grad_norm": 3.092610731007641,
      "learning_rate": 1.4356435643564355e-07,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8144494295120239,
      "num_tokens": 8098018.0,
      "step": 233
    },
    {
      "epoch": 0.043454038997214485,
      "grad_norm": 3.1394691824835625,
      "learning_rate": 1.4418316831683166e-07,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8134656548500061,
      "num_tokens": 8128605.0,
      "step": 234
    },
    {
      "epoch": 0.0436397400185701,
      "grad_norm": 2.995565288124012,
      "learning_rate": 1.448019801980198e-07,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8190504908561707,
      "num_tokens": 8164120.0,
      "step": 235
    },
    {
      "epoch": 0.04382544103992572,
      "grad_norm": 3.2849793499294653,
      "learning_rate": 1.454207920792079e-07,
      "loss": 0.695,
      "mean_token_accuracy": 0.7927666902542114,
      "num_tokens": 8197162.0,
      "step": 236
    },
    {
      "epoch": 0.04401114206128134,
      "grad_norm": 2.7857092977870312,
      "learning_rate": 1.4603960396039602e-07,
      "loss": 0.616,
      "mean_token_accuracy": 0.8128618597984314,
      "num_tokens": 8233698.0,
      "step": 237
    },
    {
      "epoch": 0.04419684308263695,
      "grad_norm": 4.662897152569415,
      "learning_rate": 1.4665841584158416e-07,
      "loss": 0.6954,
      "mean_token_accuracy": 0.7933856844902039,
      "num_tokens": 8264581.0,
      "step": 238
    },
    {
      "epoch": 0.044382544103992574,
      "grad_norm": 3.0472172309102845,
      "learning_rate": 1.4727722772277227e-07,
      "loss": 0.6584,
      "mean_token_accuracy": 0.803810715675354,
      "num_tokens": 8306046.0,
      "step": 239
    },
    {
      "epoch": 0.04456824512534819,
      "grad_norm": 2.6564322427406517,
      "learning_rate": 1.4789603960396038e-07,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8199098110198975,
      "num_tokens": 8342186.0,
      "step": 240
    },
    {
      "epoch": 0.044753946146703805,
      "grad_norm": 3.0670480917755913,
      "learning_rate": 1.485148514851485e-07,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7944965362548828,
      "num_tokens": 8377192.0,
      "step": 241
    },
    {
      "epoch": 0.04493964716805943,
      "grad_norm": 3.0979056381707775,
      "learning_rate": 1.4913366336633663e-07,
      "loss": 0.5868,
      "mean_token_accuracy": 0.824616014957428,
      "num_tokens": 8416801.0,
      "step": 242
    },
    {
      "epoch": 0.04512534818941504,
      "grad_norm": 2.9722671992342677,
      "learning_rate": 1.4975247524752474e-07,
      "loss": 0.6952,
      "mean_token_accuracy": 0.7966541647911072,
      "num_tokens": 8450058.0,
      "step": 243
    },
    {
      "epoch": 0.04531104921077066,
      "grad_norm": 3.7010740591728943,
      "learning_rate": 1.5037128712871285e-07,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8158451318740845,
      "num_tokens": 8480762.0,
      "step": 244
    },
    {
      "epoch": 0.04549675023212628,
      "grad_norm": 4.128487267547499,
      "learning_rate": 1.50990099009901e-07,
      "loss": 0.6491,
      "mean_token_accuracy": 0.8045532703399658,
      "num_tokens": 8511030.0,
      "step": 245
    },
    {
      "epoch": 0.045682451253481894,
      "grad_norm": 3.283303889140853,
      "learning_rate": 1.516089108910891e-07,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8241854906082153,
      "num_tokens": 8539856.0,
      "step": 246
    },
    {
      "epoch": 0.04586815227483751,
      "grad_norm": 3.4273467444877066,
      "learning_rate": 1.522277227722772e-07,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8193657398223877,
      "num_tokens": 8574185.0,
      "step": 247
    },
    {
      "epoch": 0.04605385329619313,
      "grad_norm": 3.4559932965860796,
      "learning_rate": 1.5284653465346535e-07,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8048984408378601,
      "num_tokens": 8612530.0,
      "step": 248
    },
    {
      "epoch": 0.046239554317548746,
      "grad_norm": 4.65844295613652,
      "learning_rate": 1.5346534653465346e-07,
      "loss": 0.6664,
      "mean_token_accuracy": 0.7997694611549377,
      "num_tokens": 8643738.0,
      "step": 249
    },
    {
      "epoch": 0.04642525533890436,
      "grad_norm": 2.396787592742213,
      "learning_rate": 1.5408415841584157e-07,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8122334480285645,
      "num_tokens": 8682181.0,
      "step": 250
    },
    {
      "epoch": 0.046610956360259984,
      "grad_norm": 3.1446687173957115,
      "learning_rate": 1.5470297029702968e-07,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8158860206604004,
      "num_tokens": 8721668.0,
      "step": 251
    },
    {
      "epoch": 0.0467966573816156,
      "grad_norm": 4.447227030216557,
      "learning_rate": 1.5532178217821782e-07,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8138111233711243,
      "num_tokens": 8756892.0,
      "step": 252
    },
    {
      "epoch": 0.046982358402971214,
      "grad_norm": 2.705381295833133,
      "learning_rate": 1.5594059405940593e-07,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8069366216659546,
      "num_tokens": 8788941.0,
      "step": 253
    },
    {
      "epoch": 0.047168059424326836,
      "grad_norm": 3.509882614034052,
      "learning_rate": 1.5655940594059404e-07,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8306561708450317,
      "num_tokens": 8824607.0,
      "step": 254
    },
    {
      "epoch": 0.04735376044568245,
      "grad_norm": 3.8925168392296507,
      "learning_rate": 1.5717821782178218e-07,
      "loss": 0.679,
      "mean_token_accuracy": 0.7951763868331909,
      "num_tokens": 8853792.0,
      "step": 255
    },
    {
      "epoch": 0.047539461467038066,
      "grad_norm": 3.415257636785023,
      "learning_rate": 1.577970297029703e-07,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8288118839263916,
      "num_tokens": 8890655.0,
      "step": 256
    },
    {
      "epoch": 0.04772516248839369,
      "grad_norm": 3.312803819240143,
      "learning_rate": 1.584158415841584e-07,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8197290301322937,
      "num_tokens": 8927364.0,
      "step": 257
    },
    {
      "epoch": 0.0479108635097493,
      "grad_norm": 3.1995943947064074,
      "learning_rate": 1.5903465346534654e-07,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8129701614379883,
      "num_tokens": 8959257.0,
      "step": 258
    },
    {
      "epoch": 0.04809656453110492,
      "grad_norm": 3.064003035043539,
      "learning_rate": 1.5965346534653465e-07,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8212976455688477,
      "num_tokens": 8994215.0,
      "step": 259
    },
    {
      "epoch": 0.04828226555246054,
      "grad_norm": 3.1695495295579503,
      "learning_rate": 1.6027227722772276e-07,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8169553875923157,
      "num_tokens": 9027898.0,
      "step": 260
    },
    {
      "epoch": 0.048467966573816156,
      "grad_norm": 3.622269635776547,
      "learning_rate": 1.608910891089109e-07,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7947181463241577,
      "num_tokens": 9064873.0,
      "step": 261
    },
    {
      "epoch": 0.04865366759517177,
      "grad_norm": 3.180860340624658,
      "learning_rate": 1.61509900990099e-07,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8215672969818115,
      "num_tokens": 9098340.0,
      "step": 262
    },
    {
      "epoch": 0.04883936861652739,
      "grad_norm": 2.5296690995258535,
      "learning_rate": 1.6212871287128712e-07,
      "loss": 0.6147,
      "mean_token_accuracy": 0.817341148853302,
      "num_tokens": 9143151.0,
      "step": 263
    },
    {
      "epoch": 0.04902506963788301,
      "grad_norm": 3.4245092024882897,
      "learning_rate": 1.6274752475247523e-07,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8109055757522583,
      "num_tokens": 9184696.0,
      "step": 264
    },
    {
      "epoch": 0.04921077065923862,
      "grad_norm": 3.1901762939745613,
      "learning_rate": 1.6336633663366337e-07,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8271304368972778,
      "num_tokens": 9219815.0,
      "step": 265
    },
    {
      "epoch": 0.049396471680594245,
      "grad_norm": 3.2332251232191154,
      "learning_rate": 1.6398514851485148e-07,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8105804324150085,
      "num_tokens": 9251723.0,
      "step": 266
    },
    {
      "epoch": 0.04958217270194986,
      "grad_norm": 2.9982784464821486,
      "learning_rate": 1.646039603960396e-07,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8145431280136108,
      "num_tokens": 9287127.0,
      "step": 267
    },
    {
      "epoch": 0.049767873723305475,
      "grad_norm": 3.5111468510325223,
      "learning_rate": 1.6522277227722773e-07,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8066461086273193,
      "num_tokens": 9322292.0,
      "step": 268
    },
    {
      "epoch": 0.0499535747446611,
      "grad_norm": 3.0519331473419538,
      "learning_rate": 1.6584158415841584e-07,
      "loss": 0.57,
      "mean_token_accuracy": 0.8253995776176453,
      "num_tokens": 9357627.0,
      "step": 269
    },
    {
      "epoch": 0.05013927576601671,
      "grad_norm": 2.9746116287373714,
      "learning_rate": 1.6646039603960396e-07,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7877525091171265,
      "num_tokens": 9391365.0,
      "step": 270
    },
    {
      "epoch": 0.05032497678737233,
      "grad_norm": 3.5760452430066825,
      "learning_rate": 1.670792079207921e-07,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8107407093048096,
      "num_tokens": 9429075.0,
      "step": 271
    },
    {
      "epoch": 0.05051067780872795,
      "grad_norm": 2.6553493352419077,
      "learning_rate": 1.676980198019802e-07,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8116996884346008,
      "num_tokens": 9466607.0,
      "step": 272
    },
    {
      "epoch": 0.050696378830083565,
      "grad_norm": 2.464944727899741,
      "learning_rate": 1.6831683168316832e-07,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8077381253242493,
      "num_tokens": 9506614.0,
      "step": 273
    },
    {
      "epoch": 0.05088207985143918,
      "grad_norm": 3.2669146176511656,
      "learning_rate": 1.6893564356435643e-07,
      "loss": 0.6131,
      "mean_token_accuracy": 0.81419837474823,
      "num_tokens": 9538934.0,
      "step": 274
    },
    {
      "epoch": 0.0510677808727948,
      "grad_norm": 5.4918220964427125,
      "learning_rate": 1.6955445544554456e-07,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8395284414291382,
      "num_tokens": 9569260.0,
      "step": 275
    },
    {
      "epoch": 0.05125348189415042,
      "grad_norm": 3.441335793337787,
      "learning_rate": 1.7017326732673268e-07,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8087707757949829,
      "num_tokens": 9604394.0,
      "step": 276
    },
    {
      "epoch": 0.05143918291550603,
      "grad_norm": 3.2066477063425647,
      "learning_rate": 1.707920792079208e-07,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8119814991950989,
      "num_tokens": 9636884.0,
      "step": 277
    },
    {
      "epoch": 0.051624883936861654,
      "grad_norm": 3.0412315308829907,
      "learning_rate": 1.7141089108910893e-07,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8143631219863892,
      "num_tokens": 9667282.0,
      "step": 278
    },
    {
      "epoch": 0.05181058495821727,
      "grad_norm": 3.7325613571466554,
      "learning_rate": 1.72029702970297e-07,
      "loss": 0.564,
      "mean_token_accuracy": 0.8316852450370789,
      "num_tokens": 9711027.0,
      "step": 279
    },
    {
      "epoch": 0.051996285979572884,
      "grad_norm": 4.026421193150852,
      "learning_rate": 1.7264851485148512e-07,
      "loss": 0.6649,
      "mean_token_accuracy": 0.8015772104263306,
      "num_tokens": 9736870.0,
      "step": 280
    },
    {
      "epoch": 0.052181987000928506,
      "grad_norm": 2.862844772270604,
      "learning_rate": 1.7326732673267326e-07,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8303108215332031,
      "num_tokens": 9776265.0,
      "step": 281
    },
    {
      "epoch": 0.05236768802228412,
      "grad_norm": 3.0650944684194563,
      "learning_rate": 1.7388613861386137e-07,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8115471601486206,
      "num_tokens": 9811246.0,
      "step": 282
    },
    {
      "epoch": 0.05255338904363974,
      "grad_norm": 4.254070549126628,
      "learning_rate": 1.7450495049504948e-07,
      "loss": 0.6527,
      "mean_token_accuracy": 0.8018236756324768,
      "num_tokens": 9845902.0,
      "step": 283
    },
    {
      "epoch": 0.05273909006499536,
      "grad_norm": 4.314913439575709,
      "learning_rate": 1.751237623762376e-07,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8096219897270203,
      "num_tokens": 9881663.0,
      "step": 284
    },
    {
      "epoch": 0.052924791086350974,
      "grad_norm": 4.0928403911047315,
      "learning_rate": 1.7574257425742573e-07,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8283827304840088,
      "num_tokens": 9915312.0,
      "step": 285
    },
    {
      "epoch": 0.053110492107706596,
      "grad_norm": 3.091485892721379,
      "learning_rate": 1.7636138613861384e-07,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8120876550674438,
      "num_tokens": 9951654.0,
      "step": 286
    },
    {
      "epoch": 0.05329619312906221,
      "grad_norm": 4.111609341929972,
      "learning_rate": 1.7698019801980195e-07,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7985255718231201,
      "num_tokens": 9980315.0,
      "step": 287
    },
    {
      "epoch": 0.053481894150417826,
      "grad_norm": 3.544136695104063,
      "learning_rate": 1.775990099009901e-07,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8187897205352783,
      "num_tokens": 10010205.0,
      "step": 288
    },
    {
      "epoch": 0.05366759517177345,
      "grad_norm": 3.36627482554172,
      "learning_rate": 1.782178217821782e-07,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8053486347198486,
      "num_tokens": 10047699.0,
      "step": 289
    },
    {
      "epoch": 0.05385329619312906,
      "grad_norm": 2.953628216583972,
      "learning_rate": 1.7883663366336631e-07,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8128364682197571,
      "num_tokens": 10081424.0,
      "step": 290
    },
    {
      "epoch": 0.05403899721448468,
      "grad_norm": 3.7272374087707223,
      "learning_rate": 1.7945544554455445e-07,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8099569082260132,
      "num_tokens": 10107512.0,
      "step": 291
    },
    {
      "epoch": 0.0542246982358403,
      "grad_norm": 2.544329510969401,
      "learning_rate": 1.8007425742574256e-07,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8243383765220642,
      "num_tokens": 10142445.0,
      "step": 292
    },
    {
      "epoch": 0.054410399257195916,
      "grad_norm": 3.494986419553735,
      "learning_rate": 1.8069306930693067e-07,
      "loss": 0.5752,
      "mean_token_accuracy": 0.8271157741546631,
      "num_tokens": 10175352.0,
      "step": 293
    },
    {
      "epoch": 0.05459610027855153,
      "grad_norm": 3.840433918680005,
      "learning_rate": 1.8131188118811879e-07,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8090956211090088,
      "num_tokens": 10207901.0,
      "step": 294
    },
    {
      "epoch": 0.05478180129990715,
      "grad_norm": 3.426204244130869,
      "learning_rate": 1.8193069306930692e-07,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8146840333938599,
      "num_tokens": 10242678.0,
      "step": 295
    },
    {
      "epoch": 0.05496750232126277,
      "grad_norm": 3.5298361949759025,
      "learning_rate": 1.8254950495049503e-07,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8217049241065979,
      "num_tokens": 10274792.0,
      "step": 296
    },
    {
      "epoch": 0.05515320334261838,
      "grad_norm": 2.502909695314435,
      "learning_rate": 1.8316831683168315e-07,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8200063109397888,
      "num_tokens": 10310215.0,
      "step": 297
    },
    {
      "epoch": 0.055338904363974005,
      "grad_norm": 2.6060500838551803,
      "learning_rate": 1.8378712871287128e-07,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8269908428192139,
      "num_tokens": 10351759.0,
      "step": 298
    },
    {
      "epoch": 0.05552460538532962,
      "grad_norm": 2.945986475751601,
      "learning_rate": 1.844059405940594e-07,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8076837658882141,
      "num_tokens": 10389138.0,
      "step": 299
    },
    {
      "epoch": 0.055710306406685235,
      "grad_norm": 3.379718251645374,
      "learning_rate": 1.850247524752475e-07,
      "loss": 0.5923,
      "mean_token_accuracy": 0.821546196937561,
      "num_tokens": 10418848.0,
      "step": 300
    },
    {
      "epoch": 0.05589600742804086,
      "grad_norm": 2.782141241593914,
      "learning_rate": 1.8564356435643564e-07,
      "loss": 0.5508,
      "mean_token_accuracy": 0.829756498336792,
      "num_tokens": 10456309.0,
      "step": 301
    },
    {
      "epoch": 0.05608170844939647,
      "grad_norm": 2.40166795174131,
      "learning_rate": 1.8626237623762376e-07,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8351085782051086,
      "num_tokens": 10495199.0,
      "step": 302
    },
    {
      "epoch": 0.05626740947075209,
      "grad_norm": 2.729758820342595,
      "learning_rate": 1.8688118811881187e-07,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8168598413467407,
      "num_tokens": 10535335.0,
      "step": 303
    },
    {
      "epoch": 0.05645311049210771,
      "grad_norm": 2.748432492682482,
      "learning_rate": 1.875e-07,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8188561797142029,
      "num_tokens": 10567627.0,
      "step": 304
    },
    {
      "epoch": 0.056638811513463325,
      "grad_norm": 3.2224388360925746,
      "learning_rate": 1.8811881188118812e-07,
      "loss": 0.6439,
      "mean_token_accuracy": 0.8032586574554443,
      "num_tokens": 10605072.0,
      "step": 305
    },
    {
      "epoch": 0.05682451253481894,
      "grad_norm": 2.629156426787523,
      "learning_rate": 1.8873762376237623e-07,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8271701335906982,
      "num_tokens": 10646202.0,
      "step": 306
    },
    {
      "epoch": 0.05701021355617456,
      "grad_norm": 2.690774566843748,
      "learning_rate": 1.8935643564356434e-07,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8388527631759644,
      "num_tokens": 10682390.0,
      "step": 307
    },
    {
      "epoch": 0.05719591457753018,
      "grad_norm": 3.2665478504747436,
      "learning_rate": 1.8997524752475248e-07,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8242343664169312,
      "num_tokens": 10718078.0,
      "step": 308
    },
    {
      "epoch": 0.05738161559888579,
      "grad_norm": 3.4307473462137033,
      "learning_rate": 1.905940594059406e-07,
      "loss": 0.5977,
      "mean_token_accuracy": 0.816184401512146,
      "num_tokens": 10751772.0,
      "step": 309
    },
    {
      "epoch": 0.057567316620241414,
      "grad_norm": 2.5433127768482118,
      "learning_rate": 1.912128712871287e-07,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7970266938209534,
      "num_tokens": 10789395.0,
      "step": 310
    },
    {
      "epoch": 0.05775301764159703,
      "grad_norm": 2.4462991381554313,
      "learning_rate": 1.9183168316831684e-07,
      "loss": 0.584,
      "mean_token_accuracy": 0.8226317763328552,
      "num_tokens": 10824015.0,
      "step": 311
    },
    {
      "epoch": 0.057938718662952644,
      "grad_norm": 3.0056439603355254,
      "learning_rate": 1.9245049504950495e-07,
      "loss": 0.6492,
      "mean_token_accuracy": 0.8066961765289307,
      "num_tokens": 10855087.0,
      "step": 312
    },
    {
      "epoch": 0.058124419684308266,
      "grad_norm": 2.2693696949477733,
      "learning_rate": 1.9306930693069306e-07,
      "loss": 0.6718,
      "mean_token_accuracy": 0.7994154691696167,
      "num_tokens": 10894738.0,
      "step": 313
    },
    {
      "epoch": 0.05831012070566388,
      "grad_norm": 2.417361191174,
      "learning_rate": 1.936881188118812e-07,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8219186067581177,
      "num_tokens": 10928478.0,
      "step": 314
    },
    {
      "epoch": 0.0584958217270195,
      "grad_norm": 3.1266004308189776,
      "learning_rate": 1.943069306930693e-07,
      "loss": 0.5648,
      "mean_token_accuracy": 0.827027440071106,
      "num_tokens": 10960994.0,
      "step": 315
    },
    {
      "epoch": 0.05868152274837512,
      "grad_norm": 2.5745954362239543,
      "learning_rate": 1.9492574257425742e-07,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8236845135688782,
      "num_tokens": 10996290.0,
      "step": 316
    },
    {
      "epoch": 0.058867223769730734,
      "grad_norm": 3.2227638324853696,
      "learning_rate": 1.9554455445544553e-07,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8242663145065308,
      "num_tokens": 11026503.0,
      "step": 317
    },
    {
      "epoch": 0.05905292479108635,
      "grad_norm": 2.5680118773224843,
      "learning_rate": 1.9616336633663367e-07,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8318099975585938,
      "num_tokens": 11062119.0,
      "step": 318
    },
    {
      "epoch": 0.05923862581244197,
      "grad_norm": 3.039950812848104,
      "learning_rate": 1.9678217821782178e-07,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8180974125862122,
      "num_tokens": 11096710.0,
      "step": 319
    },
    {
      "epoch": 0.059424326833797586,
      "grad_norm": 3.473022536106213,
      "learning_rate": 1.974009900990099e-07,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8292915225028992,
      "num_tokens": 11124604.0,
      "step": 320
    },
    {
      "epoch": 0.0596100278551532,
      "grad_norm": 2.798113876541089,
      "learning_rate": 1.9801980198019803e-07,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8038102388381958,
      "num_tokens": 11162137.0,
      "step": 321
    },
    {
      "epoch": 0.05979572887650882,
      "grad_norm": 2.521133516482071,
      "learning_rate": 1.9863861386138614e-07,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8280308246612549,
      "num_tokens": 11200158.0,
      "step": 322
    },
    {
      "epoch": 0.05998142989786444,
      "grad_norm": 2.360028804745003,
      "learning_rate": 1.9925742574257425e-07,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8117095232009888,
      "num_tokens": 11238175.0,
      "step": 323
    },
    {
      "epoch": 0.06016713091922005,
      "grad_norm": 2.5791084760305103,
      "learning_rate": 1.998762376237624e-07,
      "loss": 0.6434,
      "mean_token_accuracy": 0.8036606311798096,
      "num_tokens": 11276336.0,
      "step": 324
    },
    {
      "epoch": 0.060352831940575676,
      "grad_norm": 3.09431533372196,
      "learning_rate": 2.004950495049505e-07,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8300003409385681,
      "num_tokens": 11310789.0,
      "step": 325
    },
    {
      "epoch": 0.06053853296193129,
      "grad_norm": 2.707112839541829,
      "learning_rate": 2.011138613861386e-07,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8127666711807251,
      "num_tokens": 11346607.0,
      "step": 326
    },
    {
      "epoch": 0.060724233983286906,
      "grad_norm": 2.7584428337092595,
      "learning_rate": 2.0173267326732672e-07,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8271886110305786,
      "num_tokens": 11380506.0,
      "step": 327
    },
    {
      "epoch": 0.06090993500464253,
      "grad_norm": 2.641946585292019,
      "learning_rate": 2.0235148514851486e-07,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8148583173751831,
      "num_tokens": 11419247.0,
      "step": 328
    },
    {
      "epoch": 0.06109563602599814,
      "grad_norm": 3.171138721319672,
      "learning_rate": 2.0297029702970297e-07,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8239826560020447,
      "num_tokens": 11453229.0,
      "step": 329
    },
    {
      "epoch": 0.06128133704735376,
      "grad_norm": 2.613503270021529,
      "learning_rate": 2.0358910891089106e-07,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8207144737243652,
      "num_tokens": 11495590.0,
      "step": 330
    },
    {
      "epoch": 0.06146703806870938,
      "grad_norm": 2.4848869173777133,
      "learning_rate": 2.042079207920792e-07,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8366751670837402,
      "num_tokens": 11531576.0,
      "step": 331
    },
    {
      "epoch": 0.061652739090064995,
      "grad_norm": 3.1132008977851284,
      "learning_rate": 2.048267326732673e-07,
      "loss": 0.6211,
      "mean_token_accuracy": 0.8120222091674805,
      "num_tokens": 11565089.0,
      "step": 332
    },
    {
      "epoch": 0.06183844011142061,
      "grad_norm": 2.8499217045459453,
      "learning_rate": 2.0544554455445542e-07,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8024479746818542,
      "num_tokens": 11603682.0,
      "step": 333
    },
    {
      "epoch": 0.06202414113277623,
      "grad_norm": 2.9792600665712343,
      "learning_rate": 2.0606435643564356e-07,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8298167586326599,
      "num_tokens": 11637430.0,
      "step": 334
    },
    {
      "epoch": 0.06220984215413185,
      "grad_norm": 2.949098807806573,
      "learning_rate": 2.0668316831683167e-07,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8180980682373047,
      "num_tokens": 11672029.0,
      "step": 335
    },
    {
      "epoch": 0.06239554317548746,
      "grad_norm": 2.3771937602264264,
      "learning_rate": 2.0730198019801978e-07,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8302600979804993,
      "num_tokens": 11710899.0,
      "step": 336
    },
    {
      "epoch": 0.06258124419684308,
      "grad_norm": 3.087763425837598,
      "learning_rate": 2.079207920792079e-07,
      "loss": 0.6629,
      "mean_token_accuracy": 0.8013955950737,
      "num_tokens": 11741364.0,
      "step": 337
    },
    {
      "epoch": 0.0627669452181987,
      "grad_norm": 2.6407674765067557,
      "learning_rate": 2.0853960396039603e-07,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8266911506652832,
      "num_tokens": 11778995.0,
      "step": 338
    },
    {
      "epoch": 0.06295264623955432,
      "grad_norm": 2.537819191907084,
      "learning_rate": 2.0915841584158414e-07,
      "loss": 0.6217,
      "mean_token_accuracy": 0.816623330116272,
      "num_tokens": 11814534.0,
      "step": 339
    },
    {
      "epoch": 0.06313834726090993,
      "grad_norm": 2.9030232418116113,
      "learning_rate": 2.0977722772277225e-07,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8228936195373535,
      "num_tokens": 11852567.0,
      "step": 340
    },
    {
      "epoch": 0.06332404828226555,
      "grad_norm": 2.4499537898102046,
      "learning_rate": 2.103960396039604e-07,
      "loss": 0.557,
      "mean_token_accuracy": 0.826114296913147,
      "num_tokens": 11883675.0,
      "step": 341
    },
    {
      "epoch": 0.06350974930362117,
      "grad_norm": 2.698094819493944,
      "learning_rate": 2.110148514851485e-07,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8245593309402466,
      "num_tokens": 11911715.0,
      "step": 342
    },
    {
      "epoch": 0.06369545032497678,
      "grad_norm": 2.759019574252211,
      "learning_rate": 2.116336633663366e-07,
      "loss": 0.5949,
      "mean_token_accuracy": 0.8162686228752136,
      "num_tokens": 11939511.0,
      "step": 343
    },
    {
      "epoch": 0.0638811513463324,
      "grad_norm": 2.8104012673686585,
      "learning_rate": 2.1225247524752475e-07,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8273698091506958,
      "num_tokens": 11966888.0,
      "step": 344
    },
    {
      "epoch": 0.06406685236768803,
      "grad_norm": 2.663398339733265,
      "learning_rate": 2.1287128712871286e-07,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8269779682159424,
      "num_tokens": 12002605.0,
      "step": 345
    },
    {
      "epoch": 0.06425255338904363,
      "grad_norm": 2.2722744180595527,
      "learning_rate": 2.1349009900990097e-07,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8253858089447021,
      "num_tokens": 12040531.0,
      "step": 346
    },
    {
      "epoch": 0.06443825441039926,
      "grad_norm": 2.807897914909587,
      "learning_rate": 2.1410891089108908e-07,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8232551217079163,
      "num_tokens": 12068065.0,
      "step": 347
    },
    {
      "epoch": 0.06462395543175488,
      "grad_norm": 2.6788818243653574,
      "learning_rate": 2.1472772277227722e-07,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8380087614059448,
      "num_tokens": 12095855.0,
      "step": 348
    },
    {
      "epoch": 0.06480965645311049,
      "grad_norm": 2.695593962092814,
      "learning_rate": 2.1534653465346533e-07,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8180829882621765,
      "num_tokens": 12131763.0,
      "step": 349
    },
    {
      "epoch": 0.06499535747446611,
      "grad_norm": 2.45235800964424,
      "learning_rate": 2.1596534653465344e-07,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8369266986846924,
      "num_tokens": 12165520.0,
      "step": 350
    },
    {
      "epoch": 0.06518105849582173,
      "grad_norm": 2.413748526941341,
      "learning_rate": 2.1658415841584158e-07,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8338136672973633,
      "num_tokens": 12207809.0,
      "step": 351
    },
    {
      "epoch": 0.06536675951717734,
      "grad_norm": 2.3994471458918283,
      "learning_rate": 2.172029702970297e-07,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8123438358306885,
      "num_tokens": 12245276.0,
      "step": 352
    },
    {
      "epoch": 0.06555246053853296,
      "grad_norm": 2.506516259830913,
      "learning_rate": 2.178217821782178e-07,
      "loss": 0.5622,
      "mean_token_accuracy": 0.827427864074707,
      "num_tokens": 12280260.0,
      "step": 353
    },
    {
      "epoch": 0.06573816155988858,
      "grad_norm": 2.285459869722334,
      "learning_rate": 2.1844059405940594e-07,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8395938873291016,
      "num_tokens": 12315877.0,
      "step": 354
    },
    {
      "epoch": 0.06592386258124419,
      "grad_norm": 2.528810538198623,
      "learning_rate": 2.1905940594059405e-07,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8378129005432129,
      "num_tokens": 12356493.0,
      "step": 355
    },
    {
      "epoch": 0.06610956360259981,
      "grad_norm": 2.6641264158080786,
      "learning_rate": 2.1967821782178216e-07,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8311558961868286,
      "num_tokens": 12390178.0,
      "step": 356
    },
    {
      "epoch": 0.06629526462395544,
      "grad_norm": 2.4809304526795777,
      "learning_rate": 2.202970297029703e-07,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8277936577796936,
      "num_tokens": 12419168.0,
      "step": 357
    },
    {
      "epoch": 0.06648096564531104,
      "grad_norm": 2.1530667336265688,
      "learning_rate": 2.209158415841584e-07,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8137904405593872,
      "num_tokens": 12459064.0,
      "step": 358
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 2.6181603704808984,
      "learning_rate": 2.2153465346534652e-07,
      "loss": 0.5919,
      "mean_token_accuracy": 0.82004714012146,
      "num_tokens": 12489141.0,
      "step": 359
    },
    {
      "epoch": 0.06685236768802229,
      "grad_norm": 2.745246136986932,
      "learning_rate": 2.2215346534653464e-07,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8288031220436096,
      "num_tokens": 12515497.0,
      "step": 360
    },
    {
      "epoch": 0.0670380687093779,
      "grad_norm": 2.1611531166071676,
      "learning_rate": 2.2277227722772277e-07,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8377857208251953,
      "num_tokens": 12550721.0,
      "step": 361
    },
    {
      "epoch": 0.06722376973073352,
      "grad_norm": 2.450986596642375,
      "learning_rate": 2.2339108910891088e-07,
      "loss": 0.537,
      "mean_token_accuracy": 0.8317060470581055,
      "num_tokens": 12586846.0,
      "step": 362
    },
    {
      "epoch": 0.06740947075208914,
      "grad_norm": 2.377626849318159,
      "learning_rate": 2.24009900990099e-07,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8261464834213257,
      "num_tokens": 12619318.0,
      "step": 363
    },
    {
      "epoch": 0.06759517177344475,
      "grad_norm": 2.7359030222404925,
      "learning_rate": 2.2462871287128713e-07,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8182696104049683,
      "num_tokens": 12650358.0,
      "step": 364
    },
    {
      "epoch": 0.06778087279480037,
      "grad_norm": 2.567548829897954,
      "learning_rate": 2.2524752475247524e-07,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8277580738067627,
      "num_tokens": 12682578.0,
      "step": 365
    },
    {
      "epoch": 0.06796657381615599,
      "grad_norm": 2.6219501346917484,
      "learning_rate": 2.2586633663366336e-07,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8306049108505249,
      "num_tokens": 12715983.0,
      "step": 366
    },
    {
      "epoch": 0.0681522748375116,
      "grad_norm": 2.678279293173681,
      "learning_rate": 2.264851485148515e-07,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8288493156433105,
      "num_tokens": 12750064.0,
      "step": 367
    },
    {
      "epoch": 0.06833797585886722,
      "grad_norm": 2.516599872021391,
      "learning_rate": 2.271039603960396e-07,
      "loss": 0.6065,
      "mean_token_accuracy": 0.811511754989624,
      "num_tokens": 12780766.0,
      "step": 368
    },
    {
      "epoch": 0.06852367688022284,
      "grad_norm": 2.7089528241599954,
      "learning_rate": 2.2772277227722772e-07,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8160748481750488,
      "num_tokens": 12817975.0,
      "step": 369
    },
    {
      "epoch": 0.06870937790157845,
      "grad_norm": 2.3973959086111427,
      "learning_rate": 2.2834158415841583e-07,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8181523084640503,
      "num_tokens": 12846880.0,
      "step": 370
    },
    {
      "epoch": 0.06889507892293407,
      "grad_norm": 2.413787309196271,
      "learning_rate": 2.2896039603960397e-07,
      "loss": 0.594,
      "mean_token_accuracy": 0.8218445777893066,
      "num_tokens": 12879708.0,
      "step": 371
    },
    {
      "epoch": 0.0690807799442897,
      "grad_norm": 2.655645189040967,
      "learning_rate": 2.2957920792079208e-07,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8219832181930542,
      "num_tokens": 12910247.0,
      "step": 372
    },
    {
      "epoch": 0.0692664809656453,
      "grad_norm": 2.809998977122332,
      "learning_rate": 2.301980198019802e-07,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8418210744857788,
      "num_tokens": 12941350.0,
      "step": 373
    },
    {
      "epoch": 0.06945218198700093,
      "grad_norm": 2.031325349265952,
      "learning_rate": 2.3081683168316833e-07,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8328324556350708,
      "num_tokens": 12983220.0,
      "step": 374
    },
    {
      "epoch": 0.06963788300835655,
      "grad_norm": 2.3392776326968727,
      "learning_rate": 2.3143564356435644e-07,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8273694515228271,
      "num_tokens": 13014183.0,
      "step": 375
    },
    {
      "epoch": 0.06982358402971216,
      "grad_norm": 2.1171751846986173,
      "learning_rate": 2.3205445544554455e-07,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8246588110923767,
      "num_tokens": 13053168.0,
      "step": 376
    },
    {
      "epoch": 0.07000928505106778,
      "grad_norm": 3.122247462342369,
      "learning_rate": 2.3267326732673269e-07,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8314661979675293,
      "num_tokens": 13078726.0,
      "step": 377
    },
    {
      "epoch": 0.0701949860724234,
      "grad_norm": 2.516737210899824,
      "learning_rate": 2.332920792079208e-07,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7971106767654419,
      "num_tokens": 13113892.0,
      "step": 378
    },
    {
      "epoch": 0.07038068709377901,
      "grad_norm": 2.1633213701493355,
      "learning_rate": 2.339108910891089e-07,
      "loss": 0.6219,
      "mean_token_accuracy": 0.809738039970398,
      "num_tokens": 13152363.0,
      "step": 379
    },
    {
      "epoch": 0.07056638811513463,
      "grad_norm": 2.4316417204747953,
      "learning_rate": 2.34529702970297e-07,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8130620121955872,
      "num_tokens": 13190276.0,
      "step": 380
    },
    {
      "epoch": 0.07075208913649025,
      "grad_norm": 2.269216075509549,
      "learning_rate": 2.3514851485148513e-07,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8242106437683105,
      "num_tokens": 13225430.0,
      "step": 381
    },
    {
      "epoch": 0.07093779015784586,
      "grad_norm": 2.542725594032753,
      "learning_rate": 2.3576732673267324e-07,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8222439289093018,
      "num_tokens": 13258088.0,
      "step": 382
    },
    {
      "epoch": 0.07112349117920148,
      "grad_norm": 2.4919618939313892,
      "learning_rate": 2.3638613861386135e-07,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8247385025024414,
      "num_tokens": 13291561.0,
      "step": 383
    },
    {
      "epoch": 0.0713091922005571,
      "grad_norm": 2.7007136519728054,
      "learning_rate": 2.370049504950495e-07,
      "loss": 0.601,
      "mean_token_accuracy": 0.8122173547744751,
      "num_tokens": 13323709.0,
      "step": 384
    },
    {
      "epoch": 0.07149489322191271,
      "grad_norm": 2.2540868193736796,
      "learning_rate": 2.376237623762376e-07,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8312735557556152,
      "num_tokens": 13356472.0,
      "step": 385
    },
    {
      "epoch": 0.07168059424326834,
      "grad_norm": 2.565940993691971,
      "learning_rate": 2.3824257425742571e-07,
      "loss": 0.6316,
      "mean_token_accuracy": 0.811515212059021,
      "num_tokens": 13385401.0,
      "step": 386
    },
    {
      "epoch": 0.07186629526462396,
      "grad_norm": 2.102644063976606,
      "learning_rate": 2.388613861386138e-07,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8273032903671265,
      "num_tokens": 13418576.0,
      "step": 387
    },
    {
      "epoch": 0.07205199628597957,
      "grad_norm": 2.098044564567586,
      "learning_rate": 2.3948019801980194e-07,
      "loss": 0.608,
      "mean_token_accuracy": 0.8130183219909668,
      "num_tokens": 13453479.0,
      "step": 388
    },
    {
      "epoch": 0.07223769730733519,
      "grad_norm": 2.266627584912465,
      "learning_rate": 2.400990099009901e-07,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8362003564834595,
      "num_tokens": 13485032.0,
      "step": 389
    },
    {
      "epoch": 0.07242339832869081,
      "grad_norm": 2.2994360569342205,
      "learning_rate": 2.407178217821782e-07,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8232330083847046,
      "num_tokens": 13521766.0,
      "step": 390
    },
    {
      "epoch": 0.07260909935004642,
      "grad_norm": 2.209040325328213,
      "learning_rate": 2.413366336633663e-07,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8241219520568848,
      "num_tokens": 13557111.0,
      "step": 391
    },
    {
      "epoch": 0.07279480037140204,
      "grad_norm": 1.7781473422306038,
      "learning_rate": 2.4195544554455444e-07,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8373317718505859,
      "num_tokens": 13599058.0,
      "step": 392
    },
    {
      "epoch": 0.07298050139275766,
      "grad_norm": 1.968239496636566,
      "learning_rate": 2.4257425742574255e-07,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8280948996543884,
      "num_tokens": 13641216.0,
      "step": 393
    },
    {
      "epoch": 0.07316620241411327,
      "grad_norm": 2.365213877478072,
      "learning_rate": 2.4319306930693066e-07,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8228469491004944,
      "num_tokens": 13677213.0,
      "step": 394
    },
    {
      "epoch": 0.0733519034354689,
      "grad_norm": 2.339677235838816,
      "learning_rate": 2.4381188118811877e-07,
      "loss": 0.5494,
      "mean_token_accuracy": 0.831316351890564,
      "num_tokens": 13710612.0,
      "step": 395
    },
    {
      "epoch": 0.07353760445682452,
      "grad_norm": 1.9035776278815009,
      "learning_rate": 2.4443069306930693e-07,
      "loss": 0.5924,
      "mean_token_accuracy": 0.819975733757019,
      "num_tokens": 13750692.0,
      "step": 396
    },
    {
      "epoch": 0.07372330547818012,
      "grad_norm": 2.193079574849812,
      "learning_rate": 2.4504950495049505e-07,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8193085789680481,
      "num_tokens": 13786289.0,
      "step": 397
    },
    {
      "epoch": 0.07390900649953575,
      "grad_norm": 2.026482304976695,
      "learning_rate": 2.4566831683168316e-07,
      "loss": 0.504,
      "mean_token_accuracy": 0.8423753976821899,
      "num_tokens": 13821354.0,
      "step": 398
    },
    {
      "epoch": 0.07409470752089137,
      "grad_norm": 2.6599126664721986,
      "learning_rate": 2.4628712871287127e-07,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8186379671096802,
      "num_tokens": 13851724.0,
      "step": 399
    },
    {
      "epoch": 0.07428040854224698,
      "grad_norm": 2.093716267863664,
      "learning_rate": 2.469059405940594e-07,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8225926160812378,
      "num_tokens": 13887014.0,
      "step": 400
    },
    {
      "epoch": 0.0744661095636026,
      "grad_norm": 1.9300843615954002,
      "learning_rate": 2.475247524752475e-07,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8334250450134277,
      "num_tokens": 13924729.0,
      "step": 401
    },
    {
      "epoch": 0.07465181058495822,
      "grad_norm": 2.484510116745734,
      "learning_rate": 2.4814356435643565e-07,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8351739645004272,
      "num_tokens": 13956735.0,
      "step": 402
    },
    {
      "epoch": 0.07483751160631383,
      "grad_norm": 2.5923618949584624,
      "learning_rate": 2.4876237623762377e-07,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8350322842597961,
      "num_tokens": 13982494.0,
      "step": 403
    },
    {
      "epoch": 0.07502321262766945,
      "grad_norm": 2.1986961743892866,
      "learning_rate": 2.493811881188119e-07,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8273425698280334,
      "num_tokens": 14020511.0,
      "step": 404
    },
    {
      "epoch": 0.07520891364902507,
      "grad_norm": 2.239858026822935,
      "learning_rate": 2.5e-07,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8177433013916016,
      "num_tokens": 14054233.0,
      "step": 405
    },
    {
      "epoch": 0.07539461467038068,
      "grad_norm": 2.06479577063641,
      "learning_rate": 2.506188118811881e-07,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8266050815582275,
      "num_tokens": 14088555.0,
      "step": 406
    },
    {
      "epoch": 0.0755803156917363,
      "grad_norm": 1.8863585365014013,
      "learning_rate": 2.512376237623762e-07,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8332599401473999,
      "num_tokens": 14128994.0,
      "step": 407
    },
    {
      "epoch": 0.07576601671309192,
      "grad_norm": 2.0368450523296806,
      "learning_rate": 2.518564356435643e-07,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8197459578514099,
      "num_tokens": 14163713.0,
      "step": 408
    },
    {
      "epoch": 0.07595171773444755,
      "grad_norm": 1.9338892555739202,
      "learning_rate": 2.524752475247525e-07,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8371756076812744,
      "num_tokens": 14201973.0,
      "step": 409
    },
    {
      "epoch": 0.07613741875580315,
      "grad_norm": 2.4530972272610576,
      "learning_rate": 2.530940594059406e-07,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8404809236526489,
      "num_tokens": 14231112.0,
      "step": 410
    },
    {
      "epoch": 0.07632311977715878,
      "grad_norm": 2.3435702163991565,
      "learning_rate": 2.537128712871287e-07,
      "loss": 0.616,
      "mean_token_accuracy": 0.8086349964141846,
      "num_tokens": 14258196.0,
      "step": 411
    },
    {
      "epoch": 0.0765088207985144,
      "grad_norm": 2.0117600446021293,
      "learning_rate": 2.543316831683168e-07,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8334547877311707,
      "num_tokens": 14289167.0,
      "step": 412
    },
    {
      "epoch": 0.07669452181987,
      "grad_norm": 2.3384804281584244,
      "learning_rate": 2.5495049504950493e-07,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8340773582458496,
      "num_tokens": 14315992.0,
      "step": 413
    },
    {
      "epoch": 0.07688022284122563,
      "grad_norm": 1.9987143754621282,
      "learning_rate": 2.5556930693069304e-07,
      "loss": 0.529,
      "mean_token_accuracy": 0.8360682725906372,
      "num_tokens": 14350932.0,
      "step": 414
    },
    {
      "epoch": 0.07706592386258125,
      "grad_norm": 1.949338740485788,
      "learning_rate": 2.561881188118812e-07,
      "loss": 0.459,
      "mean_token_accuracy": 0.8540314435958862,
      "num_tokens": 14386794.0,
      "step": 415
    },
    {
      "epoch": 0.07725162488393686,
      "grad_norm": 2.222189739968826,
      "learning_rate": 2.568069306930693e-07,
      "loss": 0.489,
      "mean_token_accuracy": 0.8446543216705322,
      "num_tokens": 14421020.0,
      "step": 416
    },
    {
      "epoch": 0.07743732590529248,
      "grad_norm": 2.441811963441593,
      "learning_rate": 2.5742574257425743e-07,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8070508241653442,
      "num_tokens": 14451558.0,
      "step": 417
    },
    {
      "epoch": 0.0776230269266481,
      "grad_norm": 2.0120140084398157,
      "learning_rate": 2.5804455445544554e-07,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8381824493408203,
      "num_tokens": 14484551.0,
      "step": 418
    },
    {
      "epoch": 0.07780872794800371,
      "grad_norm": 1.7794850162358855,
      "learning_rate": 2.5866336633663365e-07,
      "loss": 0.4944,
      "mean_token_accuracy": 0.845389723777771,
      "num_tokens": 14520555.0,
      "step": 419
    },
    {
      "epoch": 0.07799442896935933,
      "grad_norm": 1.838903392227477,
      "learning_rate": 2.5928217821782176e-07,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8398743867874146,
      "num_tokens": 14562275.0,
      "step": 420
    },
    {
      "epoch": 0.07818012999071496,
      "grad_norm": 2.2444671217125185,
      "learning_rate": 2.599009900990099e-07,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8094115853309631,
      "num_tokens": 14593821.0,
      "step": 421
    },
    {
      "epoch": 0.07836583101207056,
      "grad_norm": 1.9835175702992953,
      "learning_rate": 2.6051980198019804e-07,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8306822776794434,
      "num_tokens": 14626701.0,
      "step": 422
    },
    {
      "epoch": 0.07855153203342619,
      "grad_norm": 1.9021401572809926,
      "learning_rate": 2.6113861386138615e-07,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8462146520614624,
      "num_tokens": 14663874.0,
      "step": 423
    },
    {
      "epoch": 0.07873723305478181,
      "grad_norm": 2.0973485764011213,
      "learning_rate": 2.6175742574257426e-07,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8275651931762695,
      "num_tokens": 14697311.0,
      "step": 424
    },
    {
      "epoch": 0.07892293407613742,
      "grad_norm": 1.9749241214459554,
      "learning_rate": 2.623762376237624e-07,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8234477639198303,
      "num_tokens": 14737399.0,
      "step": 425
    },
    {
      "epoch": 0.07910863509749304,
      "grad_norm": 2.782277199326522,
      "learning_rate": 2.629950495049505e-07,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8391250371932983,
      "num_tokens": 14766184.0,
      "step": 426
    },
    {
      "epoch": 0.07929433611884866,
      "grad_norm": 1.8802799233891976,
      "learning_rate": 2.636138613861386e-07,
      "loss": 0.5083,
      "mean_token_accuracy": 0.841774046421051,
      "num_tokens": 14804495.0,
      "step": 427
    },
    {
      "epoch": 0.07948003714020427,
      "grad_norm": 1.9638024533939675,
      "learning_rate": 2.642326732673267e-07,
      "loss": 0.5316,
      "mean_token_accuracy": 0.8333708047866821,
      "num_tokens": 14845845.0,
      "step": 428
    },
    {
      "epoch": 0.07966573816155989,
      "grad_norm": 2.341521122348596,
      "learning_rate": 2.6485148514851487e-07,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8312817811965942,
      "num_tokens": 14877076.0,
      "step": 429
    },
    {
      "epoch": 0.07985143918291551,
      "grad_norm": 2.0842536861935543,
      "learning_rate": 2.65470297029703e-07,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8245604038238525,
      "num_tokens": 14908994.0,
      "step": 430
    },
    {
      "epoch": 0.08003714020427112,
      "grad_norm": 2.0372324751504585,
      "learning_rate": 2.660891089108911e-07,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8435868620872498,
      "num_tokens": 14941352.0,
      "step": 431
    },
    {
      "epoch": 0.08022284122562674,
      "grad_norm": 1.9979539848941446,
      "learning_rate": 2.667079207920792e-07,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8361645936965942,
      "num_tokens": 14975201.0,
      "step": 432
    },
    {
      "epoch": 0.08040854224698236,
      "grad_norm": 1.9841797165166029,
      "learning_rate": 2.673267326732673e-07,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8386220932006836,
      "num_tokens": 15009571.0,
      "step": 433
    },
    {
      "epoch": 0.08059424326833797,
      "grad_norm": 1.848548360569142,
      "learning_rate": 2.6794554455445543e-07,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8427598476409912,
      "num_tokens": 15043991.0,
      "step": 434
    },
    {
      "epoch": 0.0807799442896936,
      "grad_norm": 2.4566187404502067,
      "learning_rate": 2.685643564356436e-07,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8333471417427063,
      "num_tokens": 15076154.0,
      "step": 435
    },
    {
      "epoch": 0.08096564531104922,
      "grad_norm": 2.0730105771806047,
      "learning_rate": 2.691831683168317e-07,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8341481685638428,
      "num_tokens": 15105950.0,
      "step": 436
    },
    {
      "epoch": 0.08115134633240483,
      "grad_norm": 2.018490711279845,
      "learning_rate": 2.698019801980198e-07,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8233605623245239,
      "num_tokens": 15138321.0,
      "step": 437
    },
    {
      "epoch": 0.08133704735376045,
      "grad_norm": 2.4543639575961618,
      "learning_rate": 2.7042079207920793e-07,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8365338444709778,
      "num_tokens": 15161692.0,
      "step": 438
    },
    {
      "epoch": 0.08152274837511607,
      "grad_norm": 2.0177188566430773,
      "learning_rate": 2.7103960396039604e-07,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8072720170021057,
      "num_tokens": 15192387.0,
      "step": 439
    },
    {
      "epoch": 0.08170844939647168,
      "grad_norm": 1.7299609671460001,
      "learning_rate": 2.7165841584158415e-07,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8507304191589355,
      "num_tokens": 15230619.0,
      "step": 440
    },
    {
      "epoch": 0.0818941504178273,
      "grad_norm": 2.002158837229134,
      "learning_rate": 2.7227722772277226e-07,
      "loss": 0.546,
      "mean_token_accuracy": 0.8311117887496948,
      "num_tokens": 15269213.0,
      "step": 441
    },
    {
      "epoch": 0.08207985143918292,
      "grad_norm": 2.2261078457982344,
      "learning_rate": 2.728960396039604e-07,
      "loss": 0.531,
      "mean_token_accuracy": 0.8358350396156311,
      "num_tokens": 15299664.0,
      "step": 442
    },
    {
      "epoch": 0.08226555246053853,
      "grad_norm": 2.2908756564555857,
      "learning_rate": 2.7351485148514854e-07,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8152558207511902,
      "num_tokens": 15326409.0,
      "step": 443
    },
    {
      "epoch": 0.08245125348189415,
      "grad_norm": 2.2466070145709636,
      "learning_rate": 2.7413366336633665e-07,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8314728736877441,
      "num_tokens": 15359312.0,
      "step": 444
    },
    {
      "epoch": 0.08263695450324977,
      "grad_norm": 2.001678350751221,
      "learning_rate": 2.7475247524752476e-07,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8434656858444214,
      "num_tokens": 15394501.0,
      "step": 445
    },
    {
      "epoch": 0.08282265552460538,
      "grad_norm": 1.906365466854046,
      "learning_rate": 2.7537128712871287e-07,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8396520614624023,
      "num_tokens": 15429288.0,
      "step": 446
    },
    {
      "epoch": 0.083008356545961,
      "grad_norm": 1.8633320685785402,
      "learning_rate": 2.75990099009901e-07,
      "loss": 0.5345,
      "mean_token_accuracy": 0.832705020904541,
      "num_tokens": 15468281.0,
      "step": 447
    },
    {
      "epoch": 0.08319405756731663,
      "grad_norm": 1.83497127460738,
      "learning_rate": 2.7660891089108915e-07,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8353453874588013,
      "num_tokens": 15508987.0,
      "step": 448
    },
    {
      "epoch": 0.08337975858867223,
      "grad_norm": 1.954441647487458,
      "learning_rate": 2.7722772277227726e-07,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8487128019332886,
      "num_tokens": 15545983.0,
      "step": 449
    },
    {
      "epoch": 0.08356545961002786,
      "grad_norm": 2.13599844253323,
      "learning_rate": 2.7784653465346537e-07,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8278701305389404,
      "num_tokens": 15586761.0,
      "step": 450
    },
    {
      "epoch": 0.08375116063138348,
      "grad_norm": 1.9156109541751347,
      "learning_rate": 2.784653465346535e-07,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8313653469085693,
      "num_tokens": 15619578.0,
      "step": 451
    },
    {
      "epoch": 0.08393686165273909,
      "grad_norm": 1.7460218362009137,
      "learning_rate": 2.790841584158416e-07,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8375782370567322,
      "num_tokens": 15656730.0,
      "step": 452
    },
    {
      "epoch": 0.08412256267409471,
      "grad_norm": 1.8674739512889673,
      "learning_rate": 2.797029702970297e-07,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8276095986366272,
      "num_tokens": 15693958.0,
      "step": 453
    },
    {
      "epoch": 0.08430826369545033,
      "grad_norm": 1.643140586752524,
      "learning_rate": 2.803217821782178e-07,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8361343741416931,
      "num_tokens": 15733608.0,
      "step": 454
    },
    {
      "epoch": 0.08449396471680594,
      "grad_norm": 2.129955098132695,
      "learning_rate": 2.80940594059406e-07,
      "loss": 0.525,
      "mean_token_accuracy": 0.8336700201034546,
      "num_tokens": 15760896.0,
      "step": 455
    },
    {
      "epoch": 0.08467966573816156,
      "grad_norm": 1.5729824722590136,
      "learning_rate": 2.8155940594059404e-07,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8456288576126099,
      "num_tokens": 15800797.0,
      "step": 456
    },
    {
      "epoch": 0.08486536675951718,
      "grad_norm": 1.7778789946995224,
      "learning_rate": 2.8217821782178215e-07,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8233253955841064,
      "num_tokens": 15838204.0,
      "step": 457
    },
    {
      "epoch": 0.08505106778087279,
      "grad_norm": 1.939596796110868,
      "learning_rate": 2.8279702970297026e-07,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8290884494781494,
      "num_tokens": 15872192.0,
      "step": 458
    },
    {
      "epoch": 0.08523676880222841,
      "grad_norm": 1.7494914634546868,
      "learning_rate": 2.8341584158415837e-07,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8258861303329468,
      "num_tokens": 15911041.0,
      "step": 459
    },
    {
      "epoch": 0.08542246982358404,
      "grad_norm": 2.0567959762290733,
      "learning_rate": 2.840346534653465e-07,
      "loss": 0.574,
      "mean_token_accuracy": 0.8201936483383179,
      "num_tokens": 15939239.0,
      "step": 460
    },
    {
      "epoch": 0.08560817084493964,
      "grad_norm": 1.7485962755484152,
      "learning_rate": 2.846534653465346e-07,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8400760293006897,
      "num_tokens": 15973687.0,
      "step": 461
    },
    {
      "epoch": 0.08579387186629527,
      "grad_norm": 1.79377036368726,
      "learning_rate": 2.8527227722772276e-07,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8268098831176758,
      "num_tokens": 16010267.0,
      "step": 462
    },
    {
      "epoch": 0.08597957288765089,
      "grad_norm": 1.9385887539982927,
      "learning_rate": 2.8589108910891087e-07,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8400081396102905,
      "num_tokens": 16043893.0,
      "step": 463
    },
    {
      "epoch": 0.0861652739090065,
      "grad_norm": 1.7905859451217396,
      "learning_rate": 2.86509900990099e-07,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8322291970252991,
      "num_tokens": 16081869.0,
      "step": 464
    },
    {
      "epoch": 0.08635097493036212,
      "grad_norm": 1.7994696026400807,
      "learning_rate": 2.871287128712871e-07,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8312927484512329,
      "num_tokens": 16117867.0,
      "step": 465
    },
    {
      "epoch": 0.08653667595171774,
      "grad_norm": 2.182429779737584,
      "learning_rate": 2.877475247524752e-07,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8277556896209717,
      "num_tokens": 16146606.0,
      "step": 466
    },
    {
      "epoch": 0.08672237697307335,
      "grad_norm": 1.728733325598923,
      "learning_rate": 2.883663366336633e-07,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8400640487670898,
      "num_tokens": 16181784.0,
      "step": 467
    },
    {
      "epoch": 0.08690807799442897,
      "grad_norm": 1.8143447146284581,
      "learning_rate": 2.889851485148514e-07,
      "loss": 0.514,
      "mean_token_accuracy": 0.8370375037193298,
      "num_tokens": 16218106.0,
      "step": 468
    },
    {
      "epoch": 0.08709377901578459,
      "grad_norm": 2.024691830809522,
      "learning_rate": 2.896039603960396e-07,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8333652019500732,
      "num_tokens": 16250568.0,
      "step": 469
    },
    {
      "epoch": 0.0872794800371402,
      "grad_norm": 1.8467719936797802,
      "learning_rate": 2.902227722772277e-07,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8298559188842773,
      "num_tokens": 16283813.0,
      "step": 470
    },
    {
      "epoch": 0.08746518105849582,
      "grad_norm": 1.8643721779911362,
      "learning_rate": 2.908415841584158e-07,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8160058259963989,
      "num_tokens": 16322346.0,
      "step": 471
    },
    {
      "epoch": 0.08765088207985144,
      "grad_norm": 1.6682050567107967,
      "learning_rate": 2.914603960396039e-07,
      "loss": 0.507,
      "mean_token_accuracy": 0.8397479057312012,
      "num_tokens": 16360729.0,
      "step": 472
    },
    {
      "epoch": 0.08783658310120705,
      "grad_norm": 2.0425275732655983,
      "learning_rate": 2.9207920792079203e-07,
      "loss": 0.5455,
      "mean_token_accuracy": 0.831763505935669,
      "num_tokens": 16390124.0,
      "step": 473
    },
    {
      "epoch": 0.08802228412256267,
      "grad_norm": 1.8675820337971158,
      "learning_rate": 2.9269801980198015e-07,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8647638559341431,
      "num_tokens": 16422240.0,
      "step": 474
    },
    {
      "epoch": 0.0882079851439183,
      "grad_norm": 1.6871790706545855,
      "learning_rate": 2.933168316831683e-07,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8606342077255249,
      "num_tokens": 16452598.0,
      "step": 475
    },
    {
      "epoch": 0.0883936861652739,
      "grad_norm": 1.694270531945463,
      "learning_rate": 2.939356435643564e-07,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8428376913070679,
      "num_tokens": 16486933.0,
      "step": 476
    },
    {
      "epoch": 0.08857938718662953,
      "grad_norm": 1.7576389004621744,
      "learning_rate": 2.9455445544554453e-07,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8180326223373413,
      "num_tokens": 16524360.0,
      "step": 477
    },
    {
      "epoch": 0.08876508820798515,
      "grad_norm": 1.9127287367997161,
      "learning_rate": 2.9517326732673264e-07,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8220419883728027,
      "num_tokens": 16559670.0,
      "step": 478
    },
    {
      "epoch": 0.08895078922934076,
      "grad_norm": 1.8512102908086707,
      "learning_rate": 2.9579207920792076e-07,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8307552933692932,
      "num_tokens": 16589166.0,
      "step": 479
    },
    {
      "epoch": 0.08913649025069638,
      "grad_norm": 1.851049583536,
      "learning_rate": 2.9641089108910887e-07,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8247278928756714,
      "num_tokens": 16620741.0,
      "step": 480
    },
    {
      "epoch": 0.089322191272052,
      "grad_norm": 1.9241992541725197,
      "learning_rate": 2.97029702970297e-07,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8314586877822876,
      "num_tokens": 16656285.0,
      "step": 481
    },
    {
      "epoch": 0.08950789229340761,
      "grad_norm": 1.8957778859157048,
      "learning_rate": 2.9764851485148514e-07,
      "loss": 0.4604,
      "mean_token_accuracy": 0.848829984664917,
      "num_tokens": 16686206.0,
      "step": 482
    },
    {
      "epoch": 0.08969359331476323,
      "grad_norm": 1.798744670613323,
      "learning_rate": 2.9826732673267325e-07,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8313331604003906,
      "num_tokens": 16727362.0,
      "step": 483
    },
    {
      "epoch": 0.08987929433611885,
      "grad_norm": 1.83940878411473,
      "learning_rate": 2.9888613861386136e-07,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8305760622024536,
      "num_tokens": 16760169.0,
      "step": 484
    },
    {
      "epoch": 0.09006499535747446,
      "grad_norm": 1.8023070187538373,
      "learning_rate": 2.995049504950495e-07,
      "loss": 0.5271,
      "mean_token_accuracy": 0.833012580871582,
      "num_tokens": 16794029.0,
      "step": 485
    },
    {
      "epoch": 0.09025069637883008,
      "grad_norm": 1.8829930634137788,
      "learning_rate": 3.001237623762376e-07,
      "loss": 0.529,
      "mean_token_accuracy": 0.8348146677017212,
      "num_tokens": 16827270.0,
      "step": 486
    },
    {
      "epoch": 0.0904363974001857,
      "grad_norm": 1.647901948947457,
      "learning_rate": 3.007425742574257e-07,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8423418402671814,
      "num_tokens": 16864923.0,
      "step": 487
    },
    {
      "epoch": 0.09062209842154131,
      "grad_norm": 2.0121719310519945,
      "learning_rate": 3.0136138613861386e-07,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8152299523353577,
      "num_tokens": 16895594.0,
      "step": 488
    },
    {
      "epoch": 0.09080779944289694,
      "grad_norm": 1.7332945388159762,
      "learning_rate": 3.01980198019802e-07,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8299543857574463,
      "num_tokens": 16928619.0,
      "step": 489
    },
    {
      "epoch": 0.09099350046425256,
      "grad_norm": 1.8765801284484886,
      "learning_rate": 3.025990099009901e-07,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8355164527893066,
      "num_tokens": 16957241.0,
      "step": 490
    },
    {
      "epoch": 0.09117920148560817,
      "grad_norm": 1.928572024872104,
      "learning_rate": 3.032178217821782e-07,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8391711711883545,
      "num_tokens": 16991500.0,
      "step": 491
    },
    {
      "epoch": 0.09136490250696379,
      "grad_norm": 1.588260560814319,
      "learning_rate": 3.038366336633663e-07,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8495410680770874,
      "num_tokens": 17027823.0,
      "step": 492
    },
    {
      "epoch": 0.09155060352831941,
      "grad_norm": 1.9270045671573106,
      "learning_rate": 3.044554455445544e-07,
      "loss": 0.483,
      "mean_token_accuracy": 0.8418212532997131,
      "num_tokens": 17055329.0,
      "step": 493
    },
    {
      "epoch": 0.09173630454967502,
      "grad_norm": 1.8312202765437195,
      "learning_rate": 3.0507425742574253e-07,
      "loss": 0.528,
      "mean_token_accuracy": 0.8244524598121643,
      "num_tokens": 17083723.0,
      "step": 494
    },
    {
      "epoch": 0.09192200557103064,
      "grad_norm": 1.936994959576631,
      "learning_rate": 3.056930693069307e-07,
      "loss": 0.624,
      "mean_token_accuracy": 0.8118287920951843,
      "num_tokens": 17115406.0,
      "step": 495
    },
    {
      "epoch": 0.09210770659238626,
      "grad_norm": 1.8417490124519431,
      "learning_rate": 3.063118811881188e-07,
      "loss": 0.5434,
      "mean_token_accuracy": 0.822861909866333,
      "num_tokens": 17150329.0,
      "step": 496
    },
    {
      "epoch": 0.09229340761374187,
      "grad_norm": 1.6730937730287725,
      "learning_rate": 3.069306930693069e-07,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8235050439834595,
      "num_tokens": 17190989.0,
      "step": 497
    },
    {
      "epoch": 0.09247910863509749,
      "grad_norm": 1.8815091219967532,
      "learning_rate": 3.0754950495049503e-07,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8209787607192993,
      "num_tokens": 17221424.0,
      "step": 498
    },
    {
      "epoch": 0.09266480965645311,
      "grad_norm": 1.845005521718087,
      "learning_rate": 3.0816831683168314e-07,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8414015173912048,
      "num_tokens": 17252708.0,
      "step": 499
    },
    {
      "epoch": 0.09285051067780872,
      "grad_norm": 1.901148926990623,
      "learning_rate": 3.0878712871287125e-07,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8329790234565735,
      "num_tokens": 17283342.0,
      "step": 500
    },
    {
      "epoch": 0.09303621169916435,
      "grad_norm": 1.8980264404199225,
      "learning_rate": 3.0940594059405936e-07,
      "loss": 0.4979,
      "mean_token_accuracy": 0.842140257358551,
      "num_tokens": 17312891.0,
      "step": 501
    },
    {
      "epoch": 0.09322191272051997,
      "grad_norm": 1.741314064399437,
      "learning_rate": 3.1002475247524753e-07,
      "loss": 0.494,
      "mean_token_accuracy": 0.8442635536193848,
      "num_tokens": 17348013.0,
      "step": 502
    },
    {
      "epoch": 0.09340761374187558,
      "grad_norm": 1.7438550365133543,
      "learning_rate": 3.1064356435643564e-07,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8370418548583984,
      "num_tokens": 17384703.0,
      "step": 503
    },
    {
      "epoch": 0.0935933147632312,
      "grad_norm": 1.8098676565264376,
      "learning_rate": 3.1126237623762375e-07,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8284441232681274,
      "num_tokens": 17421212.0,
      "step": 504
    },
    {
      "epoch": 0.09377901578458682,
      "grad_norm": 1.7519775311066235,
      "learning_rate": 3.1188118811881186e-07,
      "loss": 0.545,
      "mean_token_accuracy": 0.8270701169967651,
      "num_tokens": 17457581.0,
      "step": 505
    },
    {
      "epoch": 0.09396471680594243,
      "grad_norm": 1.6495940407685261,
      "learning_rate": 3.1249999999999997e-07,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8367671966552734,
      "num_tokens": 17495903.0,
      "step": 506
    },
    {
      "epoch": 0.09415041782729805,
      "grad_norm": 1.6588483709603314,
      "learning_rate": 3.131188118811881e-07,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8378045558929443,
      "num_tokens": 17532961.0,
      "step": 507
    },
    {
      "epoch": 0.09433611884865367,
      "grad_norm": 1.8807246589570212,
      "learning_rate": 3.1373762376237625e-07,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8213587999343872,
      "num_tokens": 17565530.0,
      "step": 508
    },
    {
      "epoch": 0.09452181987000928,
      "grad_norm": 1.6220799276985927,
      "learning_rate": 3.1435643564356436e-07,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8513814806938171,
      "num_tokens": 17602939.0,
      "step": 509
    },
    {
      "epoch": 0.0947075208913649,
      "grad_norm": 1.601079620106917,
      "learning_rate": 3.1497524752475247e-07,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8364995718002319,
      "num_tokens": 17640595.0,
      "step": 510
    },
    {
      "epoch": 0.09489322191272052,
      "grad_norm": 1.7152759664790467,
      "learning_rate": 3.155940594059406e-07,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8593779802322388,
      "num_tokens": 17676377.0,
      "step": 511
    },
    {
      "epoch": 0.09507892293407613,
      "grad_norm": 2.0286325565619565,
      "learning_rate": 3.162128712871287e-07,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8309367895126343,
      "num_tokens": 17706124.0,
      "step": 512
    },
    {
      "epoch": 0.09526462395543175,
      "grad_norm": 1.8907671850636734,
      "learning_rate": 3.168316831683168e-07,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8563295602798462,
      "num_tokens": 17733973.0,
      "step": 513
    },
    {
      "epoch": 0.09545032497678738,
      "grad_norm": 1.582633251434714,
      "learning_rate": 3.174504950495049e-07,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8340563178062439,
      "num_tokens": 17773281.0,
      "step": 514
    },
    {
      "epoch": 0.09563602599814298,
      "grad_norm": 1.6724643969866464,
      "learning_rate": 3.180693069306931e-07,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8310997486114502,
      "num_tokens": 17806377.0,
      "step": 515
    },
    {
      "epoch": 0.0958217270194986,
      "grad_norm": 2.12491099985384,
      "learning_rate": 3.186881188118812e-07,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8388565182685852,
      "num_tokens": 17832074.0,
      "step": 516
    },
    {
      "epoch": 0.09600742804085423,
      "grad_norm": 1.936069721526375,
      "learning_rate": 3.193069306930693e-07,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8319425582885742,
      "num_tokens": 17859835.0,
      "step": 517
    },
    {
      "epoch": 0.09619312906220984,
      "grad_norm": 1.6883816182608882,
      "learning_rate": 3.199257425742574e-07,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8417066335678101,
      "num_tokens": 17897628.0,
      "step": 518
    },
    {
      "epoch": 0.09637883008356546,
      "grad_norm": 1.6560442487094693,
      "learning_rate": 3.205445544554455e-07,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8376258611679077,
      "num_tokens": 17937041.0,
      "step": 519
    },
    {
      "epoch": 0.09656453110492108,
      "grad_norm": 1.9154519951864477,
      "learning_rate": 3.2116336633663364e-07,
      "loss": 0.5511,
      "mean_token_accuracy": 0.82929527759552,
      "num_tokens": 17966517.0,
      "step": 520
    },
    {
      "epoch": 0.09675023212627669,
      "grad_norm": 1.7024846440875316,
      "learning_rate": 3.217821782178218e-07,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8154300451278687,
      "num_tokens": 18006295.0,
      "step": 521
    },
    {
      "epoch": 0.09693593314763231,
      "grad_norm": 2.0004559673124316,
      "learning_rate": 3.224009900990099e-07,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8319536447525024,
      "num_tokens": 18036827.0,
      "step": 522
    },
    {
      "epoch": 0.09712163416898793,
      "grad_norm": 1.6435597271934463,
      "learning_rate": 3.23019801980198e-07,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8469040393829346,
      "num_tokens": 18072971.0,
      "step": 523
    },
    {
      "epoch": 0.09730733519034354,
      "grad_norm": 1.6675688072725274,
      "learning_rate": 3.2363861386138614e-07,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8367270231246948,
      "num_tokens": 18110288.0,
      "step": 524
    },
    {
      "epoch": 0.09749303621169916,
      "grad_norm": 1.7002842497410486,
      "learning_rate": 3.2425742574257425e-07,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8283596634864807,
      "num_tokens": 18147194.0,
      "step": 525
    },
    {
      "epoch": 0.09767873723305479,
      "grad_norm": 1.8043107046901148,
      "learning_rate": 3.2487623762376236e-07,
      "loss": 0.6309,
      "mean_token_accuracy": 0.8072329759597778,
      "num_tokens": 18185073.0,
      "step": 526
    },
    {
      "epoch": 0.0978644382544104,
      "grad_norm": 1.762887305634597,
      "learning_rate": 3.2549504950495047e-07,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8315417766571045,
      "num_tokens": 18221272.0,
      "step": 527
    },
    {
      "epoch": 0.09805013927576602,
      "grad_norm": 1.7755320576943356,
      "learning_rate": 3.2611386138613863e-07,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8191142082214355,
      "num_tokens": 18256729.0,
      "step": 528
    },
    {
      "epoch": 0.09823584029712164,
      "grad_norm": 1.7288813728149728,
      "learning_rate": 3.2673267326732674e-07,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8403612375259399,
      "num_tokens": 18288315.0,
      "step": 529
    },
    {
      "epoch": 0.09842154131847725,
      "grad_norm": 1.7447387441755149,
      "learning_rate": 3.2735148514851486e-07,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8368561267852783,
      "num_tokens": 18328453.0,
      "step": 530
    },
    {
      "epoch": 0.09860724233983287,
      "grad_norm": 1.7329905932158602,
      "learning_rate": 3.2797029702970297e-07,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8404350280761719,
      "num_tokens": 18360619.0,
      "step": 531
    },
    {
      "epoch": 0.09879294336118849,
      "grad_norm": 1.6582054100102777,
      "learning_rate": 3.285891089108911e-07,
      "loss": 0.609,
      "mean_token_accuracy": 0.8093324899673462,
      "num_tokens": 18399358.0,
      "step": 532
    },
    {
      "epoch": 0.0989786443825441,
      "grad_norm": 1.8802376572758381,
      "learning_rate": 3.292079207920792e-07,
      "loss": 0.5628,
      "mean_token_accuracy": 0.824677586555481,
      "num_tokens": 18431729.0,
      "step": 533
    },
    {
      "epoch": 0.09916434540389972,
      "grad_norm": 1.7289561309279822,
      "learning_rate": 3.298267326732673e-07,
      "loss": 0.523,
      "mean_token_accuracy": 0.8368104696273804,
      "num_tokens": 18467381.0,
      "step": 534
    },
    {
      "epoch": 0.09935004642525534,
      "grad_norm": 1.8619285837087793,
      "learning_rate": 3.3044554455445547e-07,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8501687049865723,
      "num_tokens": 18495733.0,
      "step": 535
    },
    {
      "epoch": 0.09953574744661095,
      "grad_norm": 1.6602528560044107,
      "learning_rate": 3.310643564356436e-07,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8397723436355591,
      "num_tokens": 18532631.0,
      "step": 536
    },
    {
      "epoch": 0.09972144846796657,
      "grad_norm": 1.6904679569542242,
      "learning_rate": 3.316831683168317e-07,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8372489809989929,
      "num_tokens": 18566104.0,
      "step": 537
    },
    {
      "epoch": 0.0999071494893222,
      "grad_norm": 1.9639236772164246,
      "learning_rate": 3.323019801980198e-07,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8304256796836853,
      "num_tokens": 18593958.0,
      "step": 538
    },
    {
      "epoch": 0.1000928505106778,
      "grad_norm": 1.8480684424907792,
      "learning_rate": 3.329207920792079e-07,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8290356397628784,
      "num_tokens": 18626442.0,
      "step": 539
    },
    {
      "epoch": 0.10027855153203342,
      "grad_norm": 1.8701273369202334,
      "learning_rate": 3.33539603960396e-07,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8386066555976868,
      "num_tokens": 18655435.0,
      "step": 540
    },
    {
      "epoch": 0.10046425255338905,
      "grad_norm": 1.648259213375808,
      "learning_rate": 3.341584158415842e-07,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8414132595062256,
      "num_tokens": 18693352.0,
      "step": 541
    },
    {
      "epoch": 0.10064995357474465,
      "grad_norm": 1.8287219565845108,
      "learning_rate": 3.347772277227723e-07,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8320232629776001,
      "num_tokens": 18725670.0,
      "step": 542
    },
    {
      "epoch": 0.10083565459610028,
      "grad_norm": 1.9195430545947234,
      "learning_rate": 3.353960396039604e-07,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8298889398574829,
      "num_tokens": 18753406.0,
      "step": 543
    },
    {
      "epoch": 0.1010213556174559,
      "grad_norm": 1.7319551309011734,
      "learning_rate": 3.360148514851485e-07,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8341431617736816,
      "num_tokens": 18788925.0,
      "step": 544
    },
    {
      "epoch": 0.10120705663881151,
      "grad_norm": 1.549572245050466,
      "learning_rate": 3.3663366336633663e-07,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8480817079544067,
      "num_tokens": 18827672.0,
      "step": 545
    },
    {
      "epoch": 0.10139275766016713,
      "grad_norm": 1.6676515878698168,
      "learning_rate": 3.3725247524752474e-07,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8358334898948669,
      "num_tokens": 18862853.0,
      "step": 546
    },
    {
      "epoch": 0.10157845868152275,
      "grad_norm": 1.583486934596052,
      "learning_rate": 3.3787128712871285e-07,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8493006825447083,
      "num_tokens": 18897704.0,
      "step": 547
    },
    {
      "epoch": 0.10176415970287836,
      "grad_norm": 1.6339693449284498,
      "learning_rate": 3.38490099009901e-07,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8335853815078735,
      "num_tokens": 18937330.0,
      "step": 548
    },
    {
      "epoch": 0.10194986072423398,
      "grad_norm": 1.5013153432735662,
      "learning_rate": 3.3910891089108913e-07,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8577725291252136,
      "num_tokens": 18975209.0,
      "step": 549
    },
    {
      "epoch": 0.1021355617455896,
      "grad_norm": 1.7237960784514557,
      "learning_rate": 3.3972772277227724e-07,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8549569249153137,
      "num_tokens": 19005835.0,
      "step": 550
    },
    {
      "epoch": 0.10232126276694521,
      "grad_norm": 1.6607688111438834,
      "learning_rate": 3.4034653465346535e-07,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8393492698669434,
      "num_tokens": 19041785.0,
      "step": 551
    },
    {
      "epoch": 0.10250696378830083,
      "grad_norm": 1.7645385558691098,
      "learning_rate": 3.4096534653465346e-07,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8349355459213257,
      "num_tokens": 19076228.0,
      "step": 552
    },
    {
      "epoch": 0.10269266480965646,
      "grad_norm": 1.6383316038599272,
      "learning_rate": 3.415841584158416e-07,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8360510468482971,
      "num_tokens": 19112558.0,
      "step": 553
    },
    {
      "epoch": 0.10287836583101206,
      "grad_norm": 1.755879928078858,
      "learning_rate": 3.4220297029702974e-07,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8251420259475708,
      "num_tokens": 19146470.0,
      "step": 554
    },
    {
      "epoch": 0.10306406685236769,
      "grad_norm": 1.7179571777645288,
      "learning_rate": 3.4282178217821785e-07,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8370627760887146,
      "num_tokens": 19181331.0,
      "step": 555
    },
    {
      "epoch": 0.10324976787372331,
      "grad_norm": 1.660389143982462,
      "learning_rate": 3.4344059405940596e-07,
      "loss": 0.53,
      "mean_token_accuracy": 0.8320251107215881,
      "num_tokens": 19216383.0,
      "step": 556
    },
    {
      "epoch": 0.10343546889507892,
      "grad_norm": 1.470770569081989,
      "learning_rate": 3.44059405940594e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8554314374923706,
      "num_tokens": 19258133.0,
      "step": 557
    },
    {
      "epoch": 0.10362116991643454,
      "grad_norm": 1.6712600643988247,
      "learning_rate": 3.4467821782178213e-07,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8441650867462158,
      "num_tokens": 19292388.0,
      "step": 558
    },
    {
      "epoch": 0.10380687093779016,
      "grad_norm": 1.835035412737247,
      "learning_rate": 3.4529702970297024e-07,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8389516472816467,
      "num_tokens": 19320978.0,
      "step": 559
    },
    {
      "epoch": 0.10399257195914577,
      "grad_norm": 1.7492811523037113,
      "learning_rate": 3.4591584158415835e-07,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8324436545372009,
      "num_tokens": 19353815.0,
      "step": 560
    },
    {
      "epoch": 0.10417827298050139,
      "grad_norm": 1.8789061283022885,
      "learning_rate": 3.465346534653465e-07,
      "loss": 0.51,
      "mean_token_accuracy": 0.837151050567627,
      "num_tokens": 19382157.0,
      "step": 561
    },
    {
      "epoch": 0.10436397400185701,
      "grad_norm": 1.6442745713464315,
      "learning_rate": 3.4715346534653463e-07,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8343455791473389,
      "num_tokens": 19420395.0,
      "step": 562
    },
    {
      "epoch": 0.10454967502321262,
      "grad_norm": 1.5244928951199066,
      "learning_rate": 3.4777227722772274e-07,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8521033525466919,
      "num_tokens": 19460298.0,
      "step": 563
    },
    {
      "epoch": 0.10473537604456824,
      "grad_norm": 1.7791302396463053,
      "learning_rate": 3.4839108910891085e-07,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8377941846847534,
      "num_tokens": 19494455.0,
      "step": 564
    },
    {
      "epoch": 0.10492107706592387,
      "grad_norm": 1.6202783173971886,
      "learning_rate": 3.4900990099009896e-07,
      "loss": 0.4992,
      "mean_token_accuracy": 0.840867280960083,
      "num_tokens": 19531831.0,
      "step": 565
    },
    {
      "epoch": 0.10510677808727947,
      "grad_norm": 1.6236061619799218,
      "learning_rate": 3.496287128712871e-07,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8276318311691284,
      "num_tokens": 19570634.0,
      "step": 566
    },
    {
      "epoch": 0.1052924791086351,
      "grad_norm": 1.8740402961438725,
      "learning_rate": 3.502475247524752e-07,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8320468068122864,
      "num_tokens": 19599663.0,
      "step": 567
    },
    {
      "epoch": 0.10547818012999072,
      "grad_norm": 1.865873850239097,
      "learning_rate": 3.5086633663366335e-07,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8460807800292969,
      "num_tokens": 19629175.0,
      "step": 568
    },
    {
      "epoch": 0.10566388115134633,
      "grad_norm": 1.740735864600566,
      "learning_rate": 3.5148514851485146e-07,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8241517543792725,
      "num_tokens": 19665837.0,
      "step": 569
    },
    {
      "epoch": 0.10584958217270195,
      "grad_norm": 1.6211084024841997,
      "learning_rate": 3.5210396039603957e-07,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8408594131469727,
      "num_tokens": 19703440.0,
      "step": 570
    },
    {
      "epoch": 0.10603528319405757,
      "grad_norm": 1.6420238559102327,
      "learning_rate": 3.527227722772277e-07,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8397814035415649,
      "num_tokens": 19736977.0,
      "step": 571
    },
    {
      "epoch": 0.10622098421541319,
      "grad_norm": 1.723760510398762,
      "learning_rate": 3.533415841584158e-07,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8371474742889404,
      "num_tokens": 19769622.0,
      "step": 572
    },
    {
      "epoch": 0.1064066852367688,
      "grad_norm": 1.7522572438169375,
      "learning_rate": 3.539603960396039e-07,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8450411558151245,
      "num_tokens": 19803012.0,
      "step": 573
    },
    {
      "epoch": 0.10659238625812442,
      "grad_norm": 1.8114679615001235,
      "learning_rate": 3.5457920792079207e-07,
      "loss": 0.4304,
      "mean_token_accuracy": 0.862179696559906,
      "num_tokens": 19830888.0,
      "step": 574
    },
    {
      "epoch": 0.10677808727948004,
      "grad_norm": 1.7668228246049753,
      "learning_rate": 3.551980198019802e-07,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8153091073036194,
      "num_tokens": 19867226.0,
      "step": 575
    },
    {
      "epoch": 0.10696378830083565,
      "grad_norm": 1.589456165315488,
      "learning_rate": 3.558168316831683e-07,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8320741057395935,
      "num_tokens": 19906561.0,
      "step": 576
    },
    {
      "epoch": 0.10714948932219127,
      "grad_norm": 1.7827114899496912,
      "learning_rate": 3.564356435643564e-07,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8254417181015015,
      "num_tokens": 19936916.0,
      "step": 577
    },
    {
      "epoch": 0.1073351903435469,
      "grad_norm": 1.5938507972068838,
      "learning_rate": 3.570544554455445e-07,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8482434749603271,
      "num_tokens": 19974846.0,
      "step": 578
    },
    {
      "epoch": 0.1075208913649025,
      "grad_norm": 1.566913907659804,
      "learning_rate": 3.5767326732673263e-07,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8281151056289673,
      "num_tokens": 20015748.0,
      "step": 579
    },
    {
      "epoch": 0.10770659238625813,
      "grad_norm": 1.6649237956851979,
      "learning_rate": 3.5829207920792074e-07,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8258079290390015,
      "num_tokens": 20051463.0,
      "step": 580
    },
    {
      "epoch": 0.10789229340761375,
      "grad_norm": 1.5829164167072927,
      "learning_rate": 3.589108910891089e-07,
      "loss": 0.524,
      "mean_token_accuracy": 0.8335896134376526,
      "num_tokens": 20087468.0,
      "step": 581
    },
    {
      "epoch": 0.10807799442896936,
      "grad_norm": 1.6421500424375444,
      "learning_rate": 3.59529702970297e-07,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8352010250091553,
      "num_tokens": 20121217.0,
      "step": 582
    },
    {
      "epoch": 0.10826369545032498,
      "grad_norm": 1.7762036733239643,
      "learning_rate": 3.601485148514851e-07,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8477394580841064,
      "num_tokens": 20150116.0,
      "step": 583
    },
    {
      "epoch": 0.1084493964716806,
      "grad_norm": 1.53718655231812,
      "learning_rate": 3.6076732673267324e-07,
      "loss": 0.464,
      "mean_token_accuracy": 0.8483957648277283,
      "num_tokens": 20187938.0,
      "step": 584
    },
    {
      "epoch": 0.10863509749303621,
      "grad_norm": 1.6463794837653611,
      "learning_rate": 3.6138613861386135e-07,
      "loss": 0.485,
      "mean_token_accuracy": 0.8432353734970093,
      "num_tokens": 20221611.0,
      "step": 585
    },
    {
      "epoch": 0.10882079851439183,
      "grad_norm": 1.5652653383666197,
      "learning_rate": 3.6200495049504946e-07,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8460382223129272,
      "num_tokens": 20256808.0,
      "step": 586
    },
    {
      "epoch": 0.10900649953574745,
      "grad_norm": 2.237531464277062,
      "learning_rate": 3.6262376237623757e-07,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8428577780723572,
      "num_tokens": 20286260.0,
      "step": 587
    },
    {
      "epoch": 0.10919220055710306,
      "grad_norm": 1.603569464692342,
      "learning_rate": 3.6324257425742574e-07,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8429856300354004,
      "num_tokens": 20322484.0,
      "step": 588
    },
    {
      "epoch": 0.10937790157845868,
      "grad_norm": 1.6969300376240133,
      "learning_rate": 3.6386138613861385e-07,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8174694776535034,
      "num_tokens": 20360150.0,
      "step": 589
    },
    {
      "epoch": 0.1095636025998143,
      "grad_norm": 1.5952680245858215,
      "learning_rate": 3.6448019801980196e-07,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8428344130516052,
      "num_tokens": 20398502.0,
      "step": 590
    },
    {
      "epoch": 0.10974930362116991,
      "grad_norm": 1.573239574288456,
      "learning_rate": 3.6509900990099007e-07,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8262699842453003,
      "num_tokens": 20438288.0,
      "step": 591
    },
    {
      "epoch": 0.10993500464252554,
      "grad_norm": 1.5510266586910668,
      "learning_rate": 3.657178217821782e-07,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8324899673461914,
      "num_tokens": 20476687.0,
      "step": 592
    },
    {
      "epoch": 0.11012070566388116,
      "grad_norm": 1.6874759925750518,
      "learning_rate": 3.663366336633663e-07,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8399968147277832,
      "num_tokens": 20513461.0,
      "step": 593
    },
    {
      "epoch": 0.11030640668523677,
      "grad_norm": 1.668044383965799,
      "learning_rate": 3.6695544554455446e-07,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8492591381072998,
      "num_tokens": 20545686.0,
      "step": 594
    },
    {
      "epoch": 0.11049210770659239,
      "grad_norm": 1.5886445327957934,
      "learning_rate": 3.6757425742574257e-07,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8362780213356018,
      "num_tokens": 20581906.0,
      "step": 595
    },
    {
      "epoch": 0.11067780872794801,
      "grad_norm": 1.5609274162986329,
      "learning_rate": 3.681930693069307e-07,
      "loss": 0.5152,
      "mean_token_accuracy": 0.837030291557312,
      "num_tokens": 20619559.0,
      "step": 596
    },
    {
      "epoch": 0.11086350974930362,
      "grad_norm": 1.563506890548472,
      "learning_rate": 3.688118811881188e-07,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8461557626724243,
      "num_tokens": 20657239.0,
      "step": 597
    },
    {
      "epoch": 0.11104921077065924,
      "grad_norm": 1.7641625641422611,
      "learning_rate": 3.694306930693069e-07,
      "loss": 0.572,
      "mean_token_accuracy": 0.821830153465271,
      "num_tokens": 20692892.0,
      "step": 598
    },
    {
      "epoch": 0.11123491179201486,
      "grad_norm": 1.8261896404468108,
      "learning_rate": 3.70049504950495e-07,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8217473030090332,
      "num_tokens": 20724475.0,
      "step": 599
    },
    {
      "epoch": 0.11142061281337047,
      "grad_norm": 1.8955181855002885,
      "learning_rate": 3.706683168316831e-07,
      "loss": 0.5533,
      "mean_token_accuracy": 0.826324462890625,
      "num_tokens": 20757111.0,
      "step": 600
    },
    {
      "epoch": 0.11160631383472609,
      "grad_norm": 1.6354816782523258,
      "learning_rate": 3.712871287128713e-07,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8539690971374512,
      "num_tokens": 20790246.0,
      "step": 601
    },
    {
      "epoch": 0.11179201485608171,
      "grad_norm": 1.5826972627109814,
      "learning_rate": 3.719059405940594e-07,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8439018726348877,
      "num_tokens": 20827640.0,
      "step": 602
    },
    {
      "epoch": 0.11197771587743732,
      "grad_norm": 1.7168228442621103,
      "learning_rate": 3.725247524752475e-07,
      "loss": 0.559,
      "mean_token_accuracy": 0.8242475390434265,
      "num_tokens": 20860844.0,
      "step": 603
    },
    {
      "epoch": 0.11216341689879294,
      "grad_norm": 1.9026039709207139,
      "learning_rate": 3.731435643564356e-07,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8423489332199097,
      "num_tokens": 20886233.0,
      "step": 604
    },
    {
      "epoch": 0.11234911792014857,
      "grad_norm": 1.592425356490723,
      "learning_rate": 3.7376237623762373e-07,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8553932905197144,
      "num_tokens": 20922294.0,
      "step": 605
    },
    {
      "epoch": 0.11253481894150417,
      "grad_norm": 1.739055366999337,
      "learning_rate": 3.7438118811881185e-07,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8524646759033203,
      "num_tokens": 20960000.0,
      "step": 606
    },
    {
      "epoch": 0.1127205199628598,
      "grad_norm": 1.6058147019962834,
      "learning_rate": 3.75e-07,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8249022960662842,
      "num_tokens": 20999372.0,
      "step": 607
    },
    {
      "epoch": 0.11290622098421542,
      "grad_norm": 1.652185554646287,
      "learning_rate": 3.756188118811881e-07,
      "loss": 0.539,
      "mean_token_accuracy": 0.8307642936706543,
      "num_tokens": 21033275.0,
      "step": 608
    },
    {
      "epoch": 0.11309192200557103,
      "grad_norm": 1.574644907283434,
      "learning_rate": 3.7623762376237623e-07,
      "loss": 0.5007,
      "mean_token_accuracy": 0.840788722038269,
      "num_tokens": 21070559.0,
      "step": 609
    },
    {
      "epoch": 0.11327762302692665,
      "grad_norm": 1.5747951756184584,
      "learning_rate": 3.7685643564356434e-07,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8481936454772949,
      "num_tokens": 21104419.0,
      "step": 610
    },
    {
      "epoch": 0.11346332404828227,
      "grad_norm": 1.643466236860062,
      "learning_rate": 3.7747524752475245e-07,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8390921354293823,
      "num_tokens": 21138167.0,
      "step": 611
    },
    {
      "epoch": 0.11364902506963788,
      "grad_norm": 1.6258949884855147,
      "learning_rate": 3.7809405940594057e-07,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8360673189163208,
      "num_tokens": 21169694.0,
      "step": 612
    },
    {
      "epoch": 0.1138347260909935,
      "grad_norm": 1.724352881683009,
      "learning_rate": 3.787128712871287e-07,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8366657495498657,
      "num_tokens": 21201021.0,
      "step": 613
    },
    {
      "epoch": 0.11402042711234912,
      "grad_norm": 1.6794846251785496,
      "learning_rate": 3.7933168316831684e-07,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8348981738090515,
      "num_tokens": 21235213.0,
      "step": 614
    },
    {
      "epoch": 0.11420612813370473,
      "grad_norm": 1.788872471518446,
      "learning_rate": 3.7995049504950495e-07,
      "loss": 0.4986,
      "mean_token_accuracy": 0.839112401008606,
      "num_tokens": 21262874.0,
      "step": 615
    },
    {
      "epoch": 0.11439182915506035,
      "grad_norm": 1.7210029940947047,
      "learning_rate": 3.8056930693069306e-07,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8453793525695801,
      "num_tokens": 21296206.0,
      "step": 616
    },
    {
      "epoch": 0.11457753017641598,
      "grad_norm": 1.5287213871266645,
      "learning_rate": 3.811881188118812e-07,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8553394079208374,
      "num_tokens": 21332559.0,
      "step": 617
    },
    {
      "epoch": 0.11476323119777158,
      "grad_norm": 1.697600165215139,
      "learning_rate": 3.818069306930693e-07,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8267185688018799,
      "num_tokens": 21368009.0,
      "step": 618
    },
    {
      "epoch": 0.1149489322191272,
      "grad_norm": 1.6289119251910982,
      "learning_rate": 3.824257425742574e-07,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8392457962036133,
      "num_tokens": 21400659.0,
      "step": 619
    },
    {
      "epoch": 0.11513463324048283,
      "grad_norm": 1.7459934480873018,
      "learning_rate": 3.830445544554455e-07,
      "loss": 0.475,
      "mean_token_accuracy": 0.8476681709289551,
      "num_tokens": 21431668.0,
      "step": 620
    },
    {
      "epoch": 0.11532033426183844,
      "grad_norm": 1.4987818137220816,
      "learning_rate": 3.836633663366337e-07,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8375632762908936,
      "num_tokens": 21472639.0,
      "step": 621
    },
    {
      "epoch": 0.11550603528319406,
      "grad_norm": 1.5170100834200761,
      "learning_rate": 3.842821782178218e-07,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8302626609802246,
      "num_tokens": 21515350.0,
      "step": 622
    },
    {
      "epoch": 0.11569173630454968,
      "grad_norm": 1.7451264201961685,
      "learning_rate": 3.849009900990099e-07,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8412788510322571,
      "num_tokens": 21546031.0,
      "step": 623
    },
    {
      "epoch": 0.11587743732590529,
      "grad_norm": 1.94154801445903,
      "learning_rate": 3.85519801980198e-07,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8423408269882202,
      "num_tokens": 21573083.0,
      "step": 624
    },
    {
      "epoch": 0.11606313834726091,
      "grad_norm": 1.6475168828134417,
      "learning_rate": 3.861386138613861e-07,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8311105966567993,
      "num_tokens": 21606750.0,
      "step": 625
    },
    {
      "epoch": 0.11624883936861653,
      "grad_norm": 1.6290369591990521,
      "learning_rate": 3.8675742574257423e-07,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8452330827713013,
      "num_tokens": 21643095.0,
      "step": 626
    },
    {
      "epoch": 0.11643454038997214,
      "grad_norm": 1.4647872419114,
      "learning_rate": 3.873762376237624e-07,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8366003036499023,
      "num_tokens": 21683571.0,
      "step": 627
    },
    {
      "epoch": 0.11662024141132776,
      "grad_norm": 2.0417692373681047,
      "learning_rate": 3.879950495049505e-07,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8487105965614319,
      "num_tokens": 21705701.0,
      "step": 628
    },
    {
      "epoch": 0.11680594243268339,
      "grad_norm": 1.6018368739989397,
      "learning_rate": 3.886138613861386e-07,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8504645824432373,
      "num_tokens": 21741092.0,
      "step": 629
    },
    {
      "epoch": 0.116991643454039,
      "grad_norm": 1.4237620416912589,
      "learning_rate": 3.8923267326732673e-07,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8476289510726929,
      "num_tokens": 21782944.0,
      "step": 630
    },
    {
      "epoch": 0.11717734447539462,
      "grad_norm": 1.6843181076085212,
      "learning_rate": 3.8985148514851484e-07,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8336994647979736,
      "num_tokens": 21814654.0,
      "step": 631
    },
    {
      "epoch": 0.11736304549675024,
      "grad_norm": 1.736876871628129,
      "learning_rate": 3.9047029702970295e-07,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8514808416366577,
      "num_tokens": 21842552.0,
      "step": 632
    },
    {
      "epoch": 0.11754874651810585,
      "grad_norm": 1.60070496521316,
      "learning_rate": 3.9108910891089106e-07,
      "loss": 0.5301,
      "mean_token_accuracy": 0.832700252532959,
      "num_tokens": 21879503.0,
      "step": 633
    },
    {
      "epoch": 0.11773444753946147,
      "grad_norm": 1.7372088350412989,
      "learning_rate": 3.9170792079207923e-07,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8307701945304871,
      "num_tokens": 21909161.0,
      "step": 634
    },
    {
      "epoch": 0.11792014856081709,
      "grad_norm": 1.6102838849008412,
      "learning_rate": 3.9232673267326734e-07,
      "loss": 0.4975,
      "mean_token_accuracy": 0.839328408241272,
      "num_tokens": 21942686.0,
      "step": 635
    },
    {
      "epoch": 0.1181058495821727,
      "grad_norm": 1.4512334044939057,
      "learning_rate": 3.9294554455445545e-07,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8559722900390625,
      "num_tokens": 21981145.0,
      "step": 636
    },
    {
      "epoch": 0.11829155060352832,
      "grad_norm": 1.654798307433561,
      "learning_rate": 3.9356435643564356e-07,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8369909524917603,
      "num_tokens": 22015139.0,
      "step": 637
    },
    {
      "epoch": 0.11847725162488394,
      "grad_norm": 1.6556018565839912,
      "learning_rate": 3.9418316831683167e-07,
      "loss": 0.531,
      "mean_token_accuracy": 0.8336282968521118,
      "num_tokens": 22050131.0,
      "step": 638
    },
    {
      "epoch": 0.11866295264623955,
      "grad_norm": 1.6030562560729391,
      "learning_rate": 3.948019801980198e-07,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8452305793762207,
      "num_tokens": 22085501.0,
      "step": 639
    },
    {
      "epoch": 0.11884865366759517,
      "grad_norm": 1.4793396039806943,
      "learning_rate": 3.954207920792079e-07,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8491196036338806,
      "num_tokens": 22122597.0,
      "step": 640
    },
    {
      "epoch": 0.1190343546889508,
      "grad_norm": 1.7373332265492245,
      "learning_rate": 3.9603960396039606e-07,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8372255563735962,
      "num_tokens": 22158198.0,
      "step": 641
    },
    {
      "epoch": 0.1192200557103064,
      "grad_norm": 1.6494240676539476,
      "learning_rate": 3.9665841584158417e-07,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8401894569396973,
      "num_tokens": 22196055.0,
      "step": 642
    },
    {
      "epoch": 0.11940575673166202,
      "grad_norm": 1.584534785453547,
      "learning_rate": 3.972772277227723e-07,
      "loss": 0.561,
      "mean_token_accuracy": 0.8279135227203369,
      "num_tokens": 22237026.0,
      "step": 643
    },
    {
      "epoch": 0.11959145775301765,
      "grad_norm": 1.524592573809647,
      "learning_rate": 3.978960396039604e-07,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8467869758605957,
      "num_tokens": 22274305.0,
      "step": 644
    },
    {
      "epoch": 0.11977715877437325,
      "grad_norm": 1.6596726872288137,
      "learning_rate": 3.985148514851485e-07,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8281843066215515,
      "num_tokens": 22309734.0,
      "step": 645
    },
    {
      "epoch": 0.11996285979572888,
      "grad_norm": 1.5761207720170562,
      "learning_rate": 3.991336633663366e-07,
      "loss": 0.4935,
      "mean_token_accuracy": 0.842462420463562,
      "num_tokens": 22346351.0,
      "step": 646
    },
    {
      "epoch": 0.1201485608170845,
      "grad_norm": 1.64720167640187,
      "learning_rate": 3.997524752475248e-07,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8202509880065918,
      "num_tokens": 22383589.0,
      "step": 647
    },
    {
      "epoch": 0.1203342618384401,
      "grad_norm": 1.849552401970689,
      "learning_rate": 4.003712871287129e-07,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8424457311630249,
      "num_tokens": 22409538.0,
      "step": 648
    },
    {
      "epoch": 0.12051996285979573,
      "grad_norm": 1.73417815623364,
      "learning_rate": 4.00990099009901e-07,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8580706119537354,
      "num_tokens": 22436937.0,
      "step": 649
    },
    {
      "epoch": 0.12070566388115135,
      "grad_norm": 1.5407962289385908,
      "learning_rate": 4.016089108910891e-07,
      "loss": 0.467,
      "mean_token_accuracy": 0.8526080250740051,
      "num_tokens": 22478015.0,
      "step": 650
    },
    {
      "epoch": 0.12089136490250696,
      "grad_norm": 1.4525621776864597,
      "learning_rate": 4.022277227722772e-07,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8527768850326538,
      "num_tokens": 22518719.0,
      "step": 651
    },
    {
      "epoch": 0.12107706592386258,
      "grad_norm": 1.6952437732010281,
      "learning_rate": 4.0284653465346534e-07,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8319209814071655,
      "num_tokens": 22553046.0,
      "step": 652
    },
    {
      "epoch": 0.1212627669452182,
      "grad_norm": 1.4542706679150474,
      "learning_rate": 4.0346534653465345e-07,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8372531533241272,
      "num_tokens": 22596738.0,
      "step": 653
    },
    {
      "epoch": 0.12144846796657381,
      "grad_norm": 1.690168288352865,
      "learning_rate": 4.040841584158416e-07,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8431586027145386,
      "num_tokens": 22630732.0,
      "step": 654
    },
    {
      "epoch": 0.12163416898792943,
      "grad_norm": 1.7498319509577036,
      "learning_rate": 4.047029702970297e-07,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8430590033531189,
      "num_tokens": 22664055.0,
      "step": 655
    },
    {
      "epoch": 0.12181987000928506,
      "grad_norm": 1.5948229751733403,
      "learning_rate": 4.0532178217821783e-07,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8496311902999878,
      "num_tokens": 22700824.0,
      "step": 656
    },
    {
      "epoch": 0.12200557103064066,
      "grad_norm": 1.5997767275584118,
      "learning_rate": 4.0594059405940595e-07,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8468396663665771,
      "num_tokens": 22737365.0,
      "step": 657
    },
    {
      "epoch": 0.12219127205199629,
      "grad_norm": 1.6022579458286477,
      "learning_rate": 4.06559405940594e-07,
      "loss": 0.475,
      "mean_token_accuracy": 0.845341682434082,
      "num_tokens": 22772261.0,
      "step": 658
    },
    {
      "epoch": 0.12237697307335191,
      "grad_norm": 1.696494988746222,
      "learning_rate": 4.071782178217821e-07,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8457645177841187,
      "num_tokens": 22807028.0,
      "step": 659
    },
    {
      "epoch": 0.12256267409470752,
      "grad_norm": 1.7201547456579533,
      "learning_rate": 4.0779702970297023e-07,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8230897188186646,
      "num_tokens": 22841764.0,
      "step": 660
    },
    {
      "epoch": 0.12274837511606314,
      "grad_norm": 1.7736547179393667,
      "learning_rate": 4.084158415841584e-07,
      "loss": 0.475,
      "mean_token_accuracy": 0.8448023796081543,
      "num_tokens": 22868949.0,
      "step": 661
    },
    {
      "epoch": 0.12293407613741876,
      "grad_norm": 1.6210121843012129,
      "learning_rate": 4.090346534653465e-07,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8358273506164551,
      "num_tokens": 22902846.0,
      "step": 662
    },
    {
      "epoch": 0.12311977715877437,
      "grad_norm": 1.716405260510368,
      "learning_rate": 4.096534653465346e-07,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8296217918395996,
      "num_tokens": 22937127.0,
      "step": 663
    },
    {
      "epoch": 0.12330547818012999,
      "grad_norm": 1.7484115505975144,
      "learning_rate": 4.102722772277227e-07,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8347190618515015,
      "num_tokens": 22967286.0,
      "step": 664
    },
    {
      "epoch": 0.12349117920148561,
      "grad_norm": 1.533086301977607,
      "learning_rate": 4.1089108910891084e-07,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8585121631622314,
      "num_tokens": 23002360.0,
      "step": 665
    },
    {
      "epoch": 0.12367688022284122,
      "grad_norm": 1.5122972559508328,
      "learning_rate": 4.1150990099009895e-07,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8194084167480469,
      "num_tokens": 23044647.0,
      "step": 666
    },
    {
      "epoch": 0.12386258124419684,
      "grad_norm": 1.5889066335990272,
      "learning_rate": 4.121287128712871e-07,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8519725799560547,
      "num_tokens": 23079104.0,
      "step": 667
    },
    {
      "epoch": 0.12404828226555246,
      "grad_norm": 1.7194793846951042,
      "learning_rate": 4.127475247524752e-07,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8300142288208008,
      "num_tokens": 23110322.0,
      "step": 668
    },
    {
      "epoch": 0.12423398328690807,
      "grad_norm": 1.6809009897487326,
      "learning_rate": 4.1336633663366333e-07,
      "loss": 0.4757,
      "mean_token_accuracy": 0.847261905670166,
      "num_tokens": 23142135.0,
      "step": 669
    },
    {
      "epoch": 0.1244196843082637,
      "grad_norm": 1.7032651455685974,
      "learning_rate": 4.1398514851485145e-07,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8403000235557556,
      "num_tokens": 23173961.0,
      "step": 670
    },
    {
      "epoch": 0.12460538532961932,
      "grad_norm": 1.660179076137736,
      "learning_rate": 4.1460396039603956e-07,
      "loss": 0.4873,
      "mean_token_accuracy": 0.84149169921875,
      "num_tokens": 23206747.0,
      "step": 671
    },
    {
      "epoch": 0.12479108635097493,
      "grad_norm": 1.4618140658749925,
      "learning_rate": 4.1522277227722767e-07,
      "loss": 0.4465,
      "mean_token_accuracy": 0.853990912437439,
      "num_tokens": 23244686.0,
      "step": 672
    },
    {
      "epoch": 0.12497678737233055,
      "grad_norm": 1.635100162623303,
      "learning_rate": 4.158415841584158e-07,
      "loss": 0.5277,
      "mean_token_accuracy": 0.830819845199585,
      "num_tokens": 23277438.0,
      "step": 673
    },
    {
      "epoch": 0.12516248839368616,
      "grad_norm": 1.5261880891128785,
      "learning_rate": 4.1646039603960394e-07,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8379338383674622,
      "num_tokens": 23314321.0,
      "step": 674
    },
    {
      "epoch": 0.12534818941504178,
      "grad_norm": 1.7206984383567565,
      "learning_rate": 4.1707920792079206e-07,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8331977128982544,
      "num_tokens": 23344324.0,
      "step": 675
    },
    {
      "epoch": 0.1255338904363974,
      "grad_norm": 1.6821382213163483,
      "learning_rate": 4.1769801980198017e-07,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8444552421569824,
      "num_tokens": 23376653.0,
      "step": 676
    },
    {
      "epoch": 0.12571959145775302,
      "grad_norm": 1.46740731050643,
      "learning_rate": 4.183168316831683e-07,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8431528806686401,
      "num_tokens": 23417764.0,
      "step": 677
    },
    {
      "epoch": 0.12590529247910864,
      "grad_norm": 1.5237405690157304,
      "learning_rate": 4.189356435643564e-07,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8354339599609375,
      "num_tokens": 23459058.0,
      "step": 678
    },
    {
      "epoch": 0.12609099350046427,
      "grad_norm": 1.512970211260215,
      "learning_rate": 4.195544554455445e-07,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8508057594299316,
      "num_tokens": 23496551.0,
      "step": 679
    },
    {
      "epoch": 0.12627669452181986,
      "grad_norm": 1.4802908362001197,
      "learning_rate": 4.2017326732673266e-07,
      "loss": 0.4354,
      "mean_token_accuracy": 0.855415940284729,
      "num_tokens": 23533187.0,
      "step": 680
    },
    {
      "epoch": 0.12646239554317548,
      "grad_norm": 1.6331102388842285,
      "learning_rate": 4.207920792079208e-07,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8572193384170532,
      "num_tokens": 23564630.0,
      "step": 681
    },
    {
      "epoch": 0.1266480965645311,
      "grad_norm": 1.6623900888705767,
      "learning_rate": 4.214108910891089e-07,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8261546492576599,
      "num_tokens": 23601445.0,
      "step": 682
    },
    {
      "epoch": 0.12683379758588673,
      "grad_norm": 1.8933659833404908,
      "learning_rate": 4.22029702970297e-07,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8255757093429565,
      "num_tokens": 23634278.0,
      "step": 683
    },
    {
      "epoch": 0.12701949860724235,
      "grad_norm": 1.6201449552346103,
      "learning_rate": 4.226485148514851e-07,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8375473022460938,
      "num_tokens": 23670717.0,
      "step": 684
    },
    {
      "epoch": 0.12720519962859797,
      "grad_norm": 1.5730687041802012,
      "learning_rate": 4.232673267326732e-07,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8572812080383301,
      "num_tokens": 23705810.0,
      "step": 685
    },
    {
      "epoch": 0.12739090064995356,
      "grad_norm": 1.4612304775151728,
      "learning_rate": 4.2388613861386133e-07,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8570245504379272,
      "num_tokens": 23746161.0,
      "step": 686
    },
    {
      "epoch": 0.1275766016713092,
      "grad_norm": 1.5279448619404836,
      "learning_rate": 4.245049504950495e-07,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8272365927696228,
      "num_tokens": 23788622.0,
      "step": 687
    },
    {
      "epoch": 0.1277623026926648,
      "grad_norm": 1.6235503570108523,
      "learning_rate": 4.251237623762376e-07,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8397685289382935,
      "num_tokens": 23823005.0,
      "step": 688
    },
    {
      "epoch": 0.12794800371402043,
      "grad_norm": 1.5895026671884418,
      "learning_rate": 4.257425742574257e-07,
      "loss": 0.4563,
      "mean_token_accuracy": 0.851302981376648,
      "num_tokens": 23855149.0,
      "step": 689
    },
    {
      "epoch": 0.12813370473537605,
      "grad_norm": 1.5859411174969569,
      "learning_rate": 4.2636138613861383e-07,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8369623422622681,
      "num_tokens": 23887429.0,
      "step": 690
    },
    {
      "epoch": 0.12831940575673167,
      "grad_norm": 1.5030292449750384,
      "learning_rate": 4.2698019801980194e-07,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8475391864776611,
      "num_tokens": 23924633.0,
      "step": 691
    },
    {
      "epoch": 0.12850510677808727,
      "grad_norm": 1.4281238876845785,
      "learning_rate": 4.2759900990099005e-07,
      "loss": 0.4534,
      "mean_token_accuracy": 0.855783224105835,
      "num_tokens": 23966550.0,
      "step": 692
    },
    {
      "epoch": 0.1286908077994429,
      "grad_norm": 1.544990762055361,
      "learning_rate": 4.2821782178217816e-07,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8415892720222473,
      "num_tokens": 24004948.0,
      "step": 693
    },
    {
      "epoch": 0.1288765088207985,
      "grad_norm": 1.5913078374261003,
      "learning_rate": 4.2883663366336633e-07,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8389744758605957,
      "num_tokens": 24041358.0,
      "step": 694
    },
    {
      "epoch": 0.12906220984215414,
      "grad_norm": 1.752723392831651,
      "learning_rate": 4.2945544554455444e-07,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8439443707466125,
      "num_tokens": 24071076.0,
      "step": 695
    },
    {
      "epoch": 0.12924791086350976,
      "grad_norm": 1.7576437440720447,
      "learning_rate": 4.3007425742574255e-07,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8296775817871094,
      "num_tokens": 24103065.0,
      "step": 696
    },
    {
      "epoch": 0.12943361188486538,
      "grad_norm": 1.640738446221668,
      "learning_rate": 4.3069306930693066e-07,
      "loss": 0.4606,
      "mean_token_accuracy": 0.851344108581543,
      "num_tokens": 24134651.0,
      "step": 697
    },
    {
      "epoch": 0.12961931290622097,
      "grad_norm": 1.729601640664512,
      "learning_rate": 4.313118811881188e-07,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8482546210289001,
      "num_tokens": 24166556.0,
      "step": 698
    },
    {
      "epoch": 0.1298050139275766,
      "grad_norm": 1.7649777588999713,
      "learning_rate": 4.319306930693069e-07,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8438771367073059,
      "num_tokens": 24202462.0,
      "step": 699
    },
    {
      "epoch": 0.12999071494893222,
      "grad_norm": 1.6572359952595541,
      "learning_rate": 4.3254950495049505e-07,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8519357442855835,
      "num_tokens": 24230114.0,
      "step": 700
    },
    {
      "epoch": 0.13017641597028784,
      "grad_norm": 1.5331836018404692,
      "learning_rate": 4.3316831683168316e-07,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8377468585968018,
      "num_tokens": 24268179.0,
      "step": 701
    },
    {
      "epoch": 0.13036211699164346,
      "grad_norm": 1.7131518895041722,
      "learning_rate": 4.3378712871287127e-07,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8381304144859314,
      "num_tokens": 24300411.0,
      "step": 702
    },
    {
      "epoch": 0.13054781801299908,
      "grad_norm": 1.571312023128133,
      "learning_rate": 4.344059405940594e-07,
      "loss": 0.434,
      "mean_token_accuracy": 0.8590437173843384,
      "num_tokens": 24333203.0,
      "step": 703
    },
    {
      "epoch": 0.13073351903435468,
      "grad_norm": 1.5060629710176947,
      "learning_rate": 4.350247524752475e-07,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8471345901489258,
      "num_tokens": 24372230.0,
      "step": 704
    },
    {
      "epoch": 0.1309192200557103,
      "grad_norm": 1.518661902441678,
      "learning_rate": 4.356435643564356e-07,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8471053242683411,
      "num_tokens": 24410569.0,
      "step": 705
    },
    {
      "epoch": 0.13110492107706592,
      "grad_norm": 1.8073296421936846,
      "learning_rate": 4.362623762376237e-07,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8403275609016418,
      "num_tokens": 24441255.0,
      "step": 706
    },
    {
      "epoch": 0.13129062209842154,
      "grad_norm": 1.7045923651930162,
      "learning_rate": 4.368811881188119e-07,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8373216390609741,
      "num_tokens": 24472848.0,
      "step": 707
    },
    {
      "epoch": 0.13147632311977717,
      "grad_norm": 1.7494530062214317,
      "learning_rate": 4.375e-07,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8221547603607178,
      "num_tokens": 24505152.0,
      "step": 708
    },
    {
      "epoch": 0.1316620241411328,
      "grad_norm": 1.5191566698270889,
      "learning_rate": 4.381188118811881e-07,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8478728532791138,
      "num_tokens": 24542166.0,
      "step": 709
    },
    {
      "epoch": 0.13184772516248838,
      "grad_norm": 1.6310922554315956,
      "learning_rate": 4.387376237623762e-07,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8430383205413818,
      "num_tokens": 24575912.0,
      "step": 710
    },
    {
      "epoch": 0.132033426183844,
      "grad_norm": 1.5240606388840963,
      "learning_rate": 4.3935643564356433e-07,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8513296842575073,
      "num_tokens": 24613501.0,
      "step": 711
    },
    {
      "epoch": 0.13221912720519963,
      "grad_norm": 1.5713936715340062,
      "learning_rate": 4.3997524752475244e-07,
      "loss": 0.4823,
      "mean_token_accuracy": 0.846743106842041,
      "num_tokens": 24646396.0,
      "step": 712
    },
    {
      "epoch": 0.13240482822655525,
      "grad_norm": 1.8868978117297175,
      "learning_rate": 4.405940594059406e-07,
      "loss": 0.5304,
      "mean_token_accuracy": 0.830804705619812,
      "num_tokens": 24673578.0,
      "step": 713
    },
    {
      "epoch": 0.13259052924791087,
      "grad_norm": 1.4390828679415637,
      "learning_rate": 4.412128712871287e-07,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8440700173377991,
      "num_tokens": 24714145.0,
      "step": 714
    },
    {
      "epoch": 0.1327762302692665,
      "grad_norm": 1.503440993153239,
      "learning_rate": 4.418316831683168e-07,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8486253619194031,
      "num_tokens": 24750478.0,
      "step": 715
    },
    {
      "epoch": 0.1329619312906221,
      "grad_norm": 1.5020483612736135,
      "learning_rate": 4.4245049504950494e-07,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8363718390464783,
      "num_tokens": 24788782.0,
      "step": 716
    },
    {
      "epoch": 0.1331476323119777,
      "grad_norm": 1.4517551971580231,
      "learning_rate": 4.4306930693069305e-07,
      "loss": 0.5347,
      "mean_token_accuracy": 0.8297842741012573,
      "num_tokens": 24831737.0,
      "step": 717
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 1.7223414692079773,
      "learning_rate": 4.4368811881188116e-07,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8496747016906738,
      "num_tokens": 24859671.0,
      "step": 718
    },
    {
      "epoch": 0.13351903435468895,
      "grad_norm": 1.4201464080709307,
      "learning_rate": 4.4430693069306927e-07,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8535256385803223,
      "num_tokens": 24899132.0,
      "step": 719
    },
    {
      "epoch": 0.13370473537604458,
      "grad_norm": 1.7257985956064887,
      "learning_rate": 4.4492574257425744e-07,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8301029205322266,
      "num_tokens": 24930796.0,
      "step": 720
    },
    {
      "epoch": 0.1338904363974002,
      "grad_norm": 1.5948791200315104,
      "learning_rate": 4.4554455445544555e-07,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8371648192405701,
      "num_tokens": 24963844.0,
      "step": 721
    },
    {
      "epoch": 0.1340761374187558,
      "grad_norm": 1.6188913967853704,
      "learning_rate": 4.4616336633663366e-07,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8281913995742798,
      "num_tokens": 24998941.0,
      "step": 722
    },
    {
      "epoch": 0.13426183844011141,
      "grad_norm": 1.6632499306909356,
      "learning_rate": 4.4678217821782177e-07,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8355874419212341,
      "num_tokens": 25032875.0,
      "step": 723
    },
    {
      "epoch": 0.13444753946146704,
      "grad_norm": 1.5630837764448111,
      "learning_rate": 4.474009900990099e-07,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8584202527999878,
      "num_tokens": 25070088.0,
      "step": 724
    },
    {
      "epoch": 0.13463324048282266,
      "grad_norm": 1.6511102323125548,
      "learning_rate": 4.48019801980198e-07,
      "loss": 0.4624,
      "mean_token_accuracy": 0.850671648979187,
      "num_tokens": 25104970.0,
      "step": 725
    },
    {
      "epoch": 0.13481894150417828,
      "grad_norm": 1.5162701978116442,
      "learning_rate": 4.486386138613861e-07,
      "loss": 0.5168,
      "mean_token_accuracy": 0.834302544593811,
      "num_tokens": 25144405.0,
      "step": 726
    },
    {
      "epoch": 0.1350046425255339,
      "grad_norm": 1.848832500596359,
      "learning_rate": 4.4925742574257427e-07,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8367363214492798,
      "num_tokens": 25174910.0,
      "step": 727
    },
    {
      "epoch": 0.1351903435468895,
      "grad_norm": 1.694559570955194,
      "learning_rate": 4.498762376237624e-07,
      "loss": 0.4722,
      "mean_token_accuracy": 0.854051411151886,
      "num_tokens": 25204938.0,
      "step": 728
    },
    {
      "epoch": 0.13537604456824512,
      "grad_norm": 1.676386461079287,
      "learning_rate": 4.504950495049505e-07,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8421258926391602,
      "num_tokens": 25238394.0,
      "step": 729
    },
    {
      "epoch": 0.13556174558960074,
      "grad_norm": 1.4849476391564806,
      "learning_rate": 4.511138613861386e-07,
      "loss": 0.4959,
      "mean_token_accuracy": 0.841493546962738,
      "num_tokens": 25281936.0,
      "step": 730
    },
    {
      "epoch": 0.13574744661095636,
      "grad_norm": 1.6279822920259237,
      "learning_rate": 4.517326732673267e-07,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8322476148605347,
      "num_tokens": 25314171.0,
      "step": 731
    },
    {
      "epoch": 0.13593314763231198,
      "grad_norm": 1.5606482896100455,
      "learning_rate": 4.523514851485148e-07,
      "loss": 0.504,
      "mean_token_accuracy": 0.8408046960830688,
      "num_tokens": 25353300.0,
      "step": 732
    },
    {
      "epoch": 0.1361188486536676,
      "grad_norm": 1.659961217586374,
      "learning_rate": 4.52970297029703e-07,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8476388454437256,
      "num_tokens": 25386260.0,
      "step": 733
    },
    {
      "epoch": 0.1363045496750232,
      "grad_norm": 1.6885981797055971,
      "learning_rate": 4.535891089108911e-07,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8405256867408752,
      "num_tokens": 25417716.0,
      "step": 734
    },
    {
      "epoch": 0.13649025069637882,
      "grad_norm": 1.5636468969064268,
      "learning_rate": 4.542079207920792e-07,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8456074595451355,
      "num_tokens": 25453397.0,
      "step": 735
    },
    {
      "epoch": 0.13667595171773445,
      "grad_norm": 1.550281575434872,
      "learning_rate": 4.548267326732673e-07,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8374041318893433,
      "num_tokens": 25492071.0,
      "step": 736
    },
    {
      "epoch": 0.13686165273909007,
      "grad_norm": 1.531401887554325,
      "learning_rate": 4.5544554455445543e-07,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8402677774429321,
      "num_tokens": 25530697.0,
      "step": 737
    },
    {
      "epoch": 0.1370473537604457,
      "grad_norm": 1.5196001769136631,
      "learning_rate": 4.5606435643564354e-07,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8523105382919312,
      "num_tokens": 25565472.0,
      "step": 738
    },
    {
      "epoch": 0.1372330547818013,
      "grad_norm": 1.5281206775475338,
      "learning_rate": 4.5668316831683166e-07,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8549139499664307,
      "num_tokens": 25598924.0,
      "step": 739
    },
    {
      "epoch": 0.1374187558031569,
      "grad_norm": 1.5134119463070217,
      "learning_rate": 4.573019801980198e-07,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8610981106758118,
      "num_tokens": 25632399.0,
      "step": 740
    },
    {
      "epoch": 0.13760445682451253,
      "grad_norm": 1.637243707083049,
      "learning_rate": 4.5792079207920793e-07,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8463116884231567,
      "num_tokens": 25664324.0,
      "step": 741
    },
    {
      "epoch": 0.13779015784586815,
      "grad_norm": 1.5957904853247464,
      "learning_rate": 4.5853960396039604e-07,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8384634256362915,
      "num_tokens": 25700635.0,
      "step": 742
    },
    {
      "epoch": 0.13797585886722377,
      "grad_norm": 1.6349171728079033,
      "learning_rate": 4.5915841584158415e-07,
      "loss": 0.442,
      "mean_token_accuracy": 0.8561135530471802,
      "num_tokens": 25732718.0,
      "step": 743
    },
    {
      "epoch": 0.1381615598885794,
      "grad_norm": 1.5405856953877057,
      "learning_rate": 4.5977722772277227e-07,
      "loss": 0.5003,
      "mean_token_accuracy": 0.841012179851532,
      "num_tokens": 25773146.0,
      "step": 744
    },
    {
      "epoch": 0.13834726090993502,
      "grad_norm": 1.6667419003879196,
      "learning_rate": 4.603960396039604e-07,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8448835611343384,
      "num_tokens": 25806332.0,
      "step": 745
    },
    {
      "epoch": 0.1385329619312906,
      "grad_norm": 1.4810025460543395,
      "learning_rate": 4.6101485148514854e-07,
      "loss": 0.493,
      "mean_token_accuracy": 0.8403671979904175,
      "num_tokens": 25844751.0,
      "step": 746
    },
    {
      "epoch": 0.13871866295264623,
      "grad_norm": 1.5776953175871606,
      "learning_rate": 4.6163366336633665e-07,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8540582656860352,
      "num_tokens": 25876863.0,
      "step": 747
    },
    {
      "epoch": 0.13890436397400185,
      "grad_norm": 1.5241176468856656,
      "learning_rate": 4.6225247524752476e-07,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8327944278717041,
      "num_tokens": 25915786.0,
      "step": 748
    },
    {
      "epoch": 0.13909006499535748,
      "grad_norm": 1.5577188621403295,
      "learning_rate": 4.628712871287129e-07,
      "loss": 0.508,
      "mean_token_accuracy": 0.8358557224273682,
      "num_tokens": 25952093.0,
      "step": 749
    },
    {
      "epoch": 0.1392757660167131,
      "grad_norm": 1.6236924913693553,
      "learning_rate": 4.63490099009901e-07,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8368220925331116,
      "num_tokens": 25987219.0,
      "step": 750
    },
    {
      "epoch": 0.13946146703806872,
      "grad_norm": 1.5893380951145164,
      "learning_rate": 4.641089108910891e-07,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8463961482048035,
      "num_tokens": 26025244.0,
      "step": 751
    },
    {
      "epoch": 0.13964716805942431,
      "grad_norm": 1.7692736192815373,
      "learning_rate": 4.647277227722772e-07,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8456005454063416,
      "num_tokens": 26053194.0,
      "step": 752
    },
    {
      "epoch": 0.13983286908077994,
      "grad_norm": 1.6512416139412072,
      "learning_rate": 4.6534653465346537e-07,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8452390432357788,
      "num_tokens": 26087216.0,
      "step": 753
    },
    {
      "epoch": 0.14001857010213556,
      "grad_norm": 1.4323872769940742,
      "learning_rate": 4.659653465346535e-07,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8408674597740173,
      "num_tokens": 26130305.0,
      "step": 754
    },
    {
      "epoch": 0.14020427112349118,
      "grad_norm": 1.4110878090534271,
      "learning_rate": 4.665841584158416e-07,
      "loss": 0.4724,
      "mean_token_accuracy": 0.847908616065979,
      "num_tokens": 26171562.0,
      "step": 755
    },
    {
      "epoch": 0.1403899721448468,
      "grad_norm": 1.7965824669678372,
      "learning_rate": 4.672029702970297e-07,
      "loss": 0.5204,
      "mean_token_accuracy": 0.834976077079773,
      "num_tokens": 26203434.0,
      "step": 756
    },
    {
      "epoch": 0.14057567316620243,
      "grad_norm": 1.4299450556067008,
      "learning_rate": 4.678217821782178e-07,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8420565128326416,
      "num_tokens": 26242994.0,
      "step": 757
    },
    {
      "epoch": 0.14076137418755802,
      "grad_norm": 1.4706878547730367,
      "learning_rate": 4.6844059405940593e-07,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8533326983451843,
      "num_tokens": 26281385.0,
      "step": 758
    },
    {
      "epoch": 0.14094707520891364,
      "grad_norm": 1.6669791373247216,
      "learning_rate": 4.69059405940594e-07,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8343981504440308,
      "num_tokens": 26314761.0,
      "step": 759
    },
    {
      "epoch": 0.14113277623026926,
      "grad_norm": 1.4527258319232417,
      "learning_rate": 4.6967821782178215e-07,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8547177314758301,
      "num_tokens": 26354411.0,
      "step": 760
    },
    {
      "epoch": 0.14131847725162489,
      "grad_norm": 1.50101282776237,
      "learning_rate": 4.7029702970297026e-07,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8536601662635803,
      "num_tokens": 26391549.0,
      "step": 761
    },
    {
      "epoch": 0.1415041782729805,
      "grad_norm": 1.6869502815894457,
      "learning_rate": 4.709158415841584e-07,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8474026322364807,
      "num_tokens": 26422147.0,
      "step": 762
    },
    {
      "epoch": 0.14168987929433613,
      "grad_norm": 1.5703788201981244,
      "learning_rate": 4.715346534653465e-07,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8333956003189087,
      "num_tokens": 26457475.0,
      "step": 763
    },
    {
      "epoch": 0.14187558031569172,
      "grad_norm": 1.6796613571041759,
      "learning_rate": 4.721534653465346e-07,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8393558263778687,
      "num_tokens": 26490555.0,
      "step": 764
    },
    {
      "epoch": 0.14206128133704735,
      "grad_norm": 1.476313557901013,
      "learning_rate": 4.727722772277227e-07,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8421309590339661,
      "num_tokens": 26527438.0,
      "step": 765
    },
    {
      "epoch": 0.14224698235840297,
      "grad_norm": 1.5355119413603175,
      "learning_rate": 4.733910891089108e-07,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8498889207839966,
      "num_tokens": 26563185.0,
      "step": 766
    },
    {
      "epoch": 0.1424326833797586,
      "grad_norm": 1.5684174039450867,
      "learning_rate": 4.74009900990099e-07,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8515808582305908,
      "num_tokens": 26597486.0,
      "step": 767
    },
    {
      "epoch": 0.1426183844011142,
      "grad_norm": 1.6114223984881857,
      "learning_rate": 4.746287128712871e-07,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8462027311325073,
      "num_tokens": 26630596.0,
      "step": 768
    },
    {
      "epoch": 0.14280408542246983,
      "grad_norm": 1.531473449160037,
      "learning_rate": 4.752475247524752e-07,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8422735929489136,
      "num_tokens": 26669045.0,
      "step": 769
    },
    {
      "epoch": 0.14298978644382543,
      "grad_norm": 1.711564902352359,
      "learning_rate": 4.758663366336633e-07,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8238126635551453,
      "num_tokens": 26703004.0,
      "step": 770
    },
    {
      "epoch": 0.14317548746518105,
      "grad_norm": 1.577646975841212,
      "learning_rate": 4.7648514851485143e-07,
      "loss": 0.4538,
      "mean_token_accuracy": 0.849680483341217,
      "num_tokens": 26736284.0,
      "step": 771
    },
    {
      "epoch": 0.14336118848653667,
      "grad_norm": 1.5623322302109146,
      "learning_rate": 4.771039603960396e-07,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8635645508766174,
      "num_tokens": 26768897.0,
      "step": 772
    },
    {
      "epoch": 0.1435468895078923,
      "grad_norm": 1.6248207382444744,
      "learning_rate": 4.777227722772277e-07,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8461304306983948,
      "num_tokens": 26801240.0,
      "step": 773
    },
    {
      "epoch": 0.14373259052924792,
      "grad_norm": 1.4716283847150249,
      "learning_rate": 4.783415841584158e-07,
      "loss": 0.459,
      "mean_token_accuracy": 0.8530882000923157,
      "num_tokens": 26837878.0,
      "step": 774
    },
    {
      "epoch": 0.14391829155060354,
      "grad_norm": 1.6406111930063785,
      "learning_rate": 4.789603960396039e-07,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8313868045806885,
      "num_tokens": 26872868.0,
      "step": 775
    },
    {
      "epoch": 0.14410399257195913,
      "grad_norm": 1.6845528880076686,
      "learning_rate": 4.79579207920792e-07,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8383744955062866,
      "num_tokens": 26908135.0,
      "step": 776
    },
    {
      "epoch": 0.14428969359331476,
      "grad_norm": 1.5464808886889598,
      "learning_rate": 4.801980198019802e-07,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8471494913101196,
      "num_tokens": 26943663.0,
      "step": 777
    },
    {
      "epoch": 0.14447539461467038,
      "grad_norm": 1.8676831304180614,
      "learning_rate": 4.808168316831683e-07,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8589786887168884,
      "num_tokens": 26967737.0,
      "step": 778
    },
    {
      "epoch": 0.144661095636026,
      "grad_norm": 1.7806701559513407,
      "learning_rate": 4.814356435643564e-07,
      "loss": 0.487,
      "mean_token_accuracy": 0.8469600081443787,
      "num_tokens": 26996349.0,
      "step": 779
    },
    {
      "epoch": 0.14484679665738162,
      "grad_norm": 1.6210084809924097,
      "learning_rate": 4.820544554455445e-07,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8407395482063293,
      "num_tokens": 27030376.0,
      "step": 780
    },
    {
      "epoch": 0.14503249767873724,
      "grad_norm": 1.7838360630194832,
      "learning_rate": 4.826732673267326e-07,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8431200981140137,
      "num_tokens": 27059275.0,
      "step": 781
    },
    {
      "epoch": 0.14521819870009284,
      "grad_norm": 1.6253758417655837,
      "learning_rate": 4.832920792079207e-07,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8463203310966492,
      "num_tokens": 27093217.0,
      "step": 782
    },
    {
      "epoch": 0.14540389972144846,
      "grad_norm": 1.5793893507727739,
      "learning_rate": 4.839108910891089e-07,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8526239395141602,
      "num_tokens": 27127912.0,
      "step": 783
    },
    {
      "epoch": 0.14558960074280408,
      "grad_norm": 1.5644134840813597,
      "learning_rate": 4.84529702970297e-07,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8494949340820312,
      "num_tokens": 27161896.0,
      "step": 784
    },
    {
      "epoch": 0.1457753017641597,
      "grad_norm": 1.5720868299819954,
      "learning_rate": 4.851485148514851e-07,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8400108814239502,
      "num_tokens": 27198981.0,
      "step": 785
    },
    {
      "epoch": 0.14596100278551533,
      "grad_norm": 1.644568313137088,
      "learning_rate": 4.857673267326733e-07,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8288445472717285,
      "num_tokens": 27232638.0,
      "step": 786
    },
    {
      "epoch": 0.14614670380687095,
      "grad_norm": 1.6086138922273738,
      "learning_rate": 4.863861386138613e-07,
      "loss": 0.4652,
      "mean_token_accuracy": 0.850860595703125,
      "num_tokens": 27267465.0,
      "step": 787
    },
    {
      "epoch": 0.14633240482822654,
      "grad_norm": 1.5198321562955206,
      "learning_rate": 4.870049504950495e-07,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8428761959075928,
      "num_tokens": 27303035.0,
      "step": 788
    },
    {
      "epoch": 0.14651810584958216,
      "grad_norm": 1.6186022844560453,
      "learning_rate": 4.876237623762375e-07,
      "loss": 0.5005,
      "mean_token_accuracy": 0.837733268737793,
      "num_tokens": 27341765.0,
      "step": 789
    },
    {
      "epoch": 0.1467038068709378,
      "grad_norm": 1.5321199112559223,
      "learning_rate": 4.882425742574257e-07,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8516373038291931,
      "num_tokens": 27375396.0,
      "step": 790
    },
    {
      "epoch": 0.1468895078922934,
      "grad_norm": 1.6251251682874175,
      "learning_rate": 4.888613861386139e-07,
      "loss": 0.4892,
      "mean_token_accuracy": 0.839624285697937,
      "num_tokens": 27407112.0,
      "step": 791
    },
    {
      "epoch": 0.14707520891364903,
      "grad_norm": 1.6808379719182107,
      "learning_rate": 4.894801980198019e-07,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8489964604377747,
      "num_tokens": 27439293.0,
      "step": 792
    },
    {
      "epoch": 0.14726090993500465,
      "grad_norm": 1.635584212312479,
      "learning_rate": 4.900990099009901e-07,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8666260838508606,
      "num_tokens": 27470408.0,
      "step": 793
    },
    {
      "epoch": 0.14744661095636025,
      "grad_norm": 1.79244556817016,
      "learning_rate": 4.907178217821781e-07,
      "loss": 0.483,
      "mean_token_accuracy": 0.8416004776954651,
      "num_tokens": 27501387.0,
      "step": 794
    },
    {
      "epoch": 0.14763231197771587,
      "grad_norm": 1.5951318047464187,
      "learning_rate": 4.913366336633663e-07,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8478710055351257,
      "num_tokens": 27535512.0,
      "step": 795
    },
    {
      "epoch": 0.1478180129990715,
      "grad_norm": 1.7722803637833962,
      "learning_rate": 4.919554455445545e-07,
      "loss": 0.46,
      "mean_token_accuracy": 0.850415825843811,
      "num_tokens": 27563164.0,
      "step": 796
    },
    {
      "epoch": 0.1480037140204271,
      "grad_norm": 1.439389229162918,
      "learning_rate": 4.925742574257425e-07,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8340264558792114,
      "num_tokens": 27608623.0,
      "step": 797
    },
    {
      "epoch": 0.14818941504178273,
      "grad_norm": 1.5573137574299272,
      "learning_rate": 4.931930693069307e-07,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8452757596969604,
      "num_tokens": 27645983.0,
      "step": 798
    },
    {
      "epoch": 0.14837511606313836,
      "grad_norm": 1.9869608241966152,
      "learning_rate": 4.938118811881188e-07,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8438234329223633,
      "num_tokens": 27667461.0,
      "step": 799
    },
    {
      "epoch": 0.14856081708449395,
      "grad_norm": 1.596016768996986,
      "learning_rate": 4.944306930693069e-07,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8532919883728027,
      "num_tokens": 27704095.0,
      "step": 800
    },
    {
      "epoch": 0.14874651810584957,
      "grad_norm": 1.5888149998672145,
      "learning_rate": 4.95049504950495e-07,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8436137437820435,
      "num_tokens": 27738645.0,
      "step": 801
    },
    {
      "epoch": 0.1489322191272052,
      "grad_norm": 1.62489923684428,
      "learning_rate": 4.956683168316831e-07,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8363210558891296,
      "num_tokens": 27772702.0,
      "step": 802
    },
    {
      "epoch": 0.14911792014856082,
      "grad_norm": 1.5416497972956298,
      "learning_rate": 4.962871287128713e-07,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8254544138908386,
      "num_tokens": 27812708.0,
      "step": 803
    },
    {
      "epoch": 0.14930362116991644,
      "grad_norm": 1.628617026212216,
      "learning_rate": 4.969059405940594e-07,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8410710096359253,
      "num_tokens": 27846866.0,
      "step": 804
    },
    {
      "epoch": 0.14948932219127206,
      "grad_norm": 1.7714828196468726,
      "learning_rate": 4.975247524752475e-07,
      "loss": 0.496,
      "mean_token_accuracy": 0.8409150838851929,
      "num_tokens": 27874039.0,
      "step": 805
    },
    {
      "epoch": 0.14967502321262766,
      "grad_norm": 1.3997638815637747,
      "learning_rate": 4.981435643564356e-07,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8463294506072998,
      "num_tokens": 27917363.0,
      "step": 806
    },
    {
      "epoch": 0.14986072423398328,
      "grad_norm": 1.5269778194544918,
      "learning_rate": 4.987623762376238e-07,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8419116735458374,
      "num_tokens": 27952854.0,
      "step": 807
    },
    {
      "epoch": 0.1500464252553389,
      "grad_norm": 1.5811994488454462,
      "learning_rate": 4.993811881188118e-07,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8433588147163391,
      "num_tokens": 27985909.0,
      "step": 808
    },
    {
      "epoch": 0.15023212627669452,
      "grad_norm": 1.6707476969947341,
      "learning_rate": 5e-07,
      "loss": 0.4975,
      "mean_token_accuracy": 0.834456741809845,
      "num_tokens": 28017115.0,
      "step": 809
    },
    {
      "epoch": 0.15041782729805014,
      "grad_norm": 1.631797972194364,
      "learning_rate": 5.00618811881188e-07,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8470873832702637,
      "num_tokens": 28048254.0,
      "step": 810
    },
    {
      "epoch": 0.15060352831940577,
      "grad_norm": 2.0168803327515414,
      "learning_rate": 5.012376237623762e-07,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8387752771377563,
      "num_tokens": 28083299.0,
      "step": 811
    },
    {
      "epoch": 0.15078922934076136,
      "grad_norm": 1.5886983281225087,
      "learning_rate": 5.018564356435643e-07,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8231400847434998,
      "num_tokens": 28120157.0,
      "step": 812
    },
    {
      "epoch": 0.15097493036211698,
      "grad_norm": 1.7637720290865604,
      "learning_rate": 5.024752475247524e-07,
      "loss": 0.5202,
      "mean_token_accuracy": 0.834968090057373,
      "num_tokens": 28149830.0,
      "step": 813
    },
    {
      "epoch": 0.1511606313834726,
      "grad_norm": 1.5244653542969873,
      "learning_rate": 5.030940594059405e-07,
      "loss": 0.48,
      "mean_token_accuracy": 0.8420566320419312,
      "num_tokens": 28186015.0,
      "step": 814
    },
    {
      "epoch": 0.15134633240482823,
      "grad_norm": 1.6659752876178797,
      "learning_rate": 5.037128712871286e-07,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8669904470443726,
      "num_tokens": 28213460.0,
      "step": 815
    },
    {
      "epoch": 0.15153203342618385,
      "grad_norm": 1.7368940577718737,
      "learning_rate": 5.043316831683168e-07,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8455349206924438,
      "num_tokens": 28241529.0,
      "step": 816
    },
    {
      "epoch": 0.15171773444753947,
      "grad_norm": 1.5422921741532267,
      "learning_rate": 5.04950495049505e-07,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8322489857673645,
      "num_tokens": 28282360.0,
      "step": 817
    },
    {
      "epoch": 0.1519034354688951,
      "grad_norm": 1.5464297911728837,
      "learning_rate": 5.05569306930693e-07,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8614715337753296,
      "num_tokens": 28318112.0,
      "step": 818
    },
    {
      "epoch": 0.1520891364902507,
      "grad_norm": 1.6031781524234494,
      "learning_rate": 5.061881188118812e-07,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8235148191452026,
      "num_tokens": 28353397.0,
      "step": 819
    },
    {
      "epoch": 0.1522748375116063,
      "grad_norm": 1.6622732506071614,
      "learning_rate": 5.068069306930693e-07,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8626890182495117,
      "num_tokens": 28383807.0,
      "step": 820
    },
    {
      "epoch": 0.15246053853296193,
      "grad_norm": 1.5189763858499814,
      "learning_rate": 5.074257425742574e-07,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8512115478515625,
      "num_tokens": 28420546.0,
      "step": 821
    },
    {
      "epoch": 0.15264623955431755,
      "grad_norm": 1.4260603297384806,
      "learning_rate": 5.080445544554455e-07,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8522376418113708,
      "num_tokens": 28459363.0,
      "step": 822
    },
    {
      "epoch": 0.15283194057567318,
      "grad_norm": 1.5787824395201004,
      "learning_rate": 5.086633663366336e-07,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8393505811691284,
      "num_tokens": 28493559.0,
      "step": 823
    },
    {
      "epoch": 0.1530176415970288,
      "grad_norm": 1.5669940758918013,
      "learning_rate": 5.092821782178217e-07,
      "loss": 0.495,
      "mean_token_accuracy": 0.8410491943359375,
      "num_tokens": 28529290.0,
      "step": 824
    },
    {
      "epoch": 0.1532033426183844,
      "grad_norm": 1.5056248899360098,
      "learning_rate": 5.099009900990099e-07,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8573048710823059,
      "num_tokens": 28562925.0,
      "step": 825
    },
    {
      "epoch": 0.15338904363974,
      "grad_norm": 1.5958769887761133,
      "learning_rate": 5.105198019801979e-07,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8442868590354919,
      "num_tokens": 28592912.0,
      "step": 826
    },
    {
      "epoch": 0.15357474466109564,
      "grad_norm": 1.6022074581513968,
      "learning_rate": 5.111386138613861e-07,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8376154899597168,
      "num_tokens": 28628434.0,
      "step": 827
    },
    {
      "epoch": 0.15376044568245126,
      "grad_norm": 1.4136507623496706,
      "learning_rate": 5.117574257425741e-07,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8394026756286621,
      "num_tokens": 28669058.0,
      "step": 828
    },
    {
      "epoch": 0.15394614670380688,
      "grad_norm": 1.6360276014596093,
      "learning_rate": 5.123762376237624e-07,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8565691709518433,
      "num_tokens": 28703149.0,
      "step": 829
    },
    {
      "epoch": 0.1541318477251625,
      "grad_norm": 1.6146690247896798,
      "learning_rate": 5.129950495049505e-07,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8357638120651245,
      "num_tokens": 28736375.0,
      "step": 830
    },
    {
      "epoch": 0.1543175487465181,
      "grad_norm": 1.4951269886751384,
      "learning_rate": 5.136138613861386e-07,
      "loss": 0.501,
      "mean_token_accuracy": 0.8372804522514343,
      "num_tokens": 28780822.0,
      "step": 831
    },
    {
      "epoch": 0.15450324976787372,
      "grad_norm": 1.6810600727858176,
      "learning_rate": 5.142326732673267e-07,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8442226648330688,
      "num_tokens": 28813507.0,
      "step": 832
    },
    {
      "epoch": 0.15468895078922934,
      "grad_norm": 1.4977296508009252,
      "learning_rate": 5.148514851485149e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8508682250976562,
      "num_tokens": 28847214.0,
      "step": 833
    },
    {
      "epoch": 0.15487465181058496,
      "grad_norm": 1.5245648525691877,
      "learning_rate": 5.154702970297029e-07,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8575626611709595,
      "num_tokens": 28881229.0,
      "step": 834
    },
    {
      "epoch": 0.15506035283194058,
      "grad_norm": 1.525611872199361,
      "learning_rate": 5.160891089108911e-07,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8361001014709473,
      "num_tokens": 28923496.0,
      "step": 835
    },
    {
      "epoch": 0.1552460538532962,
      "grad_norm": 1.6071132363823977,
      "learning_rate": 5.167079207920791e-07,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8443470597267151,
      "num_tokens": 28958387.0,
      "step": 836
    },
    {
      "epoch": 0.1554317548746518,
      "grad_norm": 1.6074587595034873,
      "learning_rate": 5.173267326732673e-07,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8561277389526367,
      "num_tokens": 28991312.0,
      "step": 837
    },
    {
      "epoch": 0.15561745589600742,
      "grad_norm": 1.7307889609634042,
      "learning_rate": 5.179455445544554e-07,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8531643152236938,
      "num_tokens": 29019001.0,
      "step": 838
    },
    {
      "epoch": 0.15580315691736304,
      "grad_norm": 1.5581322038838803,
      "learning_rate": 5.185643564356435e-07,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8368749022483826,
      "num_tokens": 29056190.0,
      "step": 839
    },
    {
      "epoch": 0.15598885793871867,
      "grad_norm": 1.9328865131833286,
      "learning_rate": 5.191831683168316e-07,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8340964317321777,
      "num_tokens": 29085118.0,
      "step": 840
    },
    {
      "epoch": 0.1561745589600743,
      "grad_norm": 1.4980239872085355,
      "learning_rate": 5.198019801980198e-07,
      "loss": 0.477,
      "mean_token_accuracy": 0.8460891246795654,
      "num_tokens": 29123860.0,
      "step": 841
    },
    {
      "epoch": 0.1563602599814299,
      "grad_norm": 1.5000036488364474,
      "learning_rate": 5.204207920792078e-07,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8364794850349426,
      "num_tokens": 29163105.0,
      "step": 842
    },
    {
      "epoch": 0.1565459610027855,
      "grad_norm": 1.499380513557054,
      "learning_rate": 5.210396039603961e-07,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8626536130905151,
      "num_tokens": 29194756.0,
      "step": 843
    },
    {
      "epoch": 0.15673166202414113,
      "grad_norm": 1.6778782470301583,
      "learning_rate": 5.216584158415841e-07,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8403278589248657,
      "num_tokens": 29226442.0,
      "step": 844
    },
    {
      "epoch": 0.15691736304549675,
      "grad_norm": 1.6107253339101177,
      "learning_rate": 5.222772277227723e-07,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8459293842315674,
      "num_tokens": 29258512.0,
      "step": 845
    },
    {
      "epoch": 0.15710306406685237,
      "grad_norm": 1.8274937167409249,
      "learning_rate": 5.228960396039604e-07,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8366966843605042,
      "num_tokens": 29288815.0,
      "step": 846
    },
    {
      "epoch": 0.157288765088208,
      "grad_norm": 1.5532000713886558,
      "learning_rate": 5.235148514851485e-07,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8575417995452881,
      "num_tokens": 29323389.0,
      "step": 847
    },
    {
      "epoch": 0.15747446610956362,
      "grad_norm": 1.5214666823348653,
      "learning_rate": 5.241336633663366e-07,
      "loss": 0.5201,
      "mean_token_accuracy": 0.83298659324646,
      "num_tokens": 29364021.0,
      "step": 848
    },
    {
      "epoch": 0.1576601671309192,
      "grad_norm": 1.744838422366647,
      "learning_rate": 5.247524752475247e-07,
      "loss": 0.47,
      "mean_token_accuracy": 0.8441135883331299,
      "num_tokens": 29392397.0,
      "step": 849
    },
    {
      "epoch": 0.15784586815227483,
      "grad_norm": 1.5090485112656449,
      "learning_rate": 5.253712871287128e-07,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8405015468597412,
      "num_tokens": 29429288.0,
      "step": 850
    },
    {
      "epoch": 0.15803156917363045,
      "grad_norm": 1.4878823142234079,
      "learning_rate": 5.25990099009901e-07,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8456007242202759,
      "num_tokens": 29469601.0,
      "step": 851
    },
    {
      "epoch": 0.15821727019498608,
      "grad_norm": 1.520295280319927,
      "learning_rate": 5.26608910891089e-07,
      "loss": 0.5329,
      "mean_token_accuracy": 0.831669807434082,
      "num_tokens": 29509404.0,
      "step": 852
    },
    {
      "epoch": 0.1584029712163417,
      "grad_norm": 1.3737171383307527,
      "learning_rate": 5.272277227722772e-07,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8488181233406067,
      "num_tokens": 29550068.0,
      "step": 853
    },
    {
      "epoch": 0.15858867223769732,
      "grad_norm": 1.4223420249858145,
      "learning_rate": 5.278465346534653e-07,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8531002998352051,
      "num_tokens": 29589233.0,
      "step": 854
    },
    {
      "epoch": 0.15877437325905291,
      "grad_norm": 1.610592562606911,
      "learning_rate": 5.284653465346534e-07,
      "loss": 0.4371,
      "mean_token_accuracy": 0.856243371963501,
      "num_tokens": 29620875.0,
      "step": 855
    },
    {
      "epoch": 0.15896007428040854,
      "grad_norm": 1.4330880510157658,
      "learning_rate": 5.290841584158416e-07,
      "loss": 0.459,
      "mean_token_accuracy": 0.8509941101074219,
      "num_tokens": 29657681.0,
      "step": 856
    },
    {
      "epoch": 0.15914577530176416,
      "grad_norm": 1.5071542434695542,
      "learning_rate": 5.297029702970297e-07,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8480297923088074,
      "num_tokens": 29694129.0,
      "step": 857
    },
    {
      "epoch": 0.15933147632311978,
      "grad_norm": 1.5463502831126015,
      "learning_rate": 5.303217821782178e-07,
      "loss": 0.5017,
      "mean_token_accuracy": 0.843341052532196,
      "num_tokens": 29733051.0,
      "step": 858
    },
    {
      "epoch": 0.1595171773444754,
      "grad_norm": 1.649678695351174,
      "learning_rate": 5.30940594059406e-07,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8359769582748413,
      "num_tokens": 29764080.0,
      "step": 859
    },
    {
      "epoch": 0.15970287836583102,
      "grad_norm": 1.4448331678718382,
      "learning_rate": 5.31559405940594e-07,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8529849052429199,
      "num_tokens": 29803474.0,
      "step": 860
    },
    {
      "epoch": 0.15988857938718662,
      "grad_norm": 1.7874934551929267,
      "learning_rate": 5.321782178217822e-07,
      "loss": 0.507,
      "mean_token_accuracy": 0.8422361016273499,
      "num_tokens": 29832381.0,
      "step": 861
    },
    {
      "epoch": 0.16007428040854224,
      "grad_norm": 1.449030834826118,
      "learning_rate": 5.327970297029702e-07,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8409664034843445,
      "num_tokens": 29875233.0,
      "step": 862
    },
    {
      "epoch": 0.16025998142989786,
      "grad_norm": 1.445399066640975,
      "learning_rate": 5.334158415841584e-07,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8646599054336548,
      "num_tokens": 29912164.0,
      "step": 863
    },
    {
      "epoch": 0.16044568245125349,
      "grad_norm": 1.7350005813416491,
      "learning_rate": 5.340346534653465e-07,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8380080461502075,
      "num_tokens": 29941747.0,
      "step": 864
    },
    {
      "epoch": 0.1606313834726091,
      "grad_norm": 1.7124318894920278,
      "learning_rate": 5.346534653465346e-07,
      "loss": 0.546,
      "mean_token_accuracy": 0.8285888433456421,
      "num_tokens": 29972179.0,
      "step": 865
    },
    {
      "epoch": 0.16081708449396473,
      "grad_norm": 1.547884255121044,
      "learning_rate": 5.352722772277227e-07,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8403831720352173,
      "num_tokens": 30009352.0,
      "step": 866
    },
    {
      "epoch": 0.16100278551532032,
      "grad_norm": 1.7181771551623872,
      "learning_rate": 5.358910891089109e-07,
      "loss": 0.4866,
      "mean_token_accuracy": 0.843227207660675,
      "num_tokens": 30039976.0,
      "step": 867
    },
    {
      "epoch": 0.16118848653667595,
      "grad_norm": 1.5839914649503846,
      "learning_rate": 5.365099009900989e-07,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8502339720726013,
      "num_tokens": 30071731.0,
      "step": 868
    },
    {
      "epoch": 0.16137418755803157,
      "grad_norm": 1.5821014794742945,
      "learning_rate": 5.371287128712872e-07,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8459689617156982,
      "num_tokens": 30104648.0,
      "step": 869
    },
    {
      "epoch": 0.1615598885793872,
      "grad_norm": 1.4644392708625449,
      "learning_rate": 5.377475247524752e-07,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8546414971351624,
      "num_tokens": 30141291.0,
      "step": 870
    },
    {
      "epoch": 0.1617455896007428,
      "grad_norm": 1.5396505688750695,
      "learning_rate": 5.383663366336634e-07,
      "loss": 0.462,
      "mean_token_accuracy": 0.8510095477104187,
      "num_tokens": 30178934.0,
      "step": 871
    },
    {
      "epoch": 0.16193129062209843,
      "grad_norm": 1.7865834252013135,
      "learning_rate": 5.389851485148515e-07,
      "loss": 0.525,
      "mean_token_accuracy": 0.8304857015609741,
      "num_tokens": 30210512.0,
      "step": 872
    },
    {
      "epoch": 0.16211699164345403,
      "grad_norm": 1.7549766390556902,
      "learning_rate": 5.396039603960396e-07,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8413515686988831,
      "num_tokens": 30239442.0,
      "step": 873
    },
    {
      "epoch": 0.16230269266480965,
      "grad_norm": 1.3987501417820165,
      "learning_rate": 5.402227722772277e-07,
      "loss": 0.501,
      "mean_token_accuracy": 0.8364378213882446,
      "num_tokens": 30283774.0,
      "step": 874
    },
    {
      "epoch": 0.16248839368616527,
      "grad_norm": 1.7145817320637449,
      "learning_rate": 5.408415841584159e-07,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8478389978408813,
      "num_tokens": 30312674.0,
      "step": 875
    },
    {
      "epoch": 0.1626740947075209,
      "grad_norm": 1.670543024718537,
      "learning_rate": 5.414603960396039e-07,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8449987173080444,
      "num_tokens": 30342092.0,
      "step": 876
    },
    {
      "epoch": 0.16285979572887652,
      "grad_norm": 1.4677551073261523,
      "learning_rate": 5.420792079207921e-07,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8416881561279297,
      "num_tokens": 30383826.0,
      "step": 877
    },
    {
      "epoch": 0.16304549675023214,
      "grad_norm": 1.6414411720250714,
      "learning_rate": 5.426980198019801e-07,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8461030721664429,
      "num_tokens": 30414427.0,
      "step": 878
    },
    {
      "epoch": 0.16323119777158773,
      "grad_norm": 1.6279220052138663,
      "learning_rate": 5.433168316831683e-07,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8563233613967896,
      "num_tokens": 30450553.0,
      "step": 879
    },
    {
      "epoch": 0.16341689879294335,
      "grad_norm": 1.774202541116528,
      "learning_rate": 5.439356435643564e-07,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8524595499038696,
      "num_tokens": 30482617.0,
      "step": 880
    },
    {
      "epoch": 0.16360259981429898,
      "grad_norm": 1.5859211556687327,
      "learning_rate": 5.445544554455445e-07,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8502195477485657,
      "num_tokens": 30516896.0,
      "step": 881
    },
    {
      "epoch": 0.1637883008356546,
      "grad_norm": 1.474127980493389,
      "learning_rate": 5.451732673267327e-07,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8430948257446289,
      "num_tokens": 30555500.0,
      "step": 882
    },
    {
      "epoch": 0.16397400185701022,
      "grad_norm": 1.6665934512148604,
      "learning_rate": 5.457920792079208e-07,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8347160816192627,
      "num_tokens": 30586798.0,
      "step": 883
    },
    {
      "epoch": 0.16415970287836584,
      "grad_norm": 1.8237579357027973,
      "learning_rate": 5.464108910891089e-07,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8487429618835449,
      "num_tokens": 30614655.0,
      "step": 884
    },
    {
      "epoch": 0.16434540389972144,
      "grad_norm": 1.8420557380793032,
      "learning_rate": 5.470297029702971e-07,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8348324298858643,
      "num_tokens": 30646695.0,
      "step": 885
    },
    {
      "epoch": 0.16453110492107706,
      "grad_norm": 1.7595666845902138,
      "learning_rate": 5.476485148514851e-07,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8347171545028687,
      "num_tokens": 30681342.0,
      "step": 886
    },
    {
      "epoch": 0.16471680594243268,
      "grad_norm": 1.5578368072031605,
      "learning_rate": 5.482673267326733e-07,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8601741790771484,
      "num_tokens": 30718048.0,
      "step": 887
    },
    {
      "epoch": 0.1649025069637883,
      "grad_norm": 1.5025571251052916,
      "learning_rate": 5.488861386138614e-07,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8404875993728638,
      "num_tokens": 30758027.0,
      "step": 888
    },
    {
      "epoch": 0.16508820798514393,
      "grad_norm": 1.5683824966336966,
      "learning_rate": 5.495049504950495e-07,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8544433116912842,
      "num_tokens": 30793882.0,
      "step": 889
    },
    {
      "epoch": 0.16527390900649955,
      "grad_norm": 1.6824610463313037,
      "learning_rate": 5.501237623762376e-07,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8488715291023254,
      "num_tokens": 30824420.0,
      "step": 890
    },
    {
      "epoch": 0.16545961002785514,
      "grad_norm": 1.6544593365486553,
      "learning_rate": 5.507425742574257e-07,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8641357421875,
      "num_tokens": 30856702.0,
      "step": 891
    },
    {
      "epoch": 0.16564531104921076,
      "grad_norm": 1.5473081806763993,
      "learning_rate": 5.513613861386138e-07,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8461555242538452,
      "num_tokens": 30895418.0,
      "step": 892
    },
    {
      "epoch": 0.16583101207056639,
      "grad_norm": 1.5108483089859055,
      "learning_rate": 5.51980198019802e-07,
      "loss": 0.426,
      "mean_token_accuracy": 0.859079897403717,
      "num_tokens": 30928624.0,
      "step": 893
    },
    {
      "epoch": 0.166016713091922,
      "grad_norm": 1.5457202993096184,
      "learning_rate": 5.5259900990099e-07,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8511345386505127,
      "num_tokens": 30963563.0,
      "step": 894
    },
    {
      "epoch": 0.16620241411327763,
      "grad_norm": 1.533047779963994,
      "learning_rate": 5.532178217821783e-07,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8561648726463318,
      "num_tokens": 30997987.0,
      "step": 895
    },
    {
      "epoch": 0.16638811513463325,
      "grad_norm": 1.542198935207323,
      "learning_rate": 5.538366336633663e-07,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8363523483276367,
      "num_tokens": 31037153.0,
      "step": 896
    },
    {
      "epoch": 0.16657381615598885,
      "grad_norm": 1.482255881041508,
      "learning_rate": 5.544554455445545e-07,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8537400960922241,
      "num_tokens": 31076822.0,
      "step": 897
    },
    {
      "epoch": 0.16675951717734447,
      "grad_norm": 1.5533896574420323,
      "learning_rate": 5.550742574257426e-07,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8418066501617432,
      "num_tokens": 31114673.0,
      "step": 898
    },
    {
      "epoch": 0.1669452181987001,
      "grad_norm": 1.4560367131749894,
      "learning_rate": 5.556930693069307e-07,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8507963418960571,
      "num_tokens": 31156452.0,
      "step": 899
    },
    {
      "epoch": 0.1671309192200557,
      "grad_norm": 1.5900672160143918,
      "learning_rate": 5.563118811881188e-07,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8371331691741943,
      "num_tokens": 31191672.0,
      "step": 900
    },
    {
      "epoch": 0.16731662024141133,
      "grad_norm": 1.5145864290656088,
      "learning_rate": 5.56930693069307e-07,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8451017141342163,
      "num_tokens": 31228717.0,
      "step": 901
    },
    {
      "epoch": 0.16750232126276696,
      "grad_norm": 1.783112703087107,
      "learning_rate": 5.57549504950495e-07,
      "loss": 0.4638,
      "mean_token_accuracy": 0.846142590045929,
      "num_tokens": 31259288.0,
      "step": 902
    },
    {
      "epoch": 0.16768802228412255,
      "grad_norm": 1.5087809735419282,
      "learning_rate": 5.581683168316832e-07,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8408178091049194,
      "num_tokens": 31298732.0,
      "step": 903
    },
    {
      "epoch": 0.16787372330547817,
      "grad_norm": 1.4401257615641565,
      "learning_rate": 5.587871287128712e-07,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8469799757003784,
      "num_tokens": 31342729.0,
      "step": 904
    },
    {
      "epoch": 0.1680594243268338,
      "grad_norm": 1.5634370781916023,
      "learning_rate": 5.594059405940594e-07,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8526891469955444,
      "num_tokens": 31376523.0,
      "step": 905
    },
    {
      "epoch": 0.16824512534818942,
      "grad_norm": 1.6405611276152756,
      "learning_rate": 5.600247524752475e-07,
      "loss": 0.436,
      "mean_token_accuracy": 0.8603296279907227,
      "num_tokens": 31403152.0,
      "step": 906
    },
    {
      "epoch": 0.16843082636954504,
      "grad_norm": 1.591863418606977,
      "learning_rate": 5.606435643564356e-07,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8442001938819885,
      "num_tokens": 31435844.0,
      "step": 907
    },
    {
      "epoch": 0.16861652739090066,
      "grad_norm": 1.507081324053829,
      "learning_rate": 5.612623762376237e-07,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8531844615936279,
      "num_tokens": 31471993.0,
      "step": 908
    },
    {
      "epoch": 0.16880222841225626,
      "grad_norm": 1.5353035583149954,
      "learning_rate": 5.61881188118812e-07,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8599253296852112,
      "num_tokens": 31509076.0,
      "step": 909
    },
    {
      "epoch": 0.16898792943361188,
      "grad_norm": 1.5185656251160828,
      "learning_rate": 5.625e-07,
      "loss": 0.473,
      "mean_token_accuracy": 0.850314199924469,
      "num_tokens": 31543942.0,
      "step": 910
    },
    {
      "epoch": 0.1691736304549675,
      "grad_norm": 1.6129188774109933,
      "learning_rate": 5.631188118811881e-07,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8398013114929199,
      "num_tokens": 31577826.0,
      "step": 911
    },
    {
      "epoch": 0.16935933147632312,
      "grad_norm": 1.7165900164146743,
      "learning_rate": 5.637376237623762e-07,
      "loss": 0.4864,
      "mean_token_accuracy": 0.842954158782959,
      "num_tokens": 31607488.0,
      "step": 912
    },
    {
      "epoch": 0.16954503249767874,
      "grad_norm": 1.612109417581514,
      "learning_rate": 5.643564356435643e-07,
      "loss": 0.458,
      "mean_token_accuracy": 0.8490064144134521,
      "num_tokens": 31641625.0,
      "step": 913
    },
    {
      "epoch": 0.16973073351903437,
      "grad_norm": 1.7265306329693622,
      "learning_rate": 5.649752475247525e-07,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8496452569961548,
      "num_tokens": 31670145.0,
      "step": 914
    },
    {
      "epoch": 0.16991643454038996,
      "grad_norm": 1.6702704692221004,
      "learning_rate": 5.655940594059405e-07,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8445988893508911,
      "num_tokens": 31703692.0,
      "step": 915
    },
    {
      "epoch": 0.17010213556174558,
      "grad_norm": 1.5005081284555057,
      "learning_rate": 5.662128712871287e-07,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8362478017807007,
      "num_tokens": 31740746.0,
      "step": 916
    },
    {
      "epoch": 0.1702878365831012,
      "grad_norm": 1.5809829565285207,
      "learning_rate": 5.668316831683167e-07,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8533685207366943,
      "num_tokens": 31773704.0,
      "step": 917
    },
    {
      "epoch": 0.17047353760445683,
      "grad_norm": 1.6718074954303288,
      "learning_rate": 5.674504950495049e-07,
      "loss": 0.4648,
      "mean_token_accuracy": 0.85545814037323,
      "num_tokens": 31805605.0,
      "step": 918
    },
    {
      "epoch": 0.17065923862581245,
      "grad_norm": 1.6135635402028305,
      "learning_rate": 5.68069306930693e-07,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8286429643630981,
      "num_tokens": 31846152.0,
      "step": 919
    },
    {
      "epoch": 0.17084493964716807,
      "grad_norm": 1.678477035178483,
      "learning_rate": 5.686881188118811e-07,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8385404348373413,
      "num_tokens": 31881370.0,
      "step": 920
    },
    {
      "epoch": 0.17103064066852366,
      "grad_norm": 1.6019486678396992,
      "learning_rate": 5.693069306930692e-07,
      "loss": 0.4693,
      "mean_token_accuracy": 0.850965142250061,
      "num_tokens": 31914314.0,
      "step": 921
    },
    {
      "epoch": 0.1712163416898793,
      "grad_norm": 1.6073668077271654,
      "learning_rate": 5.699257425742575e-07,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8432651162147522,
      "num_tokens": 31949431.0,
      "step": 922
    },
    {
      "epoch": 0.1714020427112349,
      "grad_norm": 1.4404869379548213,
      "learning_rate": 5.705445544554455e-07,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8537249565124512,
      "num_tokens": 31989441.0,
      "step": 923
    },
    {
      "epoch": 0.17158774373259053,
      "grad_norm": 1.5265978937144367,
      "learning_rate": 5.711633663366337e-07,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8550459146499634,
      "num_tokens": 32022567.0,
      "step": 924
    },
    {
      "epoch": 0.17177344475394615,
      "grad_norm": 1.412482154116146,
      "learning_rate": 5.717821782178217e-07,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8356281518936157,
      "num_tokens": 32063975.0,
      "step": 925
    },
    {
      "epoch": 0.17195914577530177,
      "grad_norm": 1.66316078355755,
      "learning_rate": 5.724009900990099e-07,
      "loss": 0.464,
      "mean_token_accuracy": 0.8524562120437622,
      "num_tokens": 32095679.0,
      "step": 926
    },
    {
      "epoch": 0.17214484679665737,
      "grad_norm": 1.5662528792725339,
      "learning_rate": 5.73019801980198e-07,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8561694622039795,
      "num_tokens": 32126845.0,
      "step": 927
    },
    {
      "epoch": 0.172330547818013,
      "grad_norm": 1.4768330942319643,
      "learning_rate": 5.736386138613861e-07,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8568360805511475,
      "num_tokens": 32166134.0,
      "step": 928
    },
    {
      "epoch": 0.1725162488393686,
      "grad_norm": 1.4689961767652797,
      "learning_rate": 5.742574257425742e-07,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8528995513916016,
      "num_tokens": 32202317.0,
      "step": 929
    },
    {
      "epoch": 0.17270194986072424,
      "grad_norm": 1.5562448929325507,
      "learning_rate": 5.748762376237623e-07,
      "loss": 0.471,
      "mean_token_accuracy": 0.845079779624939,
      "num_tokens": 32237759.0,
      "step": 930
    },
    {
      "epoch": 0.17288765088207986,
      "grad_norm": 1.6645264489865805,
      "learning_rate": 5.754950495049504e-07,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8421589136123657,
      "num_tokens": 32268139.0,
      "step": 931
    },
    {
      "epoch": 0.17307335190343548,
      "grad_norm": 1.6820915307998625,
      "learning_rate": 5.761138613861386e-07,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8483422994613647,
      "num_tokens": 32296251.0,
      "step": 932
    },
    {
      "epoch": 0.17325905292479107,
      "grad_norm": 1.6551663795985687,
      "learning_rate": 5.767326732673266e-07,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8541156053543091,
      "num_tokens": 32325489.0,
      "step": 933
    },
    {
      "epoch": 0.1734447539461467,
      "grad_norm": 1.5059505154557642,
      "learning_rate": 5.773514851485148e-07,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8397997617721558,
      "num_tokens": 32365288.0,
      "step": 934
    },
    {
      "epoch": 0.17363045496750232,
      "grad_norm": 1.5478426043161828,
      "learning_rate": 5.779702970297029e-07,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8468549251556396,
      "num_tokens": 32401410.0,
      "step": 935
    },
    {
      "epoch": 0.17381615598885794,
      "grad_norm": 1.672830115467317,
      "learning_rate": 5.785891089108911e-07,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8364162445068359,
      "num_tokens": 32435122.0,
      "step": 936
    },
    {
      "epoch": 0.17400185701021356,
      "grad_norm": 1.5492795014522127,
      "learning_rate": 5.792079207920792e-07,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8529731035232544,
      "num_tokens": 32470111.0,
      "step": 937
    },
    {
      "epoch": 0.17418755803156918,
      "grad_norm": 1.6355952887847507,
      "learning_rate": 5.798267326732673e-07,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8553017377853394,
      "num_tokens": 32501450.0,
      "step": 938
    },
    {
      "epoch": 0.17437325905292478,
      "grad_norm": 1.5016592757854217,
      "learning_rate": 5.804455445544554e-07,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8561818599700928,
      "num_tokens": 32536408.0,
      "step": 939
    },
    {
      "epoch": 0.1745589600742804,
      "grad_norm": 1.6710656399299446,
      "learning_rate": 5.810643564356436e-07,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8432020545005798,
      "num_tokens": 32566016.0,
      "step": 940
    },
    {
      "epoch": 0.17474466109563602,
      "grad_norm": 1.4442024422065143,
      "learning_rate": 5.816831683168316e-07,
      "loss": 0.4421,
      "mean_token_accuracy": 0.855876088142395,
      "num_tokens": 32606446.0,
      "step": 941
    },
    {
      "epoch": 0.17493036211699164,
      "grad_norm": 1.4426503499708634,
      "learning_rate": 5.823019801980198e-07,
      "loss": 0.456,
      "mean_token_accuracy": 0.8500708937644958,
      "num_tokens": 32644184.0,
      "step": 942
    },
    {
      "epoch": 0.17511606313834727,
      "grad_norm": 1.7234247932881537,
      "learning_rate": 5.829207920792078e-07,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8411916494369507,
      "num_tokens": 32675055.0,
      "step": 943
    },
    {
      "epoch": 0.1753017641597029,
      "grad_norm": 1.6315077059362812,
      "learning_rate": 5.83539603960396e-07,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8151626586914062,
      "num_tokens": 32708770.0,
      "step": 944
    },
    {
      "epoch": 0.17548746518105848,
      "grad_norm": 1.6722566731898518,
      "learning_rate": 5.841584158415841e-07,
      "loss": 0.4483,
      "mean_token_accuracy": 0.858375608921051,
      "num_tokens": 32742738.0,
      "step": 945
    },
    {
      "epoch": 0.1756731662024141,
      "grad_norm": 1.4427000041334324,
      "learning_rate": 5.847772277227722e-07,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8534311056137085,
      "num_tokens": 32780421.0,
      "step": 946
    },
    {
      "epoch": 0.17585886722376973,
      "grad_norm": 1.4950136719598413,
      "learning_rate": 5.853960396039603e-07,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8552907109260559,
      "num_tokens": 32818797.0,
      "step": 947
    },
    {
      "epoch": 0.17604456824512535,
      "grad_norm": 1.5743131645376662,
      "learning_rate": 5.860148514851486e-07,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8258138298988342,
      "num_tokens": 32858671.0,
      "step": 948
    },
    {
      "epoch": 0.17623026926648097,
      "grad_norm": 1.5362566629664216,
      "learning_rate": 5.866336633663366e-07,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8599402904510498,
      "num_tokens": 32894247.0,
      "step": 949
    },
    {
      "epoch": 0.1764159702878366,
      "grad_norm": 1.4497991219984434,
      "learning_rate": 5.872524752475248e-07,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8569241762161255,
      "num_tokens": 32936185.0,
      "step": 950
    },
    {
      "epoch": 0.1766016713091922,
      "grad_norm": 1.4337823784890587,
      "learning_rate": 5.878712871287128e-07,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8533390760421753,
      "num_tokens": 32975583.0,
      "step": 951
    },
    {
      "epoch": 0.1767873723305478,
      "grad_norm": 1.6323261964650644,
      "learning_rate": 5.88490099009901e-07,
      "loss": 0.467,
      "mean_token_accuracy": 0.849439263343811,
      "num_tokens": 33008567.0,
      "step": 952
    },
    {
      "epoch": 0.17697307335190343,
      "grad_norm": 1.5989747868404287,
      "learning_rate": 5.891089108910891e-07,
      "loss": 0.4974,
      "mean_token_accuracy": 0.840134859085083,
      "num_tokens": 33043347.0,
      "step": 953
    },
    {
      "epoch": 0.17715877437325905,
      "grad_norm": 1.8392707379539026,
      "learning_rate": 5.897277227722772e-07,
      "loss": 0.467,
      "mean_token_accuracy": 0.8512839674949646,
      "num_tokens": 33071092.0,
      "step": 954
    },
    {
      "epoch": 0.17734447539461468,
      "grad_norm": 1.6360756830321042,
      "learning_rate": 5.903465346534653e-07,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8534590005874634,
      "num_tokens": 33103699.0,
      "step": 955
    },
    {
      "epoch": 0.1775301764159703,
      "grad_norm": 1.4989774365643656,
      "learning_rate": 5.909653465346535e-07,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8601162433624268,
      "num_tokens": 33142163.0,
      "step": 956
    },
    {
      "epoch": 0.1777158774373259,
      "grad_norm": 1.4252956518154993,
      "learning_rate": 5.915841584158415e-07,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8480538129806519,
      "num_tokens": 33180495.0,
      "step": 957
    },
    {
      "epoch": 0.17790157845868151,
      "grad_norm": 1.5956001678354546,
      "learning_rate": 5.922029702970297e-07,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8386791944503784,
      "num_tokens": 33216046.0,
      "step": 958
    },
    {
      "epoch": 0.17808727948003714,
      "grad_norm": 1.4565498026688957,
      "learning_rate": 5.928217821782177e-07,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8436566591262817,
      "num_tokens": 33254114.0,
      "step": 959
    },
    {
      "epoch": 0.17827298050139276,
      "grad_norm": 1.6032988380461826,
      "learning_rate": 5.934405940594059e-07,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8640875816345215,
      "num_tokens": 33287420.0,
      "step": 960
    },
    {
      "epoch": 0.17845868152274838,
      "grad_norm": 1.4431914580279361,
      "learning_rate": 5.94059405940594e-07,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8382806777954102,
      "num_tokens": 33328962.0,
      "step": 961
    },
    {
      "epoch": 0.178644382544104,
      "grad_norm": 1.3873806274669587,
      "learning_rate": 5.946782178217822e-07,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8572474718093872,
      "num_tokens": 33368484.0,
      "step": 962
    },
    {
      "epoch": 0.1788300835654596,
      "grad_norm": 1.5796118304759068,
      "learning_rate": 5.952970297029703e-07,
      "loss": 0.443,
      "mean_token_accuracy": 0.8554178476333618,
      "num_tokens": 33398433.0,
      "step": 963
    },
    {
      "epoch": 0.17901578458681522,
      "grad_norm": 1.6076616031321622,
      "learning_rate": 5.959158415841584e-07,
      "loss": 0.448,
      "mean_token_accuracy": 0.852891206741333,
      "num_tokens": 33427403.0,
      "step": 964
    },
    {
      "epoch": 0.17920148560817084,
      "grad_norm": 1.774641655349903,
      "learning_rate": 5.965346534653465e-07,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8518486022949219,
      "num_tokens": 33454617.0,
      "step": 965
    },
    {
      "epoch": 0.17938718662952646,
      "grad_norm": 1.7488975123083936,
      "learning_rate": 5.971534653465347e-07,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8491542935371399,
      "num_tokens": 33483155.0,
      "step": 966
    },
    {
      "epoch": 0.17957288765088208,
      "grad_norm": 1.5497021075508854,
      "learning_rate": 5.977722772277227e-07,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8467762470245361,
      "num_tokens": 33516288.0,
      "step": 967
    },
    {
      "epoch": 0.1797585886722377,
      "grad_norm": 1.6735302534621286,
      "learning_rate": 5.983910891089109e-07,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8669576644897461,
      "num_tokens": 33547715.0,
      "step": 968
    },
    {
      "epoch": 0.1799442896935933,
      "grad_norm": 1.5844934185300972,
      "learning_rate": 5.99009900990099e-07,
      "loss": 0.4745,
      "mean_token_accuracy": 0.846351683139801,
      "num_tokens": 33578529.0,
      "step": 969
    },
    {
      "epoch": 0.18012999071494892,
      "grad_norm": 1.597144221364036,
      "learning_rate": 5.996287128712871e-07,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8496443033218384,
      "num_tokens": 33611628.0,
      "step": 970
    },
    {
      "epoch": 0.18031569173630455,
      "grad_norm": 1.5414120923645969,
      "learning_rate": 6.002475247524752e-07,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8597387075424194,
      "num_tokens": 33645790.0,
      "step": 971
    },
    {
      "epoch": 0.18050139275766017,
      "grad_norm": 1.5376602644502604,
      "learning_rate": 6.008663366336633e-07,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8518141508102417,
      "num_tokens": 33678992.0,
      "step": 972
    },
    {
      "epoch": 0.1806870937790158,
      "grad_norm": 1.596805439744286,
      "learning_rate": 6.014851485148514e-07,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8560202717781067,
      "num_tokens": 33711081.0,
      "step": 973
    },
    {
      "epoch": 0.1808727948003714,
      "grad_norm": 1.7805402827423105,
      "learning_rate": 6.021039603960396e-07,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8517508506774902,
      "num_tokens": 33736778.0,
      "step": 974
    },
    {
      "epoch": 0.181058495821727,
      "grad_norm": 1.5185408703312988,
      "learning_rate": 6.027227722772277e-07,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8625602126121521,
      "num_tokens": 33770484.0,
      "step": 975
    },
    {
      "epoch": 0.18124419684308263,
      "grad_norm": 1.4370037410840453,
      "learning_rate": 6.033415841584159e-07,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8507580757141113,
      "num_tokens": 33809698.0,
      "step": 976
    },
    {
      "epoch": 0.18142989786443825,
      "grad_norm": 1.5762141014701005,
      "learning_rate": 6.03960396039604e-07,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8421971797943115,
      "num_tokens": 33846805.0,
      "step": 977
    },
    {
      "epoch": 0.18161559888579387,
      "grad_norm": 1.5619687728471963,
      "learning_rate": 6.045792079207921e-07,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8592203855514526,
      "num_tokens": 33882306.0,
      "step": 978
    },
    {
      "epoch": 0.1818012999071495,
      "grad_norm": 1.4945807244909346,
      "learning_rate": 6.051980198019802e-07,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8436917662620544,
      "num_tokens": 33917122.0,
      "step": 979
    },
    {
      "epoch": 0.18198700092850512,
      "grad_norm": 1.4757521855907478,
      "learning_rate": 6.058168316831683e-07,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8533287644386292,
      "num_tokens": 33955637.0,
      "step": 980
    },
    {
      "epoch": 0.18217270194986074,
      "grad_norm": 1.6293520021933354,
      "learning_rate": 6.064356435643564e-07,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8431211113929749,
      "num_tokens": 33989654.0,
      "step": 981
    },
    {
      "epoch": 0.18235840297121633,
      "grad_norm": 1.5365345761573073,
      "learning_rate": 6.070544554455446e-07,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8520899415016174,
      "num_tokens": 34024527.0,
      "step": 982
    },
    {
      "epoch": 0.18254410399257195,
      "grad_norm": 1.6088354125479027,
      "learning_rate": 6.076732673267326e-07,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8544992208480835,
      "num_tokens": 34053795.0,
      "step": 983
    },
    {
      "epoch": 0.18272980501392758,
      "grad_norm": 1.5388407204224117,
      "learning_rate": 6.082920792079208e-07,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8367465734481812,
      "num_tokens": 34091002.0,
      "step": 984
    },
    {
      "epoch": 0.1829155060352832,
      "grad_norm": 1.5594235479679943,
      "learning_rate": 6.089108910891088e-07,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8384783864021301,
      "num_tokens": 34125554.0,
      "step": 985
    },
    {
      "epoch": 0.18310120705663882,
      "grad_norm": 1.5839627772112665,
      "learning_rate": 6.09529702970297e-07,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8328864574432373,
      "num_tokens": 34162050.0,
      "step": 986
    },
    {
      "epoch": 0.18328690807799444,
      "grad_norm": 1.6714196744745757,
      "learning_rate": 6.101485148514851e-07,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8531026840209961,
      "num_tokens": 34192621.0,
      "step": 987
    },
    {
      "epoch": 0.18347260909935004,
      "grad_norm": 1.644195907097496,
      "learning_rate": 6.107673267326733e-07,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8383974432945251,
      "num_tokens": 34230209.0,
      "step": 988
    },
    {
      "epoch": 0.18365831012070566,
      "grad_norm": 1.3789626378975302,
      "learning_rate": 6.113861386138614e-07,
      "loss": 0.448,
      "mean_token_accuracy": 0.8528175354003906,
      "num_tokens": 34274178.0,
      "step": 989
    },
    {
      "epoch": 0.18384401114206128,
      "grad_norm": 1.7487959193271,
      "learning_rate": 6.120049504950496e-07,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8518197536468506,
      "num_tokens": 34302714.0,
      "step": 990
    },
    {
      "epoch": 0.1840297121634169,
      "grad_norm": 1.5393166340907811,
      "learning_rate": 6.126237623762376e-07,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8250444531440735,
      "num_tokens": 34340516.0,
      "step": 991
    },
    {
      "epoch": 0.18421541318477253,
      "grad_norm": 1.5392212769907143,
      "learning_rate": 6.132425742574258e-07,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8387334942817688,
      "num_tokens": 34375996.0,
      "step": 992
    },
    {
      "epoch": 0.18440111420612815,
      "grad_norm": 1.4701177734109412,
      "learning_rate": 6.138613861386138e-07,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8654770255088806,
      "num_tokens": 34412104.0,
      "step": 993
    },
    {
      "epoch": 0.18458681522748374,
      "grad_norm": 1.4199145705906833,
      "learning_rate": 6.14480198019802e-07,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8481123447418213,
      "num_tokens": 34457699.0,
      "step": 994
    },
    {
      "epoch": 0.18477251624883936,
      "grad_norm": 1.5207911500680071,
      "learning_rate": 6.150990099009901e-07,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8683996200561523,
      "num_tokens": 34489430.0,
      "step": 995
    },
    {
      "epoch": 0.18495821727019499,
      "grad_norm": 1.5862085210708627,
      "learning_rate": 6.157178217821782e-07,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8532916307449341,
      "num_tokens": 34525204.0,
      "step": 996
    },
    {
      "epoch": 0.1851439182915506,
      "grad_norm": 1.616309397074694,
      "learning_rate": 6.163366336633663e-07,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8305627107620239,
      "num_tokens": 34561085.0,
      "step": 997
    },
    {
      "epoch": 0.18532961931290623,
      "grad_norm": 1.6872772779608183,
      "learning_rate": 6.169554455445544e-07,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8462559580802917,
      "num_tokens": 34592131.0,
      "step": 998
    },
    {
      "epoch": 0.18551532033426185,
      "grad_norm": 1.4342150509110172,
      "learning_rate": 6.175742574257425e-07,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8583451509475708,
      "num_tokens": 34629096.0,
      "step": 999
    },
    {
      "epoch": 0.18570102135561745,
      "grad_norm": 1.3683452437922219,
      "learning_rate": 6.181930693069307e-07,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8434200286865234,
      "num_tokens": 34672565.0,
      "step": 1000
    },
    {
      "epoch": 0.18588672237697307,
      "grad_norm": 1.5298190343140985,
      "learning_rate": 6.188118811881187e-07,
      "loss": 0.493,
      "mean_token_accuracy": 0.8405337333679199,
      "num_tokens": 34710011.0,
      "step": 1001
    },
    {
      "epoch": 0.1860724233983287,
      "grad_norm": 1.599060040528443,
      "learning_rate": 6.19430693069307e-07,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8500639200210571,
      "num_tokens": 34749152.0,
      "step": 1002
    },
    {
      "epoch": 0.1862581244196843,
      "grad_norm": 1.540939824113786,
      "learning_rate": 6.200495049504951e-07,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8410995006561279,
      "num_tokens": 34784958.0,
      "step": 1003
    },
    {
      "epoch": 0.18644382544103993,
      "grad_norm": 1.4492849103823877,
      "learning_rate": 6.206683168316832e-07,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8587119579315186,
      "num_tokens": 34820459.0,
      "step": 1004
    },
    {
      "epoch": 0.18662952646239556,
      "grad_norm": 1.4997544325083125,
      "learning_rate": 6.212871287128713e-07,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8687474727630615,
      "num_tokens": 34851716.0,
      "step": 1005
    },
    {
      "epoch": 0.18681522748375115,
      "grad_norm": 1.5788500854221819,
      "learning_rate": 6.219059405940594e-07,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8436951637268066,
      "num_tokens": 34885426.0,
      "step": 1006
    },
    {
      "epoch": 0.18700092850510677,
      "grad_norm": 1.5843832581947108,
      "learning_rate": 6.225247524752475e-07,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8539239168167114,
      "num_tokens": 34921694.0,
      "step": 1007
    },
    {
      "epoch": 0.1871866295264624,
      "grad_norm": 1.5739331076880465,
      "learning_rate": 6.231435643564357e-07,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8662839531898499,
      "num_tokens": 34950933.0,
      "step": 1008
    },
    {
      "epoch": 0.18737233054781802,
      "grad_norm": 1.6347911359778935,
      "learning_rate": 6.237623762376237e-07,
      "loss": 0.479,
      "mean_token_accuracy": 0.8436134457588196,
      "num_tokens": 34983838.0,
      "step": 1009
    },
    {
      "epoch": 0.18755803156917364,
      "grad_norm": 1.6579587818000148,
      "learning_rate": 6.243811881188119e-07,
      "loss": 0.567,
      "mean_token_accuracy": 0.8181260824203491,
      "num_tokens": 35017159.0,
      "step": 1010
    },
    {
      "epoch": 0.18774373259052926,
      "grad_norm": 1.6287728690706644,
      "learning_rate": 6.249999999999999e-07,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8450452089309692,
      "num_tokens": 35047996.0,
      "step": 1011
    },
    {
      "epoch": 0.18792943361188486,
      "grad_norm": 1.6373511679425938,
      "learning_rate": 6.25618811881188e-07,
      "loss": 0.4735,
      "mean_token_accuracy": 0.849249005317688,
      "num_tokens": 35079907.0,
      "step": 1012
    },
    {
      "epoch": 0.18811513463324048,
      "grad_norm": 1.4475576342615666,
      "learning_rate": 6.262376237623762e-07,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8595477342605591,
      "num_tokens": 35118592.0,
      "step": 1013
    },
    {
      "epoch": 0.1883008356545961,
      "grad_norm": 1.6043187398329417,
      "learning_rate": 6.268564356435642e-07,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8498246073722839,
      "num_tokens": 35148986.0,
      "step": 1014
    },
    {
      "epoch": 0.18848653667595172,
      "grad_norm": 1.5401329209999992,
      "learning_rate": 6.274752475247525e-07,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8638293743133545,
      "num_tokens": 35182210.0,
      "step": 1015
    },
    {
      "epoch": 0.18867223769730734,
      "grad_norm": 1.6998435603870987,
      "learning_rate": 6.280940594059406e-07,
      "loss": 0.4835,
      "mean_token_accuracy": 0.847346305847168,
      "num_tokens": 35212835.0,
      "step": 1016
    },
    {
      "epoch": 0.18885793871866297,
      "grad_norm": 1.5997455326913315,
      "learning_rate": 6.287128712871287e-07,
      "loss": 0.481,
      "mean_token_accuracy": 0.8419272899627686,
      "num_tokens": 35245137.0,
      "step": 1017
    },
    {
      "epoch": 0.18904363974001856,
      "grad_norm": 1.4880753895424552,
      "learning_rate": 6.293316831683168e-07,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8595248460769653,
      "num_tokens": 35281263.0,
      "step": 1018
    },
    {
      "epoch": 0.18922934076137418,
      "grad_norm": 1.6748554410045804,
      "learning_rate": 6.299504950495049e-07,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8385109901428223,
      "num_tokens": 35315088.0,
      "step": 1019
    },
    {
      "epoch": 0.1894150417827298,
      "grad_norm": 1.6184772795736388,
      "learning_rate": 6.30569306930693e-07,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8387267589569092,
      "num_tokens": 35351456.0,
      "step": 1020
    },
    {
      "epoch": 0.18960074280408543,
      "grad_norm": 1.5102150899181936,
      "learning_rate": 6.311881188118812e-07,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8458997011184692,
      "num_tokens": 35388924.0,
      "step": 1021
    },
    {
      "epoch": 0.18978644382544105,
      "grad_norm": 1.5296793930961532,
      "learning_rate": 6.318069306930692e-07,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8500503301620483,
      "num_tokens": 35425740.0,
      "step": 1022
    },
    {
      "epoch": 0.18997214484679667,
      "grad_norm": 1.8114474527114108,
      "learning_rate": 6.324257425742574e-07,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8418915271759033,
      "num_tokens": 35453361.0,
      "step": 1023
    },
    {
      "epoch": 0.19015784586815226,
      "grad_norm": 1.5535450396589565,
      "learning_rate": 6.330445544554454e-07,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8442234992980957,
      "num_tokens": 35488958.0,
      "step": 1024
    },
    {
      "epoch": 0.1903435468895079,
      "grad_norm": 1.6428100632644733,
      "learning_rate": 6.336633663366336e-07,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8347334861755371,
      "num_tokens": 35519328.0,
      "step": 1025
    },
    {
      "epoch": 0.1905292479108635,
      "grad_norm": 1.548817679936908,
      "learning_rate": 6.342821782178217e-07,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8571749925613403,
      "num_tokens": 35552985.0,
      "step": 1026
    },
    {
      "epoch": 0.19071494893221913,
      "grad_norm": 1.5762845705429323,
      "learning_rate": 6.349009900990098e-07,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8628931045532227,
      "num_tokens": 35582911.0,
      "step": 1027
    },
    {
      "epoch": 0.19090064995357475,
      "grad_norm": 1.564618543671182,
      "learning_rate": 6.35519801980198e-07,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8462890386581421,
      "num_tokens": 35615550.0,
      "step": 1028
    },
    {
      "epoch": 0.19108635097493037,
      "grad_norm": 1.7882829934422677,
      "learning_rate": 6.361386138613862e-07,
      "loss": 0.4436,
      "mean_token_accuracy": 0.85396409034729,
      "num_tokens": 35643145.0,
      "step": 1029
    },
    {
      "epoch": 0.19127205199628597,
      "grad_norm": 1.6419485441175004,
      "learning_rate": 6.367574257425742e-07,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8505288362503052,
      "num_tokens": 35671954.0,
      "step": 1030
    },
    {
      "epoch": 0.1914577530176416,
      "grad_norm": 1.5047955828072919,
      "learning_rate": 6.373762376237624e-07,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8535885214805603,
      "num_tokens": 35707698.0,
      "step": 1031
    },
    {
      "epoch": 0.1916434540389972,
      "grad_norm": 1.453682865552022,
      "learning_rate": 6.379950495049504e-07,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8757544755935669,
      "num_tokens": 35742023.0,
      "step": 1032
    },
    {
      "epoch": 0.19182915506035284,
      "grad_norm": 1.3788482688520307,
      "learning_rate": 6.386138613861386e-07,
      "loss": 0.454,
      "mean_token_accuracy": 0.8552942276000977,
      "num_tokens": 35784850.0,
      "step": 1033
    },
    {
      "epoch": 0.19201485608170846,
      "grad_norm": 1.6535506253056476,
      "learning_rate": 6.392326732673267e-07,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8491488695144653,
      "num_tokens": 35817098.0,
      "step": 1034
    },
    {
      "epoch": 0.19220055710306408,
      "grad_norm": 1.6767867946299218,
      "learning_rate": 6.398514851485148e-07,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8572200536727905,
      "num_tokens": 35845771.0,
      "step": 1035
    },
    {
      "epoch": 0.19238625812441967,
      "grad_norm": 1.661021527416388,
      "learning_rate": 6.404702970297029e-07,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8426978588104248,
      "num_tokens": 35878163.0,
      "step": 1036
    },
    {
      "epoch": 0.1925719591457753,
      "grad_norm": 1.4422389484946285,
      "learning_rate": 6.41089108910891e-07,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8375098705291748,
      "num_tokens": 35919328.0,
      "step": 1037
    },
    {
      "epoch": 0.19275766016713092,
      "grad_norm": 1.699979335146808,
      "learning_rate": 6.417079207920791e-07,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8465098738670349,
      "num_tokens": 35954241.0,
      "step": 1038
    },
    {
      "epoch": 0.19294336118848654,
      "grad_norm": 1.708649834330357,
      "learning_rate": 6.423267326732673e-07,
      "loss": 0.478,
      "mean_token_accuracy": 0.8458555340766907,
      "num_tokens": 35986053.0,
      "step": 1039
    },
    {
      "epoch": 0.19312906220984216,
      "grad_norm": 1.5228303167052173,
      "learning_rate": 6.429455445544553e-07,
      "loss": 0.4089,
      "mean_token_accuracy": 0.867418646812439,
      "num_tokens": 36017123.0,
      "step": 1040
    },
    {
      "epoch": 0.19331476323119778,
      "grad_norm": 1.5024644813154193,
      "learning_rate": 6.435643564356436e-07,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8627349734306335,
      "num_tokens": 36048739.0,
      "step": 1041
    },
    {
      "epoch": 0.19350046425255338,
      "grad_norm": 1.4382044895251018,
      "learning_rate": 6.441831683168317e-07,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8638637065887451,
      "num_tokens": 36086279.0,
      "step": 1042
    },
    {
      "epoch": 0.193686165273909,
      "grad_norm": 1.6258578885000217,
      "learning_rate": 6.448019801980198e-07,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8437725305557251,
      "num_tokens": 36120513.0,
      "step": 1043
    },
    {
      "epoch": 0.19387186629526462,
      "grad_norm": 1.6095626104009,
      "learning_rate": 6.454207920792079e-07,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8605449199676514,
      "num_tokens": 36154185.0,
      "step": 1044
    },
    {
      "epoch": 0.19405756731662024,
      "grad_norm": 1.6003464248546837,
      "learning_rate": 6.46039603960396e-07,
      "loss": 0.4402,
      "mean_token_accuracy": 0.856070876121521,
      "num_tokens": 36187842.0,
      "step": 1045
    },
    {
      "epoch": 0.19424326833797587,
      "grad_norm": 1.5119864073107778,
      "learning_rate": 6.466584158415841e-07,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8462842702865601,
      "num_tokens": 36226488.0,
      "step": 1046
    },
    {
      "epoch": 0.1944289693593315,
      "grad_norm": 1.6447554891246816,
      "learning_rate": 6.472772277227723e-07,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8529120683670044,
      "num_tokens": 36258177.0,
      "step": 1047
    },
    {
      "epoch": 0.19461467038068708,
      "grad_norm": 1.589203584942928,
      "learning_rate": 6.478960396039603e-07,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8387005925178528,
      "num_tokens": 36290163.0,
      "step": 1048
    },
    {
      "epoch": 0.1948003714020427,
      "grad_norm": 1.4836621075981793,
      "learning_rate": 6.485148514851485e-07,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8612295389175415,
      "num_tokens": 36324149.0,
      "step": 1049
    },
    {
      "epoch": 0.19498607242339833,
      "grad_norm": 1.524855662533524,
      "learning_rate": 6.491336633663366e-07,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8458570837974548,
      "num_tokens": 36363629.0,
      "step": 1050
    },
    {
      "epoch": 0.19517177344475395,
      "grad_norm": 1.5548277085937496,
      "learning_rate": 6.497524752475247e-07,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8406417965888977,
      "num_tokens": 36400515.0,
      "step": 1051
    },
    {
      "epoch": 0.19535747446610957,
      "grad_norm": 1.6737328975803416,
      "learning_rate": 6.503712871287128e-07,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8541032075881958,
      "num_tokens": 36435726.0,
      "step": 1052
    },
    {
      "epoch": 0.1955431754874652,
      "grad_norm": 1.580966921280327,
      "learning_rate": 6.509900990099009e-07,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8649584650993347,
      "num_tokens": 36468319.0,
      "step": 1053
    },
    {
      "epoch": 0.1957288765088208,
      "grad_norm": 1.3648803859607352,
      "learning_rate": 6.51608910891089e-07,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8537872433662415,
      "num_tokens": 36507275.0,
      "step": 1054
    },
    {
      "epoch": 0.1959145775301764,
      "grad_norm": 1.5163819866143706,
      "learning_rate": 6.522277227722773e-07,
      "loss": 0.3967,
      "mean_token_accuracy": 0.868969202041626,
      "num_tokens": 36539166.0,
      "step": 1055
    },
    {
      "epoch": 0.19610027855153203,
      "grad_norm": 1.7489660181741031,
      "learning_rate": 6.528465346534653e-07,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8651814460754395,
      "num_tokens": 36561966.0,
      "step": 1056
    },
    {
      "epoch": 0.19628597957288765,
      "grad_norm": 1.6473444794547816,
      "learning_rate": 6.534653465346535e-07,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8531517386436462,
      "num_tokens": 36597701.0,
      "step": 1057
    },
    {
      "epoch": 0.19647168059424328,
      "grad_norm": 1.741733568512983,
      "learning_rate": 6.540841584158415e-07,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8523199558258057,
      "num_tokens": 36625909.0,
      "step": 1058
    },
    {
      "epoch": 0.1966573816155989,
      "grad_norm": 1.9087347135224975,
      "learning_rate": 6.547029702970297e-07,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8498688340187073,
      "num_tokens": 36653528.0,
      "step": 1059
    },
    {
      "epoch": 0.1968430826369545,
      "grad_norm": 1.5725338888800353,
      "learning_rate": 6.553217821782178e-07,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8561699390411377,
      "num_tokens": 36686299.0,
      "step": 1060
    },
    {
      "epoch": 0.1970287836583101,
      "grad_norm": 1.419827161582915,
      "learning_rate": 6.559405940594059e-07,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8734182119369507,
      "num_tokens": 36722010.0,
      "step": 1061
    },
    {
      "epoch": 0.19721448467966574,
      "grad_norm": 1.602647664285472,
      "learning_rate": 6.56559405940594e-07,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8457804918289185,
      "num_tokens": 36753230.0,
      "step": 1062
    },
    {
      "epoch": 0.19740018570102136,
      "grad_norm": 1.566293148509468,
      "learning_rate": 6.571782178217822e-07,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8513582944869995,
      "num_tokens": 36784561.0,
      "step": 1063
    },
    {
      "epoch": 0.19758588672237698,
      "grad_norm": 1.5209339974874652,
      "learning_rate": 6.577970297029702e-07,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8516079187393188,
      "num_tokens": 36822734.0,
      "step": 1064
    },
    {
      "epoch": 0.1977715877437326,
      "grad_norm": 1.6486975918603188,
      "learning_rate": 6.584158415841584e-07,
      "loss": 0.5188,
      "mean_token_accuracy": 0.834179162979126,
      "num_tokens": 36855011.0,
      "step": 1065
    },
    {
      "epoch": 0.1979572887650882,
      "grad_norm": 1.6040192617803148,
      "learning_rate": 6.590346534653464e-07,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8539116978645325,
      "num_tokens": 36891377.0,
      "step": 1066
    },
    {
      "epoch": 0.19814298978644382,
      "grad_norm": 1.4217693106582499,
      "learning_rate": 6.596534653465346e-07,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8591241836547852,
      "num_tokens": 36930180.0,
      "step": 1067
    },
    {
      "epoch": 0.19832869080779944,
      "grad_norm": 1.4238486126094885,
      "learning_rate": 6.602722772277228e-07,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8635877370834351,
      "num_tokens": 36967231.0,
      "step": 1068
    },
    {
      "epoch": 0.19851439182915506,
      "grad_norm": 1.5030801493220782,
      "learning_rate": 6.608910891089109e-07,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8589342832565308,
      "num_tokens": 37003603.0,
      "step": 1069
    },
    {
      "epoch": 0.19870009285051068,
      "grad_norm": 1.4602617313404798,
      "learning_rate": 6.61509900990099e-07,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8481877446174622,
      "num_tokens": 37045294.0,
      "step": 1070
    },
    {
      "epoch": 0.1988857938718663,
      "grad_norm": 1.4959216661304298,
      "learning_rate": 6.621287128712872e-07,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8392311334609985,
      "num_tokens": 37084524.0,
      "step": 1071
    },
    {
      "epoch": 0.1990714948932219,
      "grad_norm": 1.5120684533274382,
      "learning_rate": 6.627475247524752e-07,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8603960275650024,
      "num_tokens": 37120849.0,
      "step": 1072
    },
    {
      "epoch": 0.19925719591457752,
      "grad_norm": 1.5385604613365618,
      "learning_rate": 6.633663366336634e-07,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8463382720947266,
      "num_tokens": 37159154.0,
      "step": 1073
    },
    {
      "epoch": 0.19944289693593314,
      "grad_norm": 1.4892456383187003,
      "learning_rate": 6.639851485148514e-07,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8404366374015808,
      "num_tokens": 37198296.0,
      "step": 1074
    },
    {
      "epoch": 0.19962859795728877,
      "grad_norm": 1.8203164813696293,
      "learning_rate": 6.646039603960396e-07,
      "loss": 0.5294,
      "mean_token_accuracy": 0.832427978515625,
      "num_tokens": 37229426.0,
      "step": 1075
    },
    {
      "epoch": 0.1998142989786444,
      "grad_norm": 1.3904111339683276,
      "learning_rate": 6.652227722772277e-07,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8535199165344238,
      "num_tokens": 37271114.0,
      "step": 1076
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6247216181533497,
      "learning_rate": 6.658415841584158e-07,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8490339517593384,
      "num_tokens": 37304350.0,
      "step": 1077
    },
    {
      "epoch": 0.2001857010213556,
      "grad_norm": 1.5398685711978248,
      "learning_rate": 6.664603960396039e-07,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8450025916099548,
      "num_tokens": 37338081.0,
      "step": 1078
    },
    {
      "epoch": 0.20037140204271123,
      "grad_norm": 1.6207840661308606,
      "learning_rate": 6.67079207920792e-07,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8483004570007324,
      "num_tokens": 37370643.0,
      "step": 1079
    },
    {
      "epoch": 0.20055710306406685,
      "grad_norm": 1.514450940875639,
      "learning_rate": 6.676980198019801e-07,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8566693067550659,
      "num_tokens": 37403615.0,
      "step": 1080
    },
    {
      "epoch": 0.20074280408542247,
      "grad_norm": 1.6055486906823198,
      "learning_rate": 6.683168316831684e-07,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8436105251312256,
      "num_tokens": 37440146.0,
      "step": 1081
    },
    {
      "epoch": 0.2009285051067781,
      "grad_norm": 1.540711425742739,
      "learning_rate": 6.689356435643564e-07,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8359460830688477,
      "num_tokens": 37481457.0,
      "step": 1082
    },
    {
      "epoch": 0.20111420612813372,
      "grad_norm": 1.6708800761864024,
      "learning_rate": 6.695544554455446e-07,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8332208395004272,
      "num_tokens": 37514545.0,
      "step": 1083
    },
    {
      "epoch": 0.2012999071494893,
      "grad_norm": 1.7199291508737682,
      "learning_rate": 6.701732673267327e-07,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8330401182174683,
      "num_tokens": 37545613.0,
      "step": 1084
    },
    {
      "epoch": 0.20148560817084493,
      "grad_norm": 1.7583606222338592,
      "learning_rate": 6.707920792079208e-07,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8500245809555054,
      "num_tokens": 37576955.0,
      "step": 1085
    },
    {
      "epoch": 0.20167130919220055,
      "grad_norm": 1.5508397832876195,
      "learning_rate": 6.714108910891089e-07,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8447055220603943,
      "num_tokens": 37612008.0,
      "step": 1086
    },
    {
      "epoch": 0.20185701021355618,
      "grad_norm": 1.6723852562476222,
      "learning_rate": 6.72029702970297e-07,
      "loss": 0.4982,
      "mean_token_accuracy": 0.84029221534729,
      "num_tokens": 37645922.0,
      "step": 1087
    },
    {
      "epoch": 0.2020427112349118,
      "grad_norm": 1.4646495002389586,
      "learning_rate": 6.726485148514851e-07,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8465484976768494,
      "num_tokens": 37684968.0,
      "step": 1088
    },
    {
      "epoch": 0.20222841225626742,
      "grad_norm": 1.606009812348051,
      "learning_rate": 6.732673267326733e-07,
      "loss": 0.4239,
      "mean_token_accuracy": 0.859096884727478,
      "num_tokens": 37715605.0,
      "step": 1089
    },
    {
      "epoch": 0.20241411327762301,
      "grad_norm": 1.5438014693048299,
      "learning_rate": 6.738861386138613e-07,
      "loss": 0.467,
      "mean_token_accuracy": 0.8483836650848389,
      "num_tokens": 37752979.0,
      "step": 1090
    },
    {
      "epoch": 0.20259981429897864,
      "grad_norm": 1.5713104498754455,
      "learning_rate": 6.745049504950495e-07,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8445305824279785,
      "num_tokens": 37786348.0,
      "step": 1091
    },
    {
      "epoch": 0.20278551532033426,
      "grad_norm": 1.5045813933817513,
      "learning_rate": 6.751237623762375e-07,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8438611030578613,
      "num_tokens": 37820779.0,
      "step": 1092
    },
    {
      "epoch": 0.20297121634168988,
      "grad_norm": 1.4471807692071983,
      "learning_rate": 6.757425742574257e-07,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8564512729644775,
      "num_tokens": 37857287.0,
      "step": 1093
    },
    {
      "epoch": 0.2031569173630455,
      "grad_norm": 1.62289138924886,
      "learning_rate": 6.763613861386139e-07,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8447400331497192,
      "num_tokens": 37888962.0,
      "step": 1094
    },
    {
      "epoch": 0.20334261838440112,
      "grad_norm": 1.4137340156007436,
      "learning_rate": 6.76980198019802e-07,
      "loss": 0.369,
      "mean_token_accuracy": 0.8758018016815186,
      "num_tokens": 37926062.0,
      "step": 1095
    },
    {
      "epoch": 0.20352831940575672,
      "grad_norm": 1.6167583630855125,
      "learning_rate": 6.775990099009901e-07,
      "loss": 0.4988,
      "mean_token_accuracy": 0.838111937046051,
      "num_tokens": 37961960.0,
      "step": 1096
    },
    {
      "epoch": 0.20371402042711234,
      "grad_norm": 1.4235060311280352,
      "learning_rate": 6.782178217821783e-07,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8496158719062805,
      "num_tokens": 38001190.0,
      "step": 1097
    },
    {
      "epoch": 0.20389972144846796,
      "grad_norm": 1.4935973119919794,
      "learning_rate": 6.788366336633663e-07,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8612439632415771,
      "num_tokens": 38036997.0,
      "step": 1098
    },
    {
      "epoch": 0.20408542246982359,
      "grad_norm": 1.49921825439365,
      "learning_rate": 6.794554455445545e-07,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8529655933380127,
      "num_tokens": 38071843.0,
      "step": 1099
    },
    {
      "epoch": 0.2042711234911792,
      "grad_norm": 1.5602409721639936,
      "learning_rate": 6.800742574257425e-07,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8518495559692383,
      "num_tokens": 38102477.0,
      "step": 1100
    },
    {
      "epoch": 0.20445682451253483,
      "grad_norm": 1.5814278654078144,
      "learning_rate": 6.806930693069307e-07,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8494694232940674,
      "num_tokens": 38136590.0,
      "step": 1101
    },
    {
      "epoch": 0.20464252553389042,
      "grad_norm": 1.8347310536020744,
      "learning_rate": 6.813118811881188e-07,
      "loss": 0.4543,
      "mean_token_accuracy": 0.854085385799408,
      "num_tokens": 38161672.0,
      "step": 1102
    },
    {
      "epoch": 0.20482822655524605,
      "grad_norm": 1.287289774819096,
      "learning_rate": 6.819306930693069e-07,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8720595240592957,
      "num_tokens": 38205461.0,
      "step": 1103
    },
    {
      "epoch": 0.20501392757660167,
      "grad_norm": 1.8306683937295976,
      "learning_rate": 6.82549504950495e-07,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8330202102661133,
      "num_tokens": 38235873.0,
      "step": 1104
    },
    {
      "epoch": 0.2051996285979573,
      "grad_norm": 1.5477272014673935,
      "learning_rate": 6.831683168316831e-07,
      "loss": 0.465,
      "mean_token_accuracy": 0.8473027944564819,
      "num_tokens": 38269357.0,
      "step": 1105
    },
    {
      "epoch": 0.2053853296193129,
      "grad_norm": 1.5211773671360593,
      "learning_rate": 6.837871287128712e-07,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8521249890327454,
      "num_tokens": 38302455.0,
      "step": 1106
    },
    {
      "epoch": 0.20557103064066853,
      "grad_norm": 1.5799114997266221,
      "learning_rate": 6.844059405940595e-07,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8573321104049683,
      "num_tokens": 38332496.0,
      "step": 1107
    },
    {
      "epoch": 0.20575673166202413,
      "grad_norm": 1.4601210409240566,
      "learning_rate": 6.850247524752475e-07,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8487905263900757,
      "num_tokens": 38370819.0,
      "step": 1108
    },
    {
      "epoch": 0.20594243268337975,
      "grad_norm": 1.531189874262646,
      "learning_rate": 6.856435643564357e-07,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8479712009429932,
      "num_tokens": 38405200.0,
      "step": 1109
    },
    {
      "epoch": 0.20612813370473537,
      "grad_norm": 1.6143287393252967,
      "learning_rate": 6.862623762376238e-07,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8278758525848389,
      "num_tokens": 38440524.0,
      "step": 1110
    },
    {
      "epoch": 0.206313834726091,
      "grad_norm": 1.6126546073088364,
      "learning_rate": 6.868811881188119e-07,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8511530160903931,
      "num_tokens": 38473414.0,
      "step": 1111
    },
    {
      "epoch": 0.20649953574744662,
      "grad_norm": 1.5704102484117142,
      "learning_rate": 6.875e-07,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8489397764205933,
      "num_tokens": 38507880.0,
      "step": 1112
    },
    {
      "epoch": 0.20668523676880224,
      "grad_norm": 1.4649154029877172,
      "learning_rate": 6.88118811881188e-07,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8417960405349731,
      "num_tokens": 38547117.0,
      "step": 1113
    },
    {
      "epoch": 0.20687093779015783,
      "grad_norm": 1.5152838903679948,
      "learning_rate": 6.887376237623762e-07,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8495540022850037,
      "num_tokens": 38580570.0,
      "step": 1114
    },
    {
      "epoch": 0.20705663881151345,
      "grad_norm": 1.640519310080729,
      "learning_rate": 6.893564356435643e-07,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8489342331886292,
      "num_tokens": 38612562.0,
      "step": 1115
    },
    {
      "epoch": 0.20724233983286908,
      "grad_norm": 1.5712869583171167,
      "learning_rate": 6.899752475247524e-07,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8605951070785522,
      "num_tokens": 38644705.0,
      "step": 1116
    },
    {
      "epoch": 0.2074280408542247,
      "grad_norm": 1.6578545432801943,
      "learning_rate": 6.905940594059405e-07,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8339406251907349,
      "num_tokens": 38683184.0,
      "step": 1117
    },
    {
      "epoch": 0.20761374187558032,
      "grad_norm": 1.608576314632886,
      "learning_rate": 6.912128712871287e-07,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8516817092895508,
      "num_tokens": 38718338.0,
      "step": 1118
    },
    {
      "epoch": 0.20779944289693594,
      "grad_norm": 1.4818467619952251,
      "learning_rate": 6.918316831683167e-07,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8531838059425354,
      "num_tokens": 38758263.0,
      "step": 1119
    },
    {
      "epoch": 0.20798514391829154,
      "grad_norm": 1.7285874795581029,
      "learning_rate": 6.924504950495049e-07,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8530492782592773,
      "num_tokens": 38795061.0,
      "step": 1120
    },
    {
      "epoch": 0.20817084493964716,
      "grad_norm": 1.4676653412773741,
      "learning_rate": 6.93069306930693e-07,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8497323393821716,
      "num_tokens": 38833994.0,
      "step": 1121
    },
    {
      "epoch": 0.20835654596100278,
      "grad_norm": 1.6492954994697964,
      "learning_rate": 6.936881188118812e-07,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8421061038970947,
      "num_tokens": 38864039.0,
      "step": 1122
    },
    {
      "epoch": 0.2085422469823584,
      "grad_norm": 1.5017767272322455,
      "learning_rate": 6.943069306930693e-07,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8464198112487793,
      "num_tokens": 38900923.0,
      "step": 1123
    },
    {
      "epoch": 0.20872794800371403,
      "grad_norm": 1.6096976591200738,
      "learning_rate": 6.949257425742574e-07,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8513097167015076,
      "num_tokens": 38934405.0,
      "step": 1124
    },
    {
      "epoch": 0.20891364902506965,
      "grad_norm": 1.5266411268152533,
      "learning_rate": 6.955445544554455e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8506852388381958,
      "num_tokens": 38971056.0,
      "step": 1125
    },
    {
      "epoch": 0.20909935004642524,
      "grad_norm": 1.6390032387627336,
      "learning_rate": 6.961633663366336e-07,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8418805003166199,
      "num_tokens": 39003283.0,
      "step": 1126
    },
    {
      "epoch": 0.20928505106778086,
      "grad_norm": 1.9322841713079815,
      "learning_rate": 6.967821782178217e-07,
      "loss": 0.504,
      "mean_token_accuracy": 0.8361916542053223,
      "num_tokens": 39035570.0,
      "step": 1127
    },
    {
      "epoch": 0.20947075208913649,
      "grad_norm": 1.4869107597853375,
      "learning_rate": 6.974009900990099e-07,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8524374961853027,
      "num_tokens": 39070226.0,
      "step": 1128
    },
    {
      "epoch": 0.2096564531104921,
      "grad_norm": 1.6318914109103926,
      "learning_rate": 6.980198019801979e-07,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8294348120689392,
      "num_tokens": 39105617.0,
      "step": 1129
    },
    {
      "epoch": 0.20984215413184773,
      "grad_norm": 1.3888691764178656,
      "learning_rate": 6.986386138613861e-07,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8488368988037109,
      "num_tokens": 39145592.0,
      "step": 1130
    },
    {
      "epoch": 0.21002785515320335,
      "grad_norm": 1.4873236008285637,
      "learning_rate": 6.992574257425742e-07,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8603816628456116,
      "num_tokens": 39181281.0,
      "step": 1131
    },
    {
      "epoch": 0.21021355617455895,
      "grad_norm": 1.477049603220808,
      "learning_rate": 6.998762376237623e-07,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8490806818008423,
      "num_tokens": 39217811.0,
      "step": 1132
    },
    {
      "epoch": 0.21039925719591457,
      "grad_norm": 1.6107350231201498,
      "learning_rate": 7.004950495049504e-07,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8518853187561035,
      "num_tokens": 39251617.0,
      "step": 1133
    },
    {
      "epoch": 0.2105849582172702,
      "grad_norm": 1.6444415882330163,
      "learning_rate": 7.011138613861386e-07,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8428956270217896,
      "num_tokens": 39282544.0,
      "step": 1134
    },
    {
      "epoch": 0.2107706592386258,
      "grad_norm": 1.5423298639564775,
      "learning_rate": 7.017326732673267e-07,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8676052689552307,
      "num_tokens": 39315547.0,
      "step": 1135
    },
    {
      "epoch": 0.21095636025998143,
      "grad_norm": 1.3630664528231764,
      "learning_rate": 7.023514851485149e-07,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8487175703048706,
      "num_tokens": 39359045.0,
      "step": 1136
    },
    {
      "epoch": 0.21114206128133706,
      "grad_norm": 1.6091755263069953,
      "learning_rate": 7.029702970297029e-07,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8448742628097534,
      "num_tokens": 39393697.0,
      "step": 1137
    },
    {
      "epoch": 0.21132776230269265,
      "grad_norm": 1.3901421559578961,
      "learning_rate": 7.035891089108911e-07,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8549783229827881,
      "num_tokens": 39439495.0,
      "step": 1138
    },
    {
      "epoch": 0.21151346332404827,
      "grad_norm": 1.530629195089977,
      "learning_rate": 7.042079207920791e-07,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8423123359680176,
      "num_tokens": 39480058.0,
      "step": 1139
    },
    {
      "epoch": 0.2116991643454039,
      "grad_norm": 1.5307034793501988,
      "learning_rate": 7.048267326732673e-07,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8552366495132446,
      "num_tokens": 39516867.0,
      "step": 1140
    },
    {
      "epoch": 0.21188486536675952,
      "grad_norm": 1.795688784263433,
      "learning_rate": 7.054455445544554e-07,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8374035358428955,
      "num_tokens": 39546116.0,
      "step": 1141
    },
    {
      "epoch": 0.21207056638811514,
      "grad_norm": 1.5321879524386066,
      "learning_rate": 7.060643564356435e-07,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8552563190460205,
      "num_tokens": 39582006.0,
      "step": 1142
    },
    {
      "epoch": 0.21225626740947076,
      "grad_norm": 1.832552661807456,
      "learning_rate": 7.066831683168316e-07,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8387222290039062,
      "num_tokens": 39611214.0,
      "step": 1143
    },
    {
      "epoch": 0.21244196843082638,
      "grad_norm": 1.7253171787265227,
      "learning_rate": 7.073019801980198e-07,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8583498001098633,
      "num_tokens": 39638345.0,
      "step": 1144
    },
    {
      "epoch": 0.21262766945218198,
      "grad_norm": 1.570980384932258,
      "learning_rate": 7.079207920792078e-07,
      "loss": 0.444,
      "mean_token_accuracy": 0.8546789884567261,
      "num_tokens": 39673384.0,
      "step": 1145
    },
    {
      "epoch": 0.2128133704735376,
      "grad_norm": 1.6422680566749388,
      "learning_rate": 7.08539603960396e-07,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8573323488235474,
      "num_tokens": 39706325.0,
      "step": 1146
    },
    {
      "epoch": 0.21299907149489322,
      "grad_norm": 1.4066979304721519,
      "learning_rate": 7.091584158415841e-07,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8648391962051392,
      "num_tokens": 39745275.0,
      "step": 1147
    },
    {
      "epoch": 0.21318477251624884,
      "grad_norm": 1.560961592546892,
      "learning_rate": 7.097772277227723e-07,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8548338413238525,
      "num_tokens": 39778724.0,
      "step": 1148
    },
    {
      "epoch": 0.21337047353760447,
      "grad_norm": 1.4005870908643328,
      "learning_rate": 7.103960396039604e-07,
      "loss": 0.4435,
      "mean_token_accuracy": 0.857161283493042,
      "num_tokens": 39820259.0,
      "step": 1149
    },
    {
      "epoch": 0.2135561745589601,
      "grad_norm": 1.4603192882829037,
      "learning_rate": 7.110148514851485e-07,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8530200719833374,
      "num_tokens": 39859837.0,
      "step": 1150
    },
    {
      "epoch": 0.21374187558031568,
      "grad_norm": 1.6706773256431238,
      "learning_rate": 7.116336633663366e-07,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8605363368988037,
      "num_tokens": 39893382.0,
      "step": 1151
    },
    {
      "epoch": 0.2139275766016713,
      "grad_norm": 1.597522324972759,
      "learning_rate": 7.122524752475248e-07,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8391976356506348,
      "num_tokens": 39931941.0,
      "step": 1152
    },
    {
      "epoch": 0.21411327762302693,
      "grad_norm": 1.4434294156883307,
      "learning_rate": 7.128712871287128e-07,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8526210188865662,
      "num_tokens": 39969030.0,
      "step": 1153
    },
    {
      "epoch": 0.21429897864438255,
      "grad_norm": 1.4885805726119725,
      "learning_rate": 7.13490099009901e-07,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8449684381484985,
      "num_tokens": 40006938.0,
      "step": 1154
    },
    {
      "epoch": 0.21448467966573817,
      "grad_norm": 1.5947954569755771,
      "learning_rate": 7.14108910891089e-07,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8575215935707092,
      "num_tokens": 40039374.0,
      "step": 1155
    },
    {
      "epoch": 0.2146703806870938,
      "grad_norm": 1.4698489426854489,
      "learning_rate": 7.147277227722772e-07,
      "loss": 0.3976,
      "mean_token_accuracy": 0.867982804775238,
      "num_tokens": 40075210.0,
      "step": 1156
    },
    {
      "epoch": 0.2148560817084494,
      "grad_norm": 1.564898666635383,
      "learning_rate": 7.153465346534653e-07,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8545830249786377,
      "num_tokens": 40105741.0,
      "step": 1157
    },
    {
      "epoch": 0.215041782729805,
      "grad_norm": 1.5537169241260118,
      "learning_rate": 7.159653465346534e-07,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8495890498161316,
      "num_tokens": 40140371.0,
      "step": 1158
    },
    {
      "epoch": 0.21522748375116063,
      "grad_norm": 1.6534865040991054,
      "learning_rate": 7.165841584158415e-07,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8492103815078735,
      "num_tokens": 40171036.0,
      "step": 1159
    },
    {
      "epoch": 0.21541318477251625,
      "grad_norm": 1.5660151718108637,
      "learning_rate": 7.172029702970297e-07,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8430261611938477,
      "num_tokens": 40205697.0,
      "step": 1160
    },
    {
      "epoch": 0.21559888579387188,
      "grad_norm": 1.4854075616582607,
      "learning_rate": 7.178217821782178e-07,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8605294227600098,
      "num_tokens": 40239155.0,
      "step": 1161
    },
    {
      "epoch": 0.2157845868152275,
      "grad_norm": 1.6014195685733543,
      "learning_rate": 7.18440594059406e-07,
      "loss": 0.5028,
      "mean_token_accuracy": 0.83808833360672,
      "num_tokens": 40274359.0,
      "step": 1162
    },
    {
      "epoch": 0.2159702878365831,
      "grad_norm": 1.7840998738399931,
      "learning_rate": 7.19059405940594e-07,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8487063646316528,
      "num_tokens": 40309050.0,
      "step": 1163
    },
    {
      "epoch": 0.2161559888579387,
      "grad_norm": 1.6263315195686936,
      "learning_rate": 7.196782178217822e-07,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8274312019348145,
      "num_tokens": 40346775.0,
      "step": 1164
    },
    {
      "epoch": 0.21634168987929434,
      "grad_norm": 1.4903396477522286,
      "learning_rate": 7.202970297029703e-07,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8364557027816772,
      "num_tokens": 40386437.0,
      "step": 1165
    },
    {
      "epoch": 0.21652739090064996,
      "grad_norm": 1.5984179185414154,
      "learning_rate": 7.209158415841584e-07,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8498944640159607,
      "num_tokens": 40419560.0,
      "step": 1166
    },
    {
      "epoch": 0.21671309192200558,
      "grad_norm": 1.586205999148253,
      "learning_rate": 7.215346534653465e-07,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8526652455329895,
      "num_tokens": 40454286.0,
      "step": 1167
    },
    {
      "epoch": 0.2168987929433612,
      "grad_norm": 1.62256987384519,
      "learning_rate": 7.221534653465346e-07,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8533637523651123,
      "num_tokens": 40485180.0,
      "step": 1168
    },
    {
      "epoch": 0.2170844939647168,
      "grad_norm": 1.5902326333646768,
      "learning_rate": 7.227722772277227e-07,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8593475818634033,
      "num_tokens": 40519679.0,
      "step": 1169
    },
    {
      "epoch": 0.21727019498607242,
      "grad_norm": 2.3260198069273166,
      "learning_rate": 7.233910891089109e-07,
      "loss": 0.4908,
      "mean_token_accuracy": 0.842448353767395,
      "num_tokens": 40552099.0,
      "step": 1170
    },
    {
      "epoch": 0.21745589600742804,
      "grad_norm": 1.8043404425016527,
      "learning_rate": 7.240099009900989e-07,
      "loss": 0.435,
      "mean_token_accuracy": 0.8565565347671509,
      "num_tokens": 40581938.0,
      "step": 1171
    },
    {
      "epoch": 0.21764159702878366,
      "grad_norm": 1.571398817269373,
      "learning_rate": 7.246287128712871e-07,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8576446175575256,
      "num_tokens": 40614816.0,
      "step": 1172
    },
    {
      "epoch": 0.21782729805013928,
      "grad_norm": 1.614320070650364,
      "learning_rate": 7.252475247524751e-07,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8423371315002441,
      "num_tokens": 40647822.0,
      "step": 1173
    },
    {
      "epoch": 0.2180129990714949,
      "grad_norm": 1.6842941883547458,
      "learning_rate": 7.258663366336634e-07,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8582186698913574,
      "num_tokens": 40675466.0,
      "step": 1174
    },
    {
      "epoch": 0.2181987000928505,
      "grad_norm": 1.5844494124437658,
      "learning_rate": 7.264851485148515e-07,
      "loss": 0.4405,
      "mean_token_accuracy": 0.855905294418335,
      "num_tokens": 40709161.0,
      "step": 1175
    },
    {
      "epoch": 0.21838440111420612,
      "grad_norm": 1.476939623792746,
      "learning_rate": 7.271039603960396e-07,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8619754314422607,
      "num_tokens": 40743224.0,
      "step": 1176
    },
    {
      "epoch": 0.21857010213556174,
      "grad_norm": 1.8242820167405407,
      "learning_rate": 7.277227722772277e-07,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8324992060661316,
      "num_tokens": 40772905.0,
      "step": 1177
    },
    {
      "epoch": 0.21875580315691737,
      "grad_norm": 1.7081681901226422,
      "learning_rate": 7.283415841584159e-07,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8433182835578918,
      "num_tokens": 40807257.0,
      "step": 1178
    },
    {
      "epoch": 0.218941504178273,
      "grad_norm": 1.6502207812093812,
      "learning_rate": 7.289603960396039e-07,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8434478044509888,
      "num_tokens": 40840445.0,
      "step": 1179
    },
    {
      "epoch": 0.2191272051996286,
      "grad_norm": 1.713192319774037,
      "learning_rate": 7.295792079207921e-07,
      "loss": 0.504,
      "mean_token_accuracy": 0.8375116586685181,
      "num_tokens": 40869205.0,
      "step": 1180
    },
    {
      "epoch": 0.2193129062209842,
      "grad_norm": 1.565230814297299,
      "learning_rate": 7.301980198019801e-07,
      "loss": 0.447,
      "mean_token_accuracy": 0.8525257110595703,
      "num_tokens": 40905016.0,
      "step": 1181
    },
    {
      "epoch": 0.21949860724233983,
      "grad_norm": 1.4468778020475186,
      "learning_rate": 7.308168316831683e-07,
      "loss": 0.462,
      "mean_token_accuracy": 0.849822998046875,
      "num_tokens": 40945067.0,
      "step": 1182
    },
    {
      "epoch": 0.21968430826369545,
      "grad_norm": 1.6185510007164259,
      "learning_rate": 7.314356435643564e-07,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8593991994857788,
      "num_tokens": 40976406.0,
      "step": 1183
    },
    {
      "epoch": 0.21987000928505107,
      "grad_norm": 1.560156488508912,
      "learning_rate": 7.320544554455445e-07,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8681668043136597,
      "num_tokens": 41006260.0,
      "step": 1184
    },
    {
      "epoch": 0.2200557103064067,
      "grad_norm": 1.4332076046444777,
      "learning_rate": 7.326732673267326e-07,
      "loss": 0.462,
      "mean_token_accuracy": 0.8511723875999451,
      "num_tokens": 41045612.0,
      "step": 1185
    },
    {
      "epoch": 0.22024141132776232,
      "grad_norm": 1.5113254094119544,
      "learning_rate": 7.332920792079207e-07,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8502991199493408,
      "num_tokens": 41084007.0,
      "step": 1186
    },
    {
      "epoch": 0.2204271123491179,
      "grad_norm": 1.7915521597855626,
      "learning_rate": 7.339108910891089e-07,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8372111320495605,
      "num_tokens": 41115974.0,
      "step": 1187
    },
    {
      "epoch": 0.22061281337047353,
      "grad_norm": 1.6090998955475142,
      "learning_rate": 7.345297029702971e-07,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8497635126113892,
      "num_tokens": 41145793.0,
      "step": 1188
    },
    {
      "epoch": 0.22079851439182915,
      "grad_norm": 1.4804827945277965,
      "learning_rate": 7.351485148514851e-07,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8440642356872559,
      "num_tokens": 41186728.0,
      "step": 1189
    },
    {
      "epoch": 0.22098421541318478,
      "grad_norm": 1.497971100165895,
      "learning_rate": 7.357673267326733e-07,
      "loss": 0.5052,
      "mean_token_accuracy": 0.836943507194519,
      "num_tokens": 41227635.0,
      "step": 1190
    },
    {
      "epoch": 0.2211699164345404,
      "grad_norm": 1.537993128249195,
      "learning_rate": 7.363861386138614e-07,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8621293902397156,
      "num_tokens": 41259276.0,
      "step": 1191
    },
    {
      "epoch": 0.22135561745589602,
      "grad_norm": 1.6745737870953914,
      "learning_rate": 7.370049504950495e-07,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8554417490959167,
      "num_tokens": 41286428.0,
      "step": 1192
    },
    {
      "epoch": 0.22154131847725161,
      "grad_norm": 1.6461188607318733,
      "learning_rate": 7.376237623762376e-07,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8524162769317627,
      "num_tokens": 41318159.0,
      "step": 1193
    },
    {
      "epoch": 0.22172701949860724,
      "grad_norm": 1.832037092435771,
      "learning_rate": 7.382425742574257e-07,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8461979627609253,
      "num_tokens": 41345109.0,
      "step": 1194
    },
    {
      "epoch": 0.22191272051996286,
      "grad_norm": 1.4237528227046825,
      "learning_rate": 7.388613861386138e-07,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8681115508079529,
      "num_tokens": 41379586.0,
      "step": 1195
    },
    {
      "epoch": 0.22209842154131848,
      "grad_norm": 1.5793653452244112,
      "learning_rate": 7.39480198019802e-07,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8555759787559509,
      "num_tokens": 41415112.0,
      "step": 1196
    },
    {
      "epoch": 0.2222841225626741,
      "grad_norm": 1.5840859346622211,
      "learning_rate": 7.4009900990099e-07,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8546200394630432,
      "num_tokens": 41448603.0,
      "step": 1197
    },
    {
      "epoch": 0.22246982358402972,
      "grad_norm": 1.5763812708947367,
      "learning_rate": 7.407178217821782e-07,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8544106483459473,
      "num_tokens": 41481647.0,
      "step": 1198
    },
    {
      "epoch": 0.22265552460538532,
      "grad_norm": 1.4931182087187944,
      "learning_rate": 7.413366336633662e-07,
      "loss": 0.436,
      "mean_token_accuracy": 0.8575544357299805,
      "num_tokens": 41519223.0,
      "step": 1199
    },
    {
      "epoch": 0.22284122562674094,
      "grad_norm": 1.5257913110610046,
      "learning_rate": 7.419554455445545e-07,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8633334040641785,
      "num_tokens": 41550909.0,
      "step": 1200
    },
    {
      "epoch": 0.22302692664809656,
      "grad_norm": 1.5946414543923388,
      "learning_rate": 7.425742574257426e-07,
      "loss": 0.455,
      "mean_token_accuracy": 0.8526860475540161,
      "num_tokens": 41583226.0,
      "step": 1201
    },
    {
      "epoch": 0.22321262766945218,
      "grad_norm": 1.5270918442165435,
      "learning_rate": 7.431930693069307e-07,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8546353578567505,
      "num_tokens": 41618014.0,
      "step": 1202
    },
    {
      "epoch": 0.2233983286908078,
      "grad_norm": 1.6718403831375643,
      "learning_rate": 7.438118811881188e-07,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8393911123275757,
      "num_tokens": 41649555.0,
      "step": 1203
    },
    {
      "epoch": 0.22358402971216343,
      "grad_norm": 1.577149916952268,
      "learning_rate": 7.44430693069307e-07,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8611260652542114,
      "num_tokens": 41681659.0,
      "step": 1204
    },
    {
      "epoch": 0.22376973073351902,
      "grad_norm": 1.543139624346249,
      "learning_rate": 7.45049504950495e-07,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8448835611343384,
      "num_tokens": 41717143.0,
      "step": 1205
    },
    {
      "epoch": 0.22395543175487465,
      "grad_norm": 1.532910688869514,
      "learning_rate": 7.456683168316832e-07,
      "loss": 0.4478,
      "mean_token_accuracy": 0.854144811630249,
      "num_tokens": 41758070.0,
      "step": 1206
    },
    {
      "epoch": 0.22414113277623027,
      "grad_norm": 1.5424399654775198,
      "learning_rate": 7.462871287128712e-07,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8420751690864563,
      "num_tokens": 41790806.0,
      "step": 1207
    },
    {
      "epoch": 0.2243268337975859,
      "grad_norm": 1.5342562709124732,
      "learning_rate": 7.469059405940594e-07,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8568469285964966,
      "num_tokens": 41823916.0,
      "step": 1208
    },
    {
      "epoch": 0.2245125348189415,
      "grad_norm": 1.7512766402764566,
      "learning_rate": 7.475247524752475e-07,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8423981070518494,
      "num_tokens": 41853960.0,
      "step": 1209
    },
    {
      "epoch": 0.22469823584029713,
      "grad_norm": 1.4377686056541827,
      "learning_rate": 7.481435643564356e-07,
      "loss": 0.4314,
      "mean_token_accuracy": 0.858345627784729,
      "num_tokens": 41893491.0,
      "step": 1210
    },
    {
      "epoch": 0.22488393686165273,
      "grad_norm": 1.569542525697792,
      "learning_rate": 7.487623762376237e-07,
      "loss": 0.4637,
      "mean_token_accuracy": 0.847489595413208,
      "num_tokens": 41930908.0,
      "step": 1211
    },
    {
      "epoch": 0.22506963788300835,
      "grad_norm": 1.5535630483191747,
      "learning_rate": 7.493811881188119e-07,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8611770868301392,
      "num_tokens": 41964882.0,
      "step": 1212
    },
    {
      "epoch": 0.22525533890436397,
      "grad_norm": 1.5492197755215376,
      "learning_rate": 7.5e-07,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8510763645172119,
      "num_tokens": 42003411.0,
      "step": 1213
    },
    {
      "epoch": 0.2254410399257196,
      "grad_norm": 1.5809826498474544,
      "learning_rate": 7.506188118811881e-07,
      "loss": 0.4479,
      "mean_token_accuracy": 0.853614091873169,
      "num_tokens": 42037578.0,
      "step": 1214
    },
    {
      "epoch": 0.22562674094707522,
      "grad_norm": 1.4259551612109902,
      "learning_rate": 7.512376237623762e-07,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8562127351760864,
      "num_tokens": 42078049.0,
      "step": 1215
    },
    {
      "epoch": 0.22581244196843084,
      "grad_norm": 1.5605855831447066,
      "learning_rate": 7.518564356435643e-07,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8609998226165771,
      "num_tokens": 42114050.0,
      "step": 1216
    },
    {
      "epoch": 0.22599814298978643,
      "grad_norm": 1.5431629231892192,
      "learning_rate": 7.524752475247525e-07,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8442206978797913,
      "num_tokens": 42151943.0,
      "step": 1217
    },
    {
      "epoch": 0.22618384401114205,
      "grad_norm": 1.4441496897856048,
      "learning_rate": 7.530940594059405e-07,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8620018362998962,
      "num_tokens": 42193560.0,
      "step": 1218
    },
    {
      "epoch": 0.22636954503249768,
      "grad_norm": 1.6556748293304124,
      "learning_rate": 7.537128712871287e-07,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8455885648727417,
      "num_tokens": 42225206.0,
      "step": 1219
    },
    {
      "epoch": 0.2265552460538533,
      "grad_norm": 1.5691504379361316,
      "learning_rate": 7.543316831683167e-07,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8326832056045532,
      "num_tokens": 42260103.0,
      "step": 1220
    },
    {
      "epoch": 0.22674094707520892,
      "grad_norm": 1.4784936212176634,
      "learning_rate": 7.549504950495049e-07,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8498926162719727,
      "num_tokens": 42296631.0,
      "step": 1221
    },
    {
      "epoch": 0.22692664809656454,
      "grad_norm": 1.563027442643037,
      "learning_rate": 7.55569306930693e-07,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8590352535247803,
      "num_tokens": 42331960.0,
      "step": 1222
    },
    {
      "epoch": 0.22711234911792014,
      "grad_norm": 1.7993831064388706,
      "learning_rate": 7.561881188118811e-07,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8478890061378479,
      "num_tokens": 42359550.0,
      "step": 1223
    },
    {
      "epoch": 0.22729805013927576,
      "grad_norm": 1.6618815038312837,
      "learning_rate": 7.568069306930692e-07,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8452726602554321,
      "num_tokens": 42391489.0,
      "step": 1224
    },
    {
      "epoch": 0.22748375116063138,
      "grad_norm": 1.4941346496280479,
      "learning_rate": 7.574257425742574e-07,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8533148765563965,
      "num_tokens": 42429630.0,
      "step": 1225
    },
    {
      "epoch": 0.227669452181987,
      "grad_norm": 1.346606680780393,
      "learning_rate": 7.580445544554454e-07,
      "loss": 0.4068,
      "mean_token_accuracy": 0.86599200963974,
      "num_tokens": 42469535.0,
      "step": 1226
    },
    {
      "epoch": 0.22785515320334263,
      "grad_norm": 1.5959170303504127,
      "learning_rate": 7.586633663366337e-07,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8657022714614868,
      "num_tokens": 42501400.0,
      "step": 1227
    },
    {
      "epoch": 0.22804085422469825,
      "grad_norm": 1.6053079124617788,
      "learning_rate": 7.592821782178217e-07,
      "loss": 0.4182,
      "mean_token_accuracy": 0.863528847694397,
      "num_tokens": 42529184.0,
      "step": 1228
    },
    {
      "epoch": 0.22822655524605384,
      "grad_norm": 1.4175990851331366,
      "learning_rate": 7.599009900990099e-07,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8506096005439758,
      "num_tokens": 42569213.0,
      "step": 1229
    },
    {
      "epoch": 0.22841225626740946,
      "grad_norm": 1.4583830186904554,
      "learning_rate": 7.60519801980198e-07,
      "loss": 0.449,
      "mean_token_accuracy": 0.8512363433837891,
      "num_tokens": 42608523.0,
      "step": 1230
    },
    {
      "epoch": 0.22859795728876509,
      "grad_norm": 1.5317248233937353,
      "learning_rate": 7.611386138613861e-07,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8488036394119263,
      "num_tokens": 42643097.0,
      "step": 1231
    },
    {
      "epoch": 0.2287836583101207,
      "grad_norm": 1.53771339965381,
      "learning_rate": 7.617574257425742e-07,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8531523942947388,
      "num_tokens": 42676867.0,
      "step": 1232
    },
    {
      "epoch": 0.22896935933147633,
      "grad_norm": 1.6788080748963148,
      "learning_rate": 7.623762376237624e-07,
      "loss": 0.495,
      "mean_token_accuracy": 0.8392907381057739,
      "num_tokens": 42709700.0,
      "step": 1233
    },
    {
      "epoch": 0.22915506035283195,
      "grad_norm": 1.533056858506874,
      "learning_rate": 7.629950495049504e-07,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8599913120269775,
      "num_tokens": 42745471.0,
      "step": 1234
    },
    {
      "epoch": 0.22934076137418755,
      "grad_norm": 1.5241784408773187,
      "learning_rate": 7.636138613861386e-07,
      "loss": 0.4082,
      "mean_token_accuracy": 0.863640546798706,
      "num_tokens": 42783570.0,
      "step": 1235
    },
    {
      "epoch": 0.22952646239554317,
      "grad_norm": 1.483773051341287,
      "learning_rate": 7.642326732673266e-07,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8470462560653687,
      "num_tokens": 42824169.0,
      "step": 1236
    },
    {
      "epoch": 0.2297121634168988,
      "grad_norm": 1.5624534291700032,
      "learning_rate": 7.648514851485148e-07,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8517342805862427,
      "num_tokens": 42859224.0,
      "step": 1237
    },
    {
      "epoch": 0.2298978644382544,
      "grad_norm": 1.566088155653879,
      "learning_rate": 7.654702970297029e-07,
      "loss": 0.4494,
      "mean_token_accuracy": 0.854576051235199,
      "num_tokens": 42895553.0,
      "step": 1238
    },
    {
      "epoch": 0.23008356545961003,
      "grad_norm": 1.6576769721734192,
      "learning_rate": 7.66089108910891e-07,
      "loss": 0.422,
      "mean_token_accuracy": 0.8635209202766418,
      "num_tokens": 42924134.0,
      "step": 1239
    },
    {
      "epoch": 0.23026926648096566,
      "grad_norm": 1.5872172875766124,
      "learning_rate": 7.667079207920792e-07,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8547161817550659,
      "num_tokens": 42954295.0,
      "step": 1240
    },
    {
      "epoch": 0.23045496750232125,
      "grad_norm": 1.4123736801001074,
      "learning_rate": 7.673267326732673e-07,
      "loss": 0.4119,
      "mean_token_accuracy": 0.862585186958313,
      "num_tokens": 42994797.0,
      "step": 1241
    },
    {
      "epoch": 0.23064066852367687,
      "grad_norm": 1.6119791444104352,
      "learning_rate": 7.679455445544554e-07,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8270115852355957,
      "num_tokens": 43031240.0,
      "step": 1242
    },
    {
      "epoch": 0.2308263695450325,
      "grad_norm": 1.6344064898924477,
      "learning_rate": 7.685643564356436e-07,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8471900224685669,
      "num_tokens": 43066055.0,
      "step": 1243
    },
    {
      "epoch": 0.23101207056638812,
      "grad_norm": 1.405452742437848,
      "learning_rate": 7.691831683168316e-07,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8630378246307373,
      "num_tokens": 43106384.0,
      "step": 1244
    },
    {
      "epoch": 0.23119777158774374,
      "grad_norm": 1.453984374614517,
      "learning_rate": 7.698019801980198e-07,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8663699626922607,
      "num_tokens": 43138844.0,
      "step": 1245
    },
    {
      "epoch": 0.23138347260909936,
      "grad_norm": 1.4735717408929965,
      "learning_rate": 7.704207920792079e-07,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8623356819152832,
      "num_tokens": 43177818.0,
      "step": 1246
    },
    {
      "epoch": 0.23156917363045496,
      "grad_norm": 1.5450610693766265,
      "learning_rate": 7.71039603960396e-07,
      "loss": 0.4895,
      "mean_token_accuracy": 0.843486487865448,
      "num_tokens": 43221983.0,
      "step": 1247
    },
    {
      "epoch": 0.23175487465181058,
      "grad_norm": 1.6368613602241588,
      "learning_rate": 7.716584158415841e-07,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8505792617797852,
      "num_tokens": 43255048.0,
      "step": 1248
    },
    {
      "epoch": 0.2319405756731662,
      "grad_norm": 1.4653607832108433,
      "learning_rate": 7.722772277227722e-07,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8486194610595703,
      "num_tokens": 43298405.0,
      "step": 1249
    },
    {
      "epoch": 0.23212627669452182,
      "grad_norm": 1.5332728574096812,
      "learning_rate": 7.728960396039603e-07,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8515115976333618,
      "num_tokens": 43333496.0,
      "step": 1250
    },
    {
      "epoch": 0.23231197771587744,
      "grad_norm": 1.4466749181768108,
      "learning_rate": 7.735148514851485e-07,
      "loss": 0.4863,
      "mean_token_accuracy": 0.843328595161438,
      "num_tokens": 43372619.0,
      "step": 1251
    },
    {
      "epoch": 0.23249767873723307,
      "grad_norm": 1.5653866454252785,
      "learning_rate": 7.741336633663365e-07,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8438688516616821,
      "num_tokens": 43410723.0,
      "step": 1252
    },
    {
      "epoch": 0.23268337975858866,
      "grad_norm": 1.6158522988051691,
      "learning_rate": 7.747524752475248e-07,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8421448469161987,
      "num_tokens": 43445902.0,
      "step": 1253
    },
    {
      "epoch": 0.23286908077994428,
      "grad_norm": 1.5773511294904763,
      "learning_rate": 7.753712871287128e-07,
      "loss": 0.434,
      "mean_token_accuracy": 0.8590990304946899,
      "num_tokens": 43481428.0,
      "step": 1254
    },
    {
      "epoch": 0.2330547818012999,
      "grad_norm": 1.5269122079151762,
      "learning_rate": 7.75990099009901e-07,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8604869246482849,
      "num_tokens": 43512955.0,
      "step": 1255
    },
    {
      "epoch": 0.23324048282265553,
      "grad_norm": 1.698125611002587,
      "learning_rate": 7.766089108910891e-07,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8418655395507812,
      "num_tokens": 43548225.0,
      "step": 1256
    },
    {
      "epoch": 0.23342618384401115,
      "grad_norm": 1.5366281444442942,
      "learning_rate": 7.772277227722772e-07,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8520615696907043,
      "num_tokens": 43583413.0,
      "step": 1257
    },
    {
      "epoch": 0.23361188486536677,
      "grad_norm": 1.4989042853147054,
      "learning_rate": 7.778465346534653e-07,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8571935296058655,
      "num_tokens": 43622426.0,
      "step": 1258
    },
    {
      "epoch": 0.23379758588672236,
      "grad_norm": 1.437802638033055,
      "learning_rate": 7.784653465346535e-07,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8579064011573792,
      "num_tokens": 43660541.0,
      "step": 1259
    },
    {
      "epoch": 0.233983286908078,
      "grad_norm": 1.5156633145605793,
      "learning_rate": 7.790841584158415e-07,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8281809091567993,
      "num_tokens": 43697034.0,
      "step": 1260
    },
    {
      "epoch": 0.2341689879294336,
      "grad_norm": 1.5789819206676567,
      "learning_rate": 7.797029702970297e-07,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8509747982025146,
      "num_tokens": 43731145.0,
      "step": 1261
    },
    {
      "epoch": 0.23435468895078923,
      "grad_norm": 1.498962318376808,
      "learning_rate": 7.803217821782177e-07,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8635908365249634,
      "num_tokens": 43766455.0,
      "step": 1262
    },
    {
      "epoch": 0.23454038997214485,
      "grad_norm": 1.5429146677544205,
      "learning_rate": 7.809405940594059e-07,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8287333846092224,
      "num_tokens": 43807353.0,
      "step": 1263
    },
    {
      "epoch": 0.23472609099350047,
      "grad_norm": 1.688650476653173,
      "learning_rate": 7.81559405940594e-07,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8531570434570312,
      "num_tokens": 43836735.0,
      "step": 1264
    },
    {
      "epoch": 0.23491179201485607,
      "grad_norm": 1.5956950960701652,
      "learning_rate": 7.821782178217821e-07,
      "loss": 0.4893,
      "mean_token_accuracy": 0.838935136795044,
      "num_tokens": 43871998.0,
      "step": 1265
    },
    {
      "epoch": 0.2350974930362117,
      "grad_norm": 1.4703973694950534,
      "learning_rate": 7.827970297029702e-07,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8664020895957947,
      "num_tokens": 43909232.0,
      "step": 1266
    },
    {
      "epoch": 0.2352831940575673,
      "grad_norm": 1.5629240785722838,
      "learning_rate": 7.834158415841585e-07,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8455982208251953,
      "num_tokens": 43944333.0,
      "step": 1267
    },
    {
      "epoch": 0.23546889507892294,
      "grad_norm": 1.6835545983232436,
      "learning_rate": 7.840346534653465e-07,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8660180568695068,
      "num_tokens": 43972916.0,
      "step": 1268
    },
    {
      "epoch": 0.23565459610027856,
      "grad_norm": 1.5682718085640308,
      "learning_rate": 7.846534653465347e-07,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8555411100387573,
      "num_tokens": 44008669.0,
      "step": 1269
    },
    {
      "epoch": 0.23584029712163418,
      "grad_norm": 1.566769097479499,
      "learning_rate": 7.852722772277227e-07,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8444886207580566,
      "num_tokens": 44045108.0,
      "step": 1270
    },
    {
      "epoch": 0.23602599814298977,
      "grad_norm": 1.6973450790264661,
      "learning_rate": 7.858910891089109e-07,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8695312142372131,
      "num_tokens": 44071794.0,
      "step": 1271
    },
    {
      "epoch": 0.2362116991643454,
      "grad_norm": 1.6684297811559172,
      "learning_rate": 7.86509900990099e-07,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8463692665100098,
      "num_tokens": 44101767.0,
      "step": 1272
    },
    {
      "epoch": 0.23639740018570102,
      "grad_norm": 1.4039702348431784,
      "learning_rate": 7.871287128712871e-07,
      "loss": 0.397,
      "mean_token_accuracy": 0.866526186466217,
      "num_tokens": 44140283.0,
      "step": 1273
    },
    {
      "epoch": 0.23658310120705664,
      "grad_norm": 1.5707944785330221,
      "learning_rate": 7.877475247524752e-07,
      "loss": 0.4284,
      "mean_token_accuracy": 0.856094241142273,
      "num_tokens": 44174007.0,
      "step": 1274
    },
    {
      "epoch": 0.23676880222841226,
      "grad_norm": 1.764813394291732,
      "learning_rate": 7.883663366336633e-07,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8598737716674805,
      "num_tokens": 44201967.0,
      "step": 1275
    },
    {
      "epoch": 0.23695450324976788,
      "grad_norm": 1.6398834137681435,
      "learning_rate": 7.889851485148514e-07,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8402451872825623,
      "num_tokens": 44238222.0,
      "step": 1276
    },
    {
      "epoch": 0.23714020427112348,
      "grad_norm": 1.417958519901061,
      "learning_rate": 7.896039603960396e-07,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8665996193885803,
      "num_tokens": 44275463.0,
      "step": 1277
    },
    {
      "epoch": 0.2373259052924791,
      "grad_norm": 1.5271581260477516,
      "learning_rate": 7.902227722772276e-07,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8696227073669434,
      "num_tokens": 44303731.0,
      "step": 1278
    },
    {
      "epoch": 0.23751160631383472,
      "grad_norm": 1.584488945252274,
      "learning_rate": 7.908415841584158e-07,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8606604933738708,
      "num_tokens": 44336048.0,
      "step": 1279
    },
    {
      "epoch": 0.23769730733519034,
      "grad_norm": 1.4541577990059644,
      "learning_rate": 7.91460396039604e-07,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8438353538513184,
      "num_tokens": 44379073.0,
      "step": 1280
    },
    {
      "epoch": 0.23788300835654597,
      "grad_norm": 1.6125484573541358,
      "learning_rate": 7.920792079207921e-07,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8414841294288635,
      "num_tokens": 44412328.0,
      "step": 1281
    },
    {
      "epoch": 0.2380687093779016,
      "grad_norm": 1.4701745715940746,
      "learning_rate": 7.926980198019802e-07,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8549531102180481,
      "num_tokens": 44451007.0,
      "step": 1282
    },
    {
      "epoch": 0.23825441039925718,
      "grad_norm": 1.4665134510393554,
      "learning_rate": 7.933168316831683e-07,
      "loss": 0.5,
      "mean_token_accuracy": 0.8362718224525452,
      "num_tokens": 44490849.0,
      "step": 1283
    },
    {
      "epoch": 0.2384401114206128,
      "grad_norm": 1.553343845655634,
      "learning_rate": 7.939356435643564e-07,
      "loss": 0.4535,
      "mean_token_accuracy": 0.850310206413269,
      "num_tokens": 44524817.0,
      "step": 1284
    },
    {
      "epoch": 0.23862581244196843,
      "grad_norm": 1.641377214692073,
      "learning_rate": 7.945544554455446e-07,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8452093601226807,
      "num_tokens": 44558578.0,
      "step": 1285
    },
    {
      "epoch": 0.23881151346332405,
      "grad_norm": 1.6027000523216233,
      "learning_rate": 7.951732673267326e-07,
      "loss": 0.431,
      "mean_token_accuracy": 0.8605759739875793,
      "num_tokens": 44590829.0,
      "step": 1286
    },
    {
      "epoch": 0.23899721448467967,
      "grad_norm": 1.68189232994286,
      "learning_rate": 7.957920792079208e-07,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8543205261230469,
      "num_tokens": 44620119.0,
      "step": 1287
    },
    {
      "epoch": 0.2391829155060353,
      "grad_norm": 1.538136745360778,
      "learning_rate": 7.964108910891088e-07,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8480998873710632,
      "num_tokens": 44658633.0,
      "step": 1288
    },
    {
      "epoch": 0.2393686165273909,
      "grad_norm": 1.5216608139009382,
      "learning_rate": 7.97029702970297e-07,
      "loss": 0.4545,
      "mean_token_accuracy": 0.850028395652771,
      "num_tokens": 44693843.0,
      "step": 1289
    },
    {
      "epoch": 0.2395543175487465,
      "grad_norm": 1.6311135881201326,
      "learning_rate": 7.976485148514851e-07,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8500052690505981,
      "num_tokens": 44728483.0,
      "step": 1290
    },
    {
      "epoch": 0.23974001857010213,
      "grad_norm": 1.4697255351158567,
      "learning_rate": 7.982673267326732e-07,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8745027780532837,
      "num_tokens": 44762582.0,
      "step": 1291
    },
    {
      "epoch": 0.23992571959145775,
      "grad_norm": 1.4907080429046335,
      "learning_rate": 7.988861386138613e-07,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8755248785018921,
      "num_tokens": 44791834.0,
      "step": 1292
    },
    {
      "epoch": 0.24011142061281338,
      "grad_norm": 1.7825428061539803,
      "learning_rate": 7.995049504950496e-07,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8527862429618835,
      "num_tokens": 44818993.0,
      "step": 1293
    },
    {
      "epoch": 0.240297121634169,
      "grad_norm": 1.6919404196584331,
      "learning_rate": 8.001237623762376e-07,
      "loss": 0.4732,
      "mean_token_accuracy": 0.845489501953125,
      "num_tokens": 44851562.0,
      "step": 1294
    },
    {
      "epoch": 0.2404828226555246,
      "grad_norm": 1.4738776817813999,
      "learning_rate": 8.007425742574258e-07,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8539695143699646,
      "num_tokens": 44886784.0,
      "step": 1295
    },
    {
      "epoch": 0.2406685236768802,
      "grad_norm": 1.6107133585967546,
      "learning_rate": 8.013613861386138e-07,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8482168912887573,
      "num_tokens": 44918529.0,
      "step": 1296
    },
    {
      "epoch": 0.24085422469823584,
      "grad_norm": 1.5328589970789743,
      "learning_rate": 8.01980198019802e-07,
      "loss": 0.4282,
      "mean_token_accuracy": 0.857169508934021,
      "num_tokens": 44954306.0,
      "step": 1297
    },
    {
      "epoch": 0.24103992571959146,
      "grad_norm": 1.6183165670326687,
      "learning_rate": 8.025990099009901e-07,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8538332581520081,
      "num_tokens": 44984312.0,
      "step": 1298
    },
    {
      "epoch": 0.24122562674094708,
      "grad_norm": 1.629659314665423,
      "learning_rate": 8.032178217821782e-07,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8623308539390564,
      "num_tokens": 45015366.0,
      "step": 1299
    },
    {
      "epoch": 0.2414113277623027,
      "grad_norm": 1.61788912829794,
      "learning_rate": 8.038366336633663e-07,
      "loss": 0.45,
      "mean_token_accuracy": 0.8537596464157104,
      "num_tokens": 45046542.0,
      "step": 1300
    },
    {
      "epoch": 0.2415970287836583,
      "grad_norm": 1.7039439579475293,
      "learning_rate": 8.044554455445544e-07,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8606298565864563,
      "num_tokens": 45074610.0,
      "step": 1301
    },
    {
      "epoch": 0.24178272980501392,
      "grad_norm": 1.6162567832478962,
      "learning_rate": 8.050742574257425e-07,
      "loss": 0.5009,
      "mean_token_accuracy": 0.837874174118042,
      "num_tokens": 45106923.0,
      "step": 1302
    },
    {
      "epoch": 0.24196843082636954,
      "grad_norm": 1.4940348885891606,
      "learning_rate": 8.056930693069307e-07,
      "loss": 0.4169,
      "mean_token_accuracy": 0.861068069934845,
      "num_tokens": 45141150.0,
      "step": 1303
    },
    {
      "epoch": 0.24215413184772516,
      "grad_norm": 1.6268116073454675,
      "learning_rate": 8.063118811881187e-07,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8474249839782715,
      "num_tokens": 45174107.0,
      "step": 1304
    },
    {
      "epoch": 0.24233983286908078,
      "grad_norm": 1.6487437215783354,
      "learning_rate": 8.069306930693069e-07,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8555957078933716,
      "num_tokens": 45203660.0,
      "step": 1305
    },
    {
      "epoch": 0.2425255338904364,
      "grad_norm": 1.6418339232810182,
      "learning_rate": 8.075495049504951e-07,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8362812995910645,
      "num_tokens": 45240077.0,
      "step": 1306
    },
    {
      "epoch": 0.24271123491179203,
      "grad_norm": 1.4376341353183684,
      "learning_rate": 8.081683168316832e-07,
      "loss": 0.451,
      "mean_token_accuracy": 0.8531612753868103,
      "num_tokens": 45276960.0,
      "step": 1307
    },
    {
      "epoch": 0.24289693593314762,
      "grad_norm": 1.519203555535339,
      "learning_rate": 8.087871287128713e-07,
      "loss": 0.391,
      "mean_token_accuracy": 0.8717570304870605,
      "num_tokens": 45310032.0,
      "step": 1308
    },
    {
      "epoch": 0.24308263695450325,
      "grad_norm": 1.6958776931579171,
      "learning_rate": 8.094059405940594e-07,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8491115570068359,
      "num_tokens": 45338328.0,
      "step": 1309
    },
    {
      "epoch": 0.24326833797585887,
      "grad_norm": 1.7225660960535887,
      "learning_rate": 8.100247524752475e-07,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8760210275650024,
      "num_tokens": 45364237.0,
      "step": 1310
    },
    {
      "epoch": 0.2434540389972145,
      "grad_norm": 1.4649387755369785,
      "learning_rate": 8.106435643564357e-07,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8609246015548706,
      "num_tokens": 45399315.0,
      "step": 1311
    },
    {
      "epoch": 0.2436397400185701,
      "grad_norm": 1.2715196580708745,
      "learning_rate": 8.112623762376237e-07,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8677060604095459,
      "num_tokens": 45443863.0,
      "step": 1312
    },
    {
      "epoch": 0.24382544103992573,
      "grad_norm": 1.607945333853589,
      "learning_rate": 8.118811881188119e-07,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8556628823280334,
      "num_tokens": 45476416.0,
      "step": 1313
    },
    {
      "epoch": 0.24401114206128133,
      "grad_norm": 1.5178562863704474,
      "learning_rate": 8.125e-07,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8553234338760376,
      "num_tokens": 45510529.0,
      "step": 1314
    },
    {
      "epoch": 0.24419684308263695,
      "grad_norm": 1.7242104973924295,
      "learning_rate": 8.13118811881188e-07,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8571357131004333,
      "num_tokens": 45538576.0,
      "step": 1315
    },
    {
      "epoch": 0.24438254410399257,
      "grad_norm": 1.5102496249376471,
      "learning_rate": 8.137376237623762e-07,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8599494695663452,
      "num_tokens": 45571118.0,
      "step": 1316
    },
    {
      "epoch": 0.2445682451253482,
      "grad_norm": 1.5207884535777216,
      "learning_rate": 8.143564356435642e-07,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8526326417922974,
      "num_tokens": 45605638.0,
      "step": 1317
    },
    {
      "epoch": 0.24475394614670382,
      "grad_norm": 1.460758807231145,
      "learning_rate": 8.149752475247524e-07,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8591092824935913,
      "num_tokens": 45643022.0,
      "step": 1318
    },
    {
      "epoch": 0.24493964716805944,
      "grad_norm": 1.4441925708043553,
      "learning_rate": 8.155940594059405e-07,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8598016500473022,
      "num_tokens": 45682844.0,
      "step": 1319
    },
    {
      "epoch": 0.24512534818941503,
      "grad_norm": 1.4441866105113847,
      "learning_rate": 8.162128712871287e-07,
      "loss": 0.4191,
      "mean_token_accuracy": 0.858770489692688,
      "num_tokens": 45720203.0,
      "step": 1320
    },
    {
      "epoch": 0.24531104921077065,
      "grad_norm": 1.577314003976968,
      "learning_rate": 8.168316831683168e-07,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8523455262184143,
      "num_tokens": 45751541.0,
      "step": 1321
    },
    {
      "epoch": 0.24549675023212628,
      "grad_norm": 1.5382248685784223,
      "learning_rate": 8.174504950495049e-07,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8465874195098877,
      "num_tokens": 45789122.0,
      "step": 1322
    },
    {
      "epoch": 0.2456824512534819,
      "grad_norm": 1.7246008423496053,
      "learning_rate": 8.18069306930693e-07,
      "loss": 0.439,
      "mean_token_accuracy": 0.8535223007202148,
      "num_tokens": 45816529.0,
      "step": 1323
    },
    {
      "epoch": 0.24586815227483752,
      "grad_norm": 1.691440652003835,
      "learning_rate": 8.186881188118812e-07,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8402917385101318,
      "num_tokens": 45845145.0,
      "step": 1324
    },
    {
      "epoch": 0.24605385329619314,
      "grad_norm": 1.580766266040303,
      "learning_rate": 8.193069306930692e-07,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8570464253425598,
      "num_tokens": 45877666.0,
      "step": 1325
    },
    {
      "epoch": 0.24623955431754874,
      "grad_norm": 1.4964496837244978,
      "learning_rate": 8.199257425742574e-07,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8543252944946289,
      "num_tokens": 45913085.0,
      "step": 1326
    },
    {
      "epoch": 0.24642525533890436,
      "grad_norm": 1.5982471967877874,
      "learning_rate": 8.205445544554455e-07,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8385865688323975,
      "num_tokens": 45947885.0,
      "step": 1327
    },
    {
      "epoch": 0.24661095636025998,
      "grad_norm": 1.4413616184135913,
      "learning_rate": 8.211633663366336e-07,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8411885499954224,
      "num_tokens": 45991004.0,
      "step": 1328
    },
    {
      "epoch": 0.2467966573816156,
      "grad_norm": 1.369580575584786,
      "learning_rate": 8.217821782178217e-07,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8628503680229187,
      "num_tokens": 46029006.0,
      "step": 1329
    },
    {
      "epoch": 0.24698235840297122,
      "grad_norm": 1.7053150202638538,
      "learning_rate": 8.224009900990098e-07,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8420271873474121,
      "num_tokens": 46060092.0,
      "step": 1330
    },
    {
      "epoch": 0.24716805942432685,
      "grad_norm": 1.459822229995166,
      "learning_rate": 8.230198019801979e-07,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8622419834136963,
      "num_tokens": 46097564.0,
      "step": 1331
    },
    {
      "epoch": 0.24735376044568244,
      "grad_norm": 1.4185980382523184,
      "learning_rate": 8.236386138613861e-07,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8697372674942017,
      "num_tokens": 46133117.0,
      "step": 1332
    },
    {
      "epoch": 0.24753946146703806,
      "grad_norm": 1.5672393696346685,
      "learning_rate": 8.242574257425742e-07,
      "loss": 0.442,
      "mean_token_accuracy": 0.8552073836326599,
      "num_tokens": 46167801.0,
      "step": 1333
    },
    {
      "epoch": 0.24772516248839369,
      "grad_norm": 1.5770373944072245,
      "learning_rate": 8.248762376237624e-07,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8543832898139954,
      "num_tokens": 46204011.0,
      "step": 1334
    },
    {
      "epoch": 0.2479108635097493,
      "grad_norm": 1.5710299333214535,
      "learning_rate": 8.254950495049504e-07,
      "loss": 0.462,
      "mean_token_accuracy": 0.8466540575027466,
      "num_tokens": 46237082.0,
      "step": 1335
    },
    {
      "epoch": 0.24809656453110493,
      "grad_norm": 1.5113428948284071,
      "learning_rate": 8.261138613861386e-07,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8584104180335999,
      "num_tokens": 46271426.0,
      "step": 1336
    },
    {
      "epoch": 0.24828226555246055,
      "grad_norm": 1.5282103569985914,
      "learning_rate": 8.267326732673267e-07,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8645318150520325,
      "num_tokens": 46303884.0,
      "step": 1337
    },
    {
      "epoch": 0.24846796657381615,
      "grad_norm": 1.462448755768575,
      "learning_rate": 8.273514851485148e-07,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8635975122451782,
      "num_tokens": 46341436.0,
      "step": 1338
    },
    {
      "epoch": 0.24865366759517177,
      "grad_norm": 1.6555187166560605,
      "learning_rate": 8.279702970297029e-07,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8471755385398865,
      "num_tokens": 46375581.0,
      "step": 1339
    },
    {
      "epoch": 0.2488393686165274,
      "grad_norm": 1.5290287480185802,
      "learning_rate": 8.285891089108911e-07,
      "loss": 0.3847,
      "mean_token_accuracy": 0.874136745929718,
      "num_tokens": 46406032.0,
      "step": 1340
    },
    {
      "epoch": 0.249025069637883,
      "grad_norm": 1.589037448627325,
      "learning_rate": 8.292079207920791e-07,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8487269878387451,
      "num_tokens": 46438964.0,
      "step": 1341
    },
    {
      "epoch": 0.24921077065923863,
      "grad_norm": 1.583066128834297,
      "learning_rate": 8.298267326732673e-07,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8518291115760803,
      "num_tokens": 46469674.0,
      "step": 1342
    },
    {
      "epoch": 0.24939647168059426,
      "grad_norm": 1.4592694298821824,
      "learning_rate": 8.304455445544553e-07,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8493900299072266,
      "num_tokens": 46510400.0,
      "step": 1343
    },
    {
      "epoch": 0.24958217270194985,
      "grad_norm": 1.4146930284118213,
      "learning_rate": 8.310643564356435e-07,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8437148332595825,
      "num_tokens": 46550519.0,
      "step": 1344
    },
    {
      "epoch": 0.24976787372330547,
      "grad_norm": 1.5619538190520041,
      "learning_rate": 8.316831683168316e-07,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8583122491836548,
      "num_tokens": 46586504.0,
      "step": 1345
    },
    {
      "epoch": 0.2499535747446611,
      "grad_norm": 1.5609451611098764,
      "learning_rate": 8.323019801980198e-07,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8684574365615845,
      "num_tokens": 46616359.0,
      "step": 1346
    },
    {
      "epoch": 0.2501392757660167,
      "grad_norm": 1.5872757635452115,
      "learning_rate": 8.329207920792079e-07,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8438045382499695,
      "num_tokens": 46649600.0,
      "step": 1347
    },
    {
      "epoch": 0.2503249767873723,
      "grad_norm": 1.5774113692030653,
      "learning_rate": 8.33539603960396e-07,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8508197665214539,
      "num_tokens": 46679757.0,
      "step": 1348
    },
    {
      "epoch": 0.25051067780872793,
      "grad_norm": 1.4074691305298013,
      "learning_rate": 8.341584158415841e-07,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8522273302078247,
      "num_tokens": 46718433.0,
      "step": 1349
    },
    {
      "epoch": 0.25069637883008355,
      "grad_norm": 1.4952166584056656,
      "learning_rate": 8.347772277227723e-07,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8539680242538452,
      "num_tokens": 46755552.0,
      "step": 1350
    },
    {
      "epoch": 0.2508820798514392,
      "grad_norm": 1.5364560988403335,
      "learning_rate": 8.353960396039603e-07,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8554613590240479,
      "num_tokens": 46788608.0,
      "step": 1351
    },
    {
      "epoch": 0.2510677808727948,
      "grad_norm": 1.598428234168956,
      "learning_rate": 8.360148514851485e-07,
      "loss": 0.5054,
      "mean_token_accuracy": 0.839907705783844,
      "num_tokens": 46826374.0,
      "step": 1352
    },
    {
      "epoch": 0.2512534818941504,
      "grad_norm": 1.5473079939633219,
      "learning_rate": 8.366336633663366e-07,
      "loss": 0.409,
      "mean_token_accuracy": 0.8659999370574951,
      "num_tokens": 46861433.0,
      "step": 1353
    },
    {
      "epoch": 0.25143918291550604,
      "grad_norm": 1.9655601001505363,
      "learning_rate": 8.372524752475247e-07,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8643558025360107,
      "num_tokens": 46884218.0,
      "step": 1354
    },
    {
      "epoch": 0.25162488393686167,
      "grad_norm": 1.559111766563012,
      "learning_rate": 8.378712871287128e-07,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8484867215156555,
      "num_tokens": 46919472.0,
      "step": 1355
    },
    {
      "epoch": 0.2518105849582173,
      "grad_norm": 1.8804802206275104,
      "learning_rate": 8.384900990099009e-07,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8429271578788757,
      "num_tokens": 46946743.0,
      "step": 1356
    },
    {
      "epoch": 0.2519962859795729,
      "grad_norm": 1.598278892950347,
      "learning_rate": 8.39108910891089e-07,
      "loss": 0.5011,
      "mean_token_accuracy": 0.838919997215271,
      "num_tokens": 46980891.0,
      "step": 1357
    },
    {
      "epoch": 0.25218198700092853,
      "grad_norm": 1.5345304380132787,
      "learning_rate": 8.397277227722772e-07,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8422071933746338,
      "num_tokens": 47017708.0,
      "step": 1358
    },
    {
      "epoch": 0.2523676880222841,
      "grad_norm": 2.014570587247415,
      "learning_rate": 8.403465346534653e-07,
      "loss": 0.394,
      "mean_token_accuracy": 0.8704643249511719,
      "num_tokens": 47042493.0,
      "step": 1359
    },
    {
      "epoch": 0.2525533890436397,
      "grad_norm": 1.8429898054755303,
      "learning_rate": 8.409653465346535e-07,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8419761061668396,
      "num_tokens": 47076384.0,
      "step": 1360
    },
    {
      "epoch": 0.25273909006499534,
      "grad_norm": 1.6174926503134373,
      "learning_rate": 8.415841584158416e-07,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8478950262069702,
      "num_tokens": 47108596.0,
      "step": 1361
    },
    {
      "epoch": 0.25292479108635096,
      "grad_norm": 1.6937976471571004,
      "learning_rate": 8.422029702970297e-07,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8404625654220581,
      "num_tokens": 47136911.0,
      "step": 1362
    },
    {
      "epoch": 0.2531104921077066,
      "grad_norm": 1.6032798557510415,
      "learning_rate": 8.428217821782178e-07,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8583331108093262,
      "num_tokens": 47166342.0,
      "step": 1363
    },
    {
      "epoch": 0.2532961931290622,
      "grad_norm": 1.5564388106058795,
      "learning_rate": 8.434405940594059e-07,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8589590787887573,
      "num_tokens": 47203878.0,
      "step": 1364
    },
    {
      "epoch": 0.25348189415041783,
      "grad_norm": 1.5240514772491183,
      "learning_rate": 8.44059405940594e-07,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8648791909217834,
      "num_tokens": 47237231.0,
      "step": 1365
    },
    {
      "epoch": 0.25366759517177345,
      "grad_norm": 1.5012700033063244,
      "learning_rate": 8.446782178217822e-07,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8572463393211365,
      "num_tokens": 47275287.0,
      "step": 1366
    },
    {
      "epoch": 0.2538532961931291,
      "grad_norm": 1.6725338027226462,
      "learning_rate": 8.452970297029702e-07,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8513641953468323,
      "num_tokens": 47304467.0,
      "step": 1367
    },
    {
      "epoch": 0.2540389972144847,
      "grad_norm": 1.535339784945426,
      "learning_rate": 8.459158415841584e-07,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8344494700431824,
      "num_tokens": 47343092.0,
      "step": 1368
    },
    {
      "epoch": 0.2542246982358403,
      "grad_norm": 1.3971401339973484,
      "learning_rate": 8.465346534653464e-07,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8710381984710693,
      "num_tokens": 47382206.0,
      "step": 1369
    },
    {
      "epoch": 0.25441039925719594,
      "grad_norm": 1.5321414177890673,
      "learning_rate": 8.471534653465346e-07,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8617369532585144,
      "num_tokens": 47414572.0,
      "step": 1370
    },
    {
      "epoch": 0.2545961002785515,
      "grad_norm": 1.4637621426301268,
      "learning_rate": 8.477722772277227e-07,
      "loss": 0.4294,
      "mean_token_accuracy": 0.858750581741333,
      "num_tokens": 47449453.0,
      "step": 1371
    },
    {
      "epoch": 0.25478180129990713,
      "grad_norm": 1.4384204817073063,
      "learning_rate": 8.483910891089109e-07,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8554471135139465,
      "num_tokens": 47491615.0,
      "step": 1372
    },
    {
      "epoch": 0.25496750232126275,
      "grad_norm": 1.6308318724269406,
      "learning_rate": 8.49009900990099e-07,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8728386163711548,
      "num_tokens": 47522911.0,
      "step": 1373
    },
    {
      "epoch": 0.2551532033426184,
      "grad_norm": 1.4666822206669585,
      "learning_rate": 8.496287128712872e-07,
      "loss": 0.4298,
      "mean_token_accuracy": 0.859982967376709,
      "num_tokens": 47560014.0,
      "step": 1374
    },
    {
      "epoch": 0.255338904363974,
      "grad_norm": 1.5657700803962908,
      "learning_rate": 8.502475247524752e-07,
      "loss": 0.415,
      "mean_token_accuracy": 0.86253821849823,
      "num_tokens": 47596739.0,
      "step": 1375
    },
    {
      "epoch": 0.2555246053853296,
      "grad_norm": 1.5734838047540058,
      "learning_rate": 8.508663366336634e-07,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8448677062988281,
      "num_tokens": 47631109.0,
      "step": 1376
    },
    {
      "epoch": 0.25571030640668524,
      "grad_norm": 1.5015090850983117,
      "learning_rate": 8.514851485148514e-07,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8519116640090942,
      "num_tokens": 47673131.0,
      "step": 1377
    },
    {
      "epoch": 0.25589600742804086,
      "grad_norm": 1.4835581227632237,
      "learning_rate": 8.521039603960396e-07,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8492234945297241,
      "num_tokens": 47709863.0,
      "step": 1378
    },
    {
      "epoch": 0.2560817084493965,
      "grad_norm": 1.5440481065945533,
      "learning_rate": 8.527227722772277e-07,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8755685091018677,
      "num_tokens": 47740117.0,
      "step": 1379
    },
    {
      "epoch": 0.2562674094707521,
      "grad_norm": 1.6391862985771681,
      "learning_rate": 8.533415841584158e-07,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8556711077690125,
      "num_tokens": 47769888.0,
      "step": 1380
    },
    {
      "epoch": 0.2564531104921077,
      "grad_norm": 1.680595976593828,
      "learning_rate": 8.539603960396039e-07,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8530405759811401,
      "num_tokens": 47799376.0,
      "step": 1381
    },
    {
      "epoch": 0.25663881151346335,
      "grad_norm": 1.5504056173459166,
      "learning_rate": 8.54579207920792e-07,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8419657349586487,
      "num_tokens": 47839329.0,
      "step": 1382
    },
    {
      "epoch": 0.2568245125348189,
      "grad_norm": 1.452384821308949,
      "learning_rate": 8.551980198019801e-07,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8409804105758667,
      "num_tokens": 47882440.0,
      "step": 1383
    },
    {
      "epoch": 0.25701021355617454,
      "grad_norm": 1.4818141747406086,
      "learning_rate": 8.558168316831683e-07,
      "loss": 0.4516,
      "mean_token_accuracy": 0.857457160949707,
      "num_tokens": 47921501.0,
      "step": 1384
    },
    {
      "epoch": 0.25719591457753016,
      "grad_norm": 1.4689231038929242,
      "learning_rate": 8.564356435643563e-07,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8384074568748474,
      "num_tokens": 47962015.0,
      "step": 1385
    },
    {
      "epoch": 0.2573816155988858,
      "grad_norm": 1.7480420586247085,
      "learning_rate": 8.570544554455446e-07,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8468829393386841,
      "num_tokens": 47993862.0,
      "step": 1386
    },
    {
      "epoch": 0.2575673166202414,
      "grad_norm": 1.6434611912975383,
      "learning_rate": 8.576732673267327e-07,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8590904474258423,
      "num_tokens": 48026815.0,
      "step": 1387
    },
    {
      "epoch": 0.257753017641597,
      "grad_norm": 1.4656920716175965,
      "learning_rate": 8.582920792079208e-07,
      "loss": 0.4422,
      "mean_token_accuracy": 0.857994556427002,
      "num_tokens": 48067511.0,
      "step": 1388
    },
    {
      "epoch": 0.25793871866295265,
      "grad_norm": 1.6935346122488375,
      "learning_rate": 8.589108910891089e-07,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8373141288757324,
      "num_tokens": 48099934.0,
      "step": 1389
    },
    {
      "epoch": 0.25812441968430827,
      "grad_norm": 1.4304034022238759,
      "learning_rate": 8.59529702970297e-07,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8505696058273315,
      "num_tokens": 48138417.0,
      "step": 1390
    },
    {
      "epoch": 0.2583101207056639,
      "grad_norm": 1.8114547126943754,
      "learning_rate": 8.601485148514851e-07,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8429552912712097,
      "num_tokens": 48166131.0,
      "step": 1391
    },
    {
      "epoch": 0.2584958217270195,
      "grad_norm": 1.4431449763372806,
      "learning_rate": 8.607673267326733e-07,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8659839630126953,
      "num_tokens": 48207520.0,
      "step": 1392
    },
    {
      "epoch": 0.25868152274837514,
      "grad_norm": 1.5881834304208087,
      "learning_rate": 8.613861386138613e-07,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8439054489135742,
      "num_tokens": 48240853.0,
      "step": 1393
    },
    {
      "epoch": 0.25886722376973076,
      "grad_norm": 1.5976363914253884,
      "learning_rate": 8.620049504950495e-07,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8428181409835815,
      "num_tokens": 48272298.0,
      "step": 1394
    },
    {
      "epoch": 0.2590529247910863,
      "grad_norm": 1.58132676785745,
      "learning_rate": 8.626237623762375e-07,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8443631529808044,
      "num_tokens": 48304079.0,
      "step": 1395
    },
    {
      "epoch": 0.25923862581244195,
      "grad_norm": 1.5672130023215518,
      "learning_rate": 8.632425742574257e-07,
      "loss": 0.4541,
      "mean_token_accuracy": 0.851902961730957,
      "num_tokens": 48338632.0,
      "step": 1396
    },
    {
      "epoch": 0.25942432683379757,
      "grad_norm": 1.5951809446847525,
      "learning_rate": 8.638613861386138e-07,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8532363176345825,
      "num_tokens": 48372823.0,
      "step": 1397
    },
    {
      "epoch": 0.2596100278551532,
      "grad_norm": 1.5279944539449002,
      "learning_rate": 8.644801980198019e-07,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8698827624320984,
      "num_tokens": 48403494.0,
      "step": 1398
    },
    {
      "epoch": 0.2597957288765088,
      "grad_norm": 1.506262904854444,
      "learning_rate": 8.650990099009901e-07,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8561246991157532,
      "num_tokens": 48440191.0,
      "step": 1399
    },
    {
      "epoch": 0.25998142989786444,
      "grad_norm": 1.5649462958195488,
      "learning_rate": 8.657178217821783e-07,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8620270490646362,
      "num_tokens": 48475770.0,
      "step": 1400
    },
    {
      "epoch": 0.26016713091922006,
      "grad_norm": 1.5995705468290042,
      "learning_rate": 8.663366336633663e-07,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8540681600570679,
      "num_tokens": 48510879.0,
      "step": 1401
    },
    {
      "epoch": 0.2603528319405757,
      "grad_norm": 1.443407254238291,
      "learning_rate": 8.669554455445545e-07,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8504015207290649,
      "num_tokens": 48550301.0,
      "step": 1402
    },
    {
      "epoch": 0.2605385329619313,
      "grad_norm": 1.4675608118512506,
      "learning_rate": 8.675742574257425e-07,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8520214557647705,
      "num_tokens": 48589787.0,
      "step": 1403
    },
    {
      "epoch": 0.2607242339832869,
      "grad_norm": 1.557621596137894,
      "learning_rate": 8.681930693069307e-07,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8528103828430176,
      "num_tokens": 48624052.0,
      "step": 1404
    },
    {
      "epoch": 0.26090993500464255,
      "grad_norm": 1.6178832497493978,
      "learning_rate": 8.688118811881188e-07,
      "loss": 0.4781,
      "mean_token_accuracy": 0.84864342212677,
      "num_tokens": 48657283.0,
      "step": 1405
    },
    {
      "epoch": 0.26109563602599817,
      "grad_norm": 1.5726325443697848,
      "learning_rate": 8.694306930693069e-07,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8473718166351318,
      "num_tokens": 48693047.0,
      "step": 1406
    },
    {
      "epoch": 0.26128133704735373,
      "grad_norm": 1.618567496512048,
      "learning_rate": 8.70049504950495e-07,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8444145917892456,
      "num_tokens": 48728538.0,
      "step": 1407
    },
    {
      "epoch": 0.26146703806870936,
      "grad_norm": 1.439709570239515,
      "learning_rate": 8.706683168316832e-07,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8682746887207031,
      "num_tokens": 48765305.0,
      "step": 1408
    },
    {
      "epoch": 0.261652739090065,
      "grad_norm": 1.3254700562646335,
      "learning_rate": 8.712871287128712e-07,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8503859043121338,
      "num_tokens": 48812740.0,
      "step": 1409
    },
    {
      "epoch": 0.2618384401114206,
      "grad_norm": 1.5698702565977463,
      "learning_rate": 8.719059405940594e-07,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8506802320480347,
      "num_tokens": 48847897.0,
      "step": 1410
    },
    {
      "epoch": 0.2620241411327762,
      "grad_norm": 1.4931958709407542,
      "learning_rate": 8.725247524752474e-07,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8408224582672119,
      "num_tokens": 48884409.0,
      "step": 1411
    },
    {
      "epoch": 0.26220984215413184,
      "grad_norm": 1.4968476824954104,
      "learning_rate": 8.731435643564357e-07,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8435868620872498,
      "num_tokens": 48921389.0,
      "step": 1412
    },
    {
      "epoch": 0.26239554317548747,
      "grad_norm": 1.714584589009384,
      "learning_rate": 8.737623762376238e-07,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8461573719978333,
      "num_tokens": 48949746.0,
      "step": 1413
    },
    {
      "epoch": 0.2625812441968431,
      "grad_norm": 1.4302645404824956,
      "learning_rate": 8.743811881188119e-07,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8691296577453613,
      "num_tokens": 48987374.0,
      "step": 1414
    },
    {
      "epoch": 0.2627669452181987,
      "grad_norm": 1.452925848505187,
      "learning_rate": 8.75e-07,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8462897539138794,
      "num_tokens": 49027214.0,
      "step": 1415
    },
    {
      "epoch": 0.26295264623955433,
      "grad_norm": 1.4273405271730026,
      "learning_rate": 8.75618811881188e-07,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8558064699172974,
      "num_tokens": 49068407.0,
      "step": 1416
    },
    {
      "epoch": 0.26313834726090995,
      "grad_norm": 1.4586937021053408,
      "learning_rate": 8.762376237623762e-07,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8570005297660828,
      "num_tokens": 49106246.0,
      "step": 1417
    },
    {
      "epoch": 0.2633240482822656,
      "grad_norm": 1.4933320776086956,
      "learning_rate": 8.768564356435643e-07,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8537981510162354,
      "num_tokens": 49139031.0,
      "step": 1418
    },
    {
      "epoch": 0.26350974930362114,
      "grad_norm": 1.4807232119383054,
      "learning_rate": 8.774752475247524e-07,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8551313877105713,
      "num_tokens": 49177144.0,
      "step": 1419
    },
    {
      "epoch": 0.26369545032497677,
      "grad_norm": 1.4060991111449879,
      "learning_rate": 8.780940594059405e-07,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8603770732879639,
      "num_tokens": 49214791.0,
      "step": 1420
    },
    {
      "epoch": 0.2638811513463324,
      "grad_norm": 1.4777896735824791,
      "learning_rate": 8.787128712871287e-07,
      "loss": 0.468,
      "mean_token_accuracy": 0.8446487784385681,
      "num_tokens": 49254008.0,
      "step": 1421
    },
    {
      "epoch": 0.264066852367688,
      "grad_norm": 1.403929875499993,
      "learning_rate": 8.793316831683167e-07,
      "loss": 0.4267,
      "mean_token_accuracy": 0.857720673084259,
      "num_tokens": 49296180.0,
      "step": 1422
    },
    {
      "epoch": 0.26425255338904363,
      "grad_norm": 1.563816446163302,
      "learning_rate": 8.799504950495049e-07,
      "loss": 0.417,
      "mean_token_accuracy": 0.8580013513565063,
      "num_tokens": 49329075.0,
      "step": 1423
    },
    {
      "epoch": 0.26443825441039925,
      "grad_norm": 1.4901390492002722,
      "learning_rate": 8.805693069306929e-07,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8365944623947144,
      "num_tokens": 49369505.0,
      "step": 1424
    },
    {
      "epoch": 0.2646239554317549,
      "grad_norm": 1.7036277993332536,
      "learning_rate": 8.811881188118812e-07,
      "loss": 0.51,
      "mean_token_accuracy": 0.8321709036827087,
      "num_tokens": 49400282.0,
      "step": 1425
    },
    {
      "epoch": 0.2648096564531105,
      "grad_norm": 1.4497475010983947,
      "learning_rate": 8.818069306930693e-07,
      "loss": 0.457,
      "mean_token_accuracy": 0.8508737683296204,
      "num_tokens": 49437947.0,
      "step": 1426
    },
    {
      "epoch": 0.2649953574744661,
      "grad_norm": 1.3953295034056075,
      "learning_rate": 8.824257425742574e-07,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8533949255943298,
      "num_tokens": 49478519.0,
      "step": 1427
    },
    {
      "epoch": 0.26518105849582174,
      "grad_norm": 1.6913071555269719,
      "learning_rate": 8.830445544554455e-07,
      "loss": 0.5,
      "mean_token_accuracy": 0.8410594463348389,
      "num_tokens": 49512453.0,
      "step": 1428
    },
    {
      "epoch": 0.26536675951717736,
      "grad_norm": 1.4628931655690454,
      "learning_rate": 8.836633663366337e-07,
      "loss": 0.457,
      "mean_token_accuracy": 0.8504528999328613,
      "num_tokens": 49553777.0,
      "step": 1429
    },
    {
      "epoch": 0.265552460538533,
      "grad_norm": 1.4470852495324196,
      "learning_rate": 8.842821782178217e-07,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8654018044471741,
      "num_tokens": 49588811.0,
      "step": 1430
    },
    {
      "epoch": 0.26573816155988855,
      "grad_norm": 1.5948906905203974,
      "learning_rate": 8.849009900990099e-07,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8523499369621277,
      "num_tokens": 49624928.0,
      "step": 1431
    },
    {
      "epoch": 0.2659238625812442,
      "grad_norm": 1.4333389831991266,
      "learning_rate": 8.855198019801979e-07,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8610829710960388,
      "num_tokens": 49659501.0,
      "step": 1432
    },
    {
      "epoch": 0.2661095636025998,
      "grad_norm": 1.6798500223245127,
      "learning_rate": 8.861386138613861e-07,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8471201062202454,
      "num_tokens": 49687861.0,
      "step": 1433
    },
    {
      "epoch": 0.2662952646239554,
      "grad_norm": 1.5173873618364555,
      "learning_rate": 8.867574257425742e-07,
      "loss": 0.433,
      "mean_token_accuracy": 0.8573607206344604,
      "num_tokens": 49724070.0,
      "step": 1434
    },
    {
      "epoch": 0.26648096564531104,
      "grad_norm": 1.4814446343768362,
      "learning_rate": 8.873762376237623e-07,
      "loss": 0.4973,
      "mean_token_accuracy": 0.8402504920959473,
      "num_tokens": 49762403.0,
      "step": 1435
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 1.3859832326061987,
      "learning_rate": 8.879950495049504e-07,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8660027384757996,
      "num_tokens": 49801430.0,
      "step": 1436
    },
    {
      "epoch": 0.2668523676880223,
      "grad_norm": 1.400242003705028,
      "learning_rate": 8.886138613861385e-07,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8625071048736572,
      "num_tokens": 49842101.0,
      "step": 1437
    },
    {
      "epoch": 0.2670380687093779,
      "grad_norm": 1.4691610380650446,
      "learning_rate": 8.892326732673266e-07,
      "loss": 0.4003,
      "mean_token_accuracy": 0.866743803024292,
      "num_tokens": 49876865.0,
      "step": 1438
    },
    {
      "epoch": 0.26722376973073353,
      "grad_norm": 1.4416761777614484,
      "learning_rate": 8.898514851485149e-07,
      "loss": 0.409,
      "mean_token_accuracy": 0.8635343313217163,
      "num_tokens": 49914416.0,
      "step": 1439
    },
    {
      "epoch": 0.26740947075208915,
      "grad_norm": 1.70808766644718,
      "learning_rate": 8.904702970297029e-07,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8641009330749512,
      "num_tokens": 49943875.0,
      "step": 1440
    },
    {
      "epoch": 0.2675951717734448,
      "grad_norm": 1.4706637523201165,
      "learning_rate": 8.910891089108911e-07,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8759996891021729,
      "num_tokens": 49976114.0,
      "step": 1441
    },
    {
      "epoch": 0.2677808727948004,
      "grad_norm": 1.6089719250704453,
      "learning_rate": 8.917079207920792e-07,
      "loss": 0.45,
      "mean_token_accuracy": 0.8528346419334412,
      "num_tokens": 50006952.0,
      "step": 1442
    },
    {
      "epoch": 0.26796657381615596,
      "grad_norm": 1.4278189130721475,
      "learning_rate": 8.923267326732673e-07,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8615573048591614,
      "num_tokens": 50048677.0,
      "step": 1443
    },
    {
      "epoch": 0.2681522748375116,
      "grad_norm": 1.5737024600407274,
      "learning_rate": 8.929455445544554e-07,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8663566708564758,
      "num_tokens": 50082504.0,
      "step": 1444
    },
    {
      "epoch": 0.2683379758588672,
      "grad_norm": 1.5410774973613799,
      "learning_rate": 8.935643564356435e-07,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8531256914138794,
      "num_tokens": 50125192.0,
      "step": 1445
    },
    {
      "epoch": 0.26852367688022283,
      "grad_norm": 1.7273038699316658,
      "learning_rate": 8.941831683168316e-07,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8654356002807617,
      "num_tokens": 50153075.0,
      "step": 1446
    },
    {
      "epoch": 0.26870937790157845,
      "grad_norm": 1.6352246130122607,
      "learning_rate": 8.948019801980198e-07,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8695622682571411,
      "num_tokens": 50181032.0,
      "step": 1447
    },
    {
      "epoch": 0.26889507892293407,
      "grad_norm": 1.6492172711833097,
      "learning_rate": 8.954207920792078e-07,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8367576599121094,
      "num_tokens": 50213638.0,
      "step": 1448
    },
    {
      "epoch": 0.2690807799442897,
      "grad_norm": 1.5647313205492916,
      "learning_rate": 8.96039603960396e-07,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8497979044914246,
      "num_tokens": 50248742.0,
      "step": 1449
    },
    {
      "epoch": 0.2692664809656453,
      "grad_norm": 1.603317146997351,
      "learning_rate": 8.96658415841584e-07,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8412394523620605,
      "num_tokens": 50284109.0,
      "step": 1450
    },
    {
      "epoch": 0.26945218198700094,
      "grad_norm": 1.5372127396619717,
      "learning_rate": 8.972772277227722e-07,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8565968871116638,
      "num_tokens": 50317671.0,
      "step": 1451
    },
    {
      "epoch": 0.26963788300835656,
      "grad_norm": 1.624033137423352,
      "learning_rate": 8.978960396039604e-07,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8656845092773438,
      "num_tokens": 50349029.0,
      "step": 1452
    },
    {
      "epoch": 0.2698235840297122,
      "grad_norm": 1.5295955219920574,
      "learning_rate": 8.985148514851485e-07,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8529348969459534,
      "num_tokens": 50386006.0,
      "step": 1453
    },
    {
      "epoch": 0.2700092850510678,
      "grad_norm": 1.7310942462817993,
      "learning_rate": 8.991336633663366e-07,
      "loss": 0.4491,
      "mean_token_accuracy": 0.858873188495636,
      "num_tokens": 50415411.0,
      "step": 1454
    },
    {
      "epoch": 0.27019498607242337,
      "grad_norm": 1.5823504607672831,
      "learning_rate": 8.997524752475248e-07,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8393266201019287,
      "num_tokens": 50449807.0,
      "step": 1455
    },
    {
      "epoch": 0.270380687093779,
      "grad_norm": 1.4821302636989135,
      "learning_rate": 9.003712871287128e-07,
      "loss": 0.459,
      "mean_token_accuracy": 0.8499143123626709,
      "num_tokens": 50485909.0,
      "step": 1456
    },
    {
      "epoch": 0.2705663881151346,
      "grad_norm": 1.3995683178358667,
      "learning_rate": 9.00990099009901e-07,
      "loss": 0.4392,
      "mean_token_accuracy": 0.851758599281311,
      "num_tokens": 50525638.0,
      "step": 1457
    },
    {
      "epoch": 0.27075208913649024,
      "grad_norm": 1.7633395048211893,
      "learning_rate": 9.01608910891089e-07,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8460904359817505,
      "num_tokens": 50557216.0,
      "step": 1458
    },
    {
      "epoch": 0.27093779015784586,
      "grad_norm": 1.4571358583929657,
      "learning_rate": 9.022277227722772e-07,
      "loss": 0.465,
      "mean_token_accuracy": 0.8485798835754395,
      "num_tokens": 50594230.0,
      "step": 1459
    },
    {
      "epoch": 0.2711234911792015,
      "grad_norm": 1.425950565483366,
      "learning_rate": 9.028465346534653e-07,
      "loss": 0.4273,
      "mean_token_accuracy": 0.859272301197052,
      "num_tokens": 50634446.0,
      "step": 1460
    },
    {
      "epoch": 0.2713091922005571,
      "grad_norm": 1.519184751897664,
      "learning_rate": 9.034653465346534e-07,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8505730032920837,
      "num_tokens": 50673941.0,
      "step": 1461
    },
    {
      "epoch": 0.2714948932219127,
      "grad_norm": 1.4694121884870024,
      "learning_rate": 9.040841584158415e-07,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8512129783630371,
      "num_tokens": 50709654.0,
      "step": 1462
    },
    {
      "epoch": 0.27168059424326835,
      "grad_norm": 1.512268693040425,
      "learning_rate": 9.047029702970296e-07,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8524814248085022,
      "num_tokens": 50744580.0,
      "step": 1463
    },
    {
      "epoch": 0.27186629526462397,
      "grad_norm": 1.4947757101166106,
      "learning_rate": 9.053217821782177e-07,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8544759750366211,
      "num_tokens": 50781742.0,
      "step": 1464
    },
    {
      "epoch": 0.2720519962859796,
      "grad_norm": 1.537965912587723,
      "learning_rate": 9.05940594059406e-07,
      "loss": 0.396,
      "mean_token_accuracy": 0.8665444254875183,
      "num_tokens": 50814657.0,
      "step": 1465
    },
    {
      "epoch": 0.2722376973073352,
      "grad_norm": 1.6059887896870575,
      "learning_rate": 9.06559405940594e-07,
      "loss": 0.4461,
      "mean_token_accuracy": 0.853954553604126,
      "num_tokens": 50848903.0,
      "step": 1466
    },
    {
      "epoch": 0.2724233983286908,
      "grad_norm": 1.7078967692252545,
      "learning_rate": 9.071782178217822e-07,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8687546849250793,
      "num_tokens": 50878377.0,
      "step": 1467
    },
    {
      "epoch": 0.2726090993500464,
      "grad_norm": 1.8365810586519564,
      "learning_rate": 9.077970297029703e-07,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8481787443161011,
      "num_tokens": 50903676.0,
      "step": 1468
    },
    {
      "epoch": 0.272794800371402,
      "grad_norm": 1.4985481458701044,
      "learning_rate": 9.084158415841584e-07,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8628315925598145,
      "num_tokens": 50937134.0,
      "step": 1469
    },
    {
      "epoch": 0.27298050139275765,
      "grad_norm": 1.4308197263715021,
      "learning_rate": 9.090346534653465e-07,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8761985898017883,
      "num_tokens": 50975482.0,
      "step": 1470
    },
    {
      "epoch": 0.27316620241411327,
      "grad_norm": 1.5126861310736943,
      "learning_rate": 9.096534653465346e-07,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8526997566223145,
      "num_tokens": 51011167.0,
      "step": 1471
    },
    {
      "epoch": 0.2733519034354689,
      "grad_norm": 1.4976107688951554,
      "learning_rate": 9.102722772277227e-07,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8517584800720215,
      "num_tokens": 51050305.0,
      "step": 1472
    },
    {
      "epoch": 0.2735376044568245,
      "grad_norm": 1.537266619145517,
      "learning_rate": 9.108910891089109e-07,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8611933588981628,
      "num_tokens": 51082500.0,
      "step": 1473
    },
    {
      "epoch": 0.27372330547818013,
      "grad_norm": 1.6314938638029297,
      "learning_rate": 9.115099009900989e-07,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8441532254219055,
      "num_tokens": 51118299.0,
      "step": 1474
    },
    {
      "epoch": 0.27390900649953576,
      "grad_norm": 1.470705524878013,
      "learning_rate": 9.121287128712871e-07,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8579714298248291,
      "num_tokens": 51155191.0,
      "step": 1475
    },
    {
      "epoch": 0.2740947075208914,
      "grad_norm": 1.7306260593476275,
      "learning_rate": 9.127475247524751e-07,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8333508968353271,
      "num_tokens": 51185821.0,
      "step": 1476
    },
    {
      "epoch": 0.274280408542247,
      "grad_norm": 1.4125529084787498,
      "learning_rate": 9.133663366336633e-07,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8695699572563171,
      "num_tokens": 51223489.0,
      "step": 1477
    },
    {
      "epoch": 0.2744661095636026,
      "grad_norm": 1.5422074181490364,
      "learning_rate": 9.139851485148514e-07,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8475085496902466,
      "num_tokens": 51260968.0,
      "step": 1478
    },
    {
      "epoch": 0.27465181058495824,
      "grad_norm": 1.560344430710139,
      "learning_rate": 9.146039603960396e-07,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8599528074264526,
      "num_tokens": 51293778.0,
      "step": 1479
    },
    {
      "epoch": 0.2748375116063138,
      "grad_norm": 1.4272516798375297,
      "learning_rate": 9.152227722772277e-07,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8535805940628052,
      "num_tokens": 51332596.0,
      "step": 1480
    },
    {
      "epoch": 0.27502321262766943,
      "grad_norm": 1.587953015267628,
      "learning_rate": 9.158415841584159e-07,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8456383943557739,
      "num_tokens": 51365069.0,
      "step": 1481
    },
    {
      "epoch": 0.27520891364902506,
      "grad_norm": 1.5572213568785034,
      "learning_rate": 9.164603960396039e-07,
      "loss": 0.4187,
      "mean_token_accuracy": 0.864626944065094,
      "num_tokens": 51401890.0,
      "step": 1482
    },
    {
      "epoch": 0.2753946146703807,
      "grad_norm": 1.500922164809947,
      "learning_rate": 9.170792079207921e-07,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8509494066238403,
      "num_tokens": 51437838.0,
      "step": 1483
    },
    {
      "epoch": 0.2755803156917363,
      "grad_norm": 1.8238885865652086,
      "learning_rate": 9.176980198019801e-07,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8406743407249451,
      "num_tokens": 51467501.0,
      "step": 1484
    },
    {
      "epoch": 0.2757660167130919,
      "grad_norm": 1.5578252974106996,
      "learning_rate": 9.183168316831683e-07,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8497030138969421,
      "num_tokens": 51502449.0,
      "step": 1485
    },
    {
      "epoch": 0.27595171773444754,
      "grad_norm": 1.511952859070437,
      "learning_rate": 9.189356435643564e-07,
      "loss": 0.44,
      "mean_token_accuracy": 0.8547980189323425,
      "num_tokens": 51538834.0,
      "step": 1486
    },
    {
      "epoch": 0.27613741875580317,
      "grad_norm": 1.7823834525076956,
      "learning_rate": 9.195544554455445e-07,
      "loss": 0.485,
      "mean_token_accuracy": 0.8377702236175537,
      "num_tokens": 51567847.0,
      "step": 1487
    },
    {
      "epoch": 0.2763231197771588,
      "grad_norm": 1.8760356871972927,
      "learning_rate": 9.201732673267326e-07,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8680088520050049,
      "num_tokens": 51589140.0,
      "step": 1488
    },
    {
      "epoch": 0.2765088207985144,
      "grad_norm": 1.5905333028184654,
      "learning_rate": 9.207920792079208e-07,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8560540676116943,
      "num_tokens": 51623408.0,
      "step": 1489
    },
    {
      "epoch": 0.27669452181987003,
      "grad_norm": 1.5388179816979957,
      "learning_rate": 9.214108910891088e-07,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8540897369384766,
      "num_tokens": 51657721.0,
      "step": 1490
    },
    {
      "epoch": 0.27688022284122565,
      "grad_norm": 1.451936587029742,
      "learning_rate": 9.220297029702971e-07,
      "loss": 0.433,
      "mean_token_accuracy": 0.8567852973937988,
      "num_tokens": 51696238.0,
      "step": 1491
    },
    {
      "epoch": 0.2770659238625812,
      "grad_norm": 1.499918425437426,
      "learning_rate": 9.226485148514851e-07,
      "loss": 0.447,
      "mean_token_accuracy": 0.8532488346099854,
      "num_tokens": 51728278.0,
      "step": 1492
    },
    {
      "epoch": 0.27725162488393684,
      "grad_norm": 1.5528986104624545,
      "learning_rate": 9.232673267326733e-07,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8601458072662354,
      "num_tokens": 51765969.0,
      "step": 1493
    },
    {
      "epoch": 0.27743732590529246,
      "grad_norm": 1.4726316675151014,
      "learning_rate": 9.238861386138614e-07,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8642586469650269,
      "num_tokens": 51800449.0,
      "step": 1494
    },
    {
      "epoch": 0.2776230269266481,
      "grad_norm": 1.674402852505918,
      "learning_rate": 9.245049504950495e-07,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8621277809143066,
      "num_tokens": 51828501.0,
      "step": 1495
    },
    {
      "epoch": 0.2778087279480037,
      "grad_norm": 1.459396243206349,
      "learning_rate": 9.251237623762376e-07,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8628758192062378,
      "num_tokens": 51869444.0,
      "step": 1496
    },
    {
      "epoch": 0.27799442896935933,
      "grad_norm": 1.7843030381599643,
      "learning_rate": 9.257425742574257e-07,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8649873733520508,
      "num_tokens": 51898349.0,
      "step": 1497
    },
    {
      "epoch": 0.27818012999071495,
      "grad_norm": 1.4739131008125985,
      "learning_rate": 9.263613861386138e-07,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8502154350280762,
      "num_tokens": 51935653.0,
      "step": 1498
    },
    {
      "epoch": 0.2783658310120706,
      "grad_norm": 1.6443255106963959,
      "learning_rate": 9.26980198019802e-07,
      "loss": 0.4378,
      "mean_token_accuracy": 0.856210470199585,
      "num_tokens": 51964552.0,
      "step": 1499
    },
    {
      "epoch": 0.2785515320334262,
      "grad_norm": 1.6134907813806283,
      "learning_rate": 9.2759900990099e-07,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8528034687042236,
      "num_tokens": 51997135.0,
      "step": 1500
    },
    {
      "epoch": 0.2787372330547818,
      "grad_norm": 1.440223920875212,
      "learning_rate": 9.282178217821782e-07,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8687004446983337,
      "num_tokens": 52031886.0,
      "step": 1501
    },
    {
      "epoch": 0.27892293407613744,
      "grad_norm": 1.5408704466880176,
      "learning_rate": 9.288366336633663e-07,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8528019189834595,
      "num_tokens": 52069527.0,
      "step": 1502
    },
    {
      "epoch": 0.27910863509749306,
      "grad_norm": 1.507773228061987,
      "learning_rate": 9.294554455445544e-07,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8410671353340149,
      "num_tokens": 52102864.0,
      "step": 1503
    },
    {
      "epoch": 0.27929433611884863,
      "grad_norm": 1.3915218761991055,
      "learning_rate": 9.300742574257425e-07,
      "loss": 0.391,
      "mean_token_accuracy": 0.8710852861404419,
      "num_tokens": 52141696.0,
      "step": 1504
    },
    {
      "epoch": 0.27948003714020425,
      "grad_norm": 1.4298790792229332,
      "learning_rate": 9.306930693069307e-07,
      "loss": 0.4207,
      "mean_token_accuracy": 0.861454963684082,
      "num_tokens": 52183312.0,
      "step": 1505
    },
    {
      "epoch": 0.2796657381615599,
      "grad_norm": 1.5577657618098306,
      "learning_rate": 9.313118811881188e-07,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8520326018333435,
      "num_tokens": 52217721.0,
      "step": 1506
    },
    {
      "epoch": 0.2798514391829155,
      "grad_norm": 1.553278818953818,
      "learning_rate": 9.31930693069307e-07,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8544172048568726,
      "num_tokens": 52251560.0,
      "step": 1507
    },
    {
      "epoch": 0.2800371402042711,
      "grad_norm": 1.6310793851918373,
      "learning_rate": 9.32549504950495e-07,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8521353006362915,
      "num_tokens": 52285554.0,
      "step": 1508
    },
    {
      "epoch": 0.28022284122562674,
      "grad_norm": 1.5708511334194946,
      "learning_rate": 9.331683168316832e-07,
      "loss": 0.4379,
      "mean_token_accuracy": 0.852426290512085,
      "num_tokens": 52319947.0,
      "step": 1509
    },
    {
      "epoch": 0.28040854224698236,
      "grad_norm": 1.5784937848217957,
      "learning_rate": 9.337871287128712e-07,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8548639416694641,
      "num_tokens": 52351952.0,
      "step": 1510
    },
    {
      "epoch": 0.280594243268338,
      "grad_norm": 1.4893698608910182,
      "learning_rate": 9.344059405940594e-07,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8560822010040283,
      "num_tokens": 52387939.0,
      "step": 1511
    },
    {
      "epoch": 0.2807799442896936,
      "grad_norm": 1.4906387164267076,
      "learning_rate": 9.350247524752475e-07,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8607094287872314,
      "num_tokens": 52424105.0,
      "step": 1512
    },
    {
      "epoch": 0.28096564531104923,
      "grad_norm": 1.7029161139243965,
      "learning_rate": 9.356435643564356e-07,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8510751128196716,
      "num_tokens": 52459870.0,
      "step": 1513
    },
    {
      "epoch": 0.28115134633240485,
      "grad_norm": 1.5879296367735498,
      "learning_rate": 9.362623762376237e-07,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8440607190132141,
      "num_tokens": 52494397.0,
      "step": 1514
    },
    {
      "epoch": 0.28133704735376047,
      "grad_norm": 1.5097938479129367,
      "learning_rate": 9.368811881188119e-07,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8507677912712097,
      "num_tokens": 52535115.0,
      "step": 1515
    },
    {
      "epoch": 0.28152274837511604,
      "grad_norm": 1.4692007937085654,
      "learning_rate": 9.374999999999999e-07,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8648730516433716,
      "num_tokens": 52570425.0,
      "step": 1516
    },
    {
      "epoch": 0.28170844939647166,
      "grad_norm": 1.69148877837692,
      "learning_rate": 9.38118811881188e-07,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8590021133422852,
      "num_tokens": 52596386.0,
      "step": 1517
    },
    {
      "epoch": 0.2818941504178273,
      "grad_norm": 1.6146882442688286,
      "learning_rate": 9.387376237623762e-07,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8628541231155396,
      "num_tokens": 52623966.0,
      "step": 1518
    },
    {
      "epoch": 0.2820798514391829,
      "grad_norm": 1.6878415589191702,
      "learning_rate": 9.393564356435643e-07,
      "loss": 0.432,
      "mean_token_accuracy": 0.8537059426307678,
      "num_tokens": 52653179.0,
      "step": 1519
    },
    {
      "epoch": 0.2822655524605385,
      "grad_norm": 1.5507234805854104,
      "learning_rate": 9.399752475247525e-07,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8524718284606934,
      "num_tokens": 52686543.0,
      "step": 1520
    },
    {
      "epoch": 0.28245125348189415,
      "grad_norm": 1.3678886755306945,
      "learning_rate": 9.405940594059405e-07,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8595590591430664,
      "num_tokens": 52725383.0,
      "step": 1521
    },
    {
      "epoch": 0.28263695450324977,
      "grad_norm": 1.442211346223254,
      "learning_rate": 9.412128712871287e-07,
      "loss": 0.4253,
      "mean_token_accuracy": 0.860120415687561,
      "num_tokens": 52764037.0,
      "step": 1522
    },
    {
      "epoch": 0.2828226555246054,
      "grad_norm": 1.6724003571900716,
      "learning_rate": 9.418316831683168e-07,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8389846682548523,
      "num_tokens": 52795907.0,
      "step": 1523
    },
    {
      "epoch": 0.283008356545961,
      "grad_norm": 1.5556688397157346,
      "learning_rate": 9.424504950495049e-07,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8538561463356018,
      "num_tokens": 52829062.0,
      "step": 1524
    },
    {
      "epoch": 0.28319405756731664,
      "grad_norm": 1.489921370539228,
      "learning_rate": 9.43069306930693e-07,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8675803542137146,
      "num_tokens": 52863079.0,
      "step": 1525
    },
    {
      "epoch": 0.28337975858867226,
      "grad_norm": 1.5780083571027124,
      "learning_rate": 9.436881188118811e-07,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8507876396179199,
      "num_tokens": 52894129.0,
      "step": 1526
    },
    {
      "epoch": 0.2835654596100279,
      "grad_norm": 1.4454659273389314,
      "learning_rate": 9.443069306930692e-07,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8732753396034241,
      "num_tokens": 52928001.0,
      "step": 1527
    },
    {
      "epoch": 0.28375116063138345,
      "grad_norm": 1.665098967150834,
      "learning_rate": 9.449257425742574e-07,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8544598817825317,
      "num_tokens": 52960869.0,
      "step": 1528
    },
    {
      "epoch": 0.28393686165273907,
      "grad_norm": 1.5684443627903424,
      "learning_rate": 9.455445544554454e-07,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8421293497085571,
      "num_tokens": 52999549.0,
      "step": 1529
    },
    {
      "epoch": 0.2841225626740947,
      "grad_norm": 1.5973527026972145,
      "learning_rate": 9.461633663366336e-07,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8581384420394897,
      "num_tokens": 53030820.0,
      "step": 1530
    },
    {
      "epoch": 0.2843082636954503,
      "grad_norm": 1.4189281926808173,
      "learning_rate": 9.467821782178216e-07,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8593242168426514,
      "num_tokens": 53071091.0,
      "step": 1531
    },
    {
      "epoch": 0.28449396471680594,
      "grad_norm": 1.3917442769514934,
      "learning_rate": 9.474009900990099e-07,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8791007399559021,
      "num_tokens": 53108077.0,
      "step": 1532
    },
    {
      "epoch": 0.28467966573816156,
      "grad_norm": 1.6737528666694583,
      "learning_rate": 9.48019801980198e-07,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8502947688102722,
      "num_tokens": 53138926.0,
      "step": 1533
    },
    {
      "epoch": 0.2848653667595172,
      "grad_norm": 2.0386674321345346,
      "learning_rate": 9.486386138613861e-07,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8609585762023926,
      "num_tokens": 53177070.0,
      "step": 1534
    },
    {
      "epoch": 0.2850510677808728,
      "grad_norm": 1.7384688782037157,
      "learning_rate": 9.492574257425742e-07,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8545978665351868,
      "num_tokens": 53208269.0,
      "step": 1535
    },
    {
      "epoch": 0.2852367688022284,
      "grad_norm": 1.7328098899494466,
      "learning_rate": 9.498762376237624e-07,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8532299995422363,
      "num_tokens": 53237653.0,
      "step": 1536
    },
    {
      "epoch": 0.28542246982358405,
      "grad_norm": 1.6068951214257425,
      "learning_rate": 9.504950495049504e-07,
      "loss": 0.459,
      "mean_token_accuracy": 0.8563694953918457,
      "num_tokens": 53268344.0,
      "step": 1537
    },
    {
      "epoch": 0.28560817084493967,
      "grad_norm": 1.5635606415741643,
      "learning_rate": 9.511138613861386e-07,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8626534938812256,
      "num_tokens": 53302171.0,
      "step": 1538
    },
    {
      "epoch": 0.2857938718662953,
      "grad_norm": 1.7813026220174502,
      "learning_rate": 9.517326732673266e-07,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8548761606216431,
      "num_tokens": 53330706.0,
      "step": 1539
    },
    {
      "epoch": 0.28597957288765086,
      "grad_norm": 1.424217512922141,
      "learning_rate": 9.523514851485148e-07,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8615955710411072,
      "num_tokens": 53367405.0,
      "step": 1540
    },
    {
      "epoch": 0.2861652739090065,
      "grad_norm": 1.6650466117971312,
      "learning_rate": 9.529702970297029e-07,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8498550653457642,
      "num_tokens": 53402058.0,
      "step": 1541
    },
    {
      "epoch": 0.2863509749303621,
      "grad_norm": 1.4147471334069581,
      "learning_rate": 9.53589108910891e-07,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8620789051055908,
      "num_tokens": 53440278.0,
      "step": 1542
    },
    {
      "epoch": 0.2865366759517177,
      "grad_norm": 1.4248661515316152,
      "learning_rate": 9.542079207920792e-07,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8739221096038818,
      "num_tokens": 53476260.0,
      "step": 1543
    },
    {
      "epoch": 0.28672237697307335,
      "grad_norm": 1.6353552004174523,
      "learning_rate": 9.548267326732672e-07,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8542624711990356,
      "num_tokens": 53506131.0,
      "step": 1544
    },
    {
      "epoch": 0.28690807799442897,
      "grad_norm": 1.5475043306038072,
      "learning_rate": 9.554455445544553e-07,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8576887845993042,
      "num_tokens": 53538139.0,
      "step": 1545
    },
    {
      "epoch": 0.2870937790157846,
      "grad_norm": 1.4605628454898496,
      "learning_rate": 9.560643564356436e-07,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8509612083435059,
      "num_tokens": 53575818.0,
      "step": 1546
    },
    {
      "epoch": 0.2872794800371402,
      "grad_norm": 1.785788436884738,
      "learning_rate": 9.566831683168316e-07,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8474735617637634,
      "num_tokens": 53602918.0,
      "step": 1547
    },
    {
      "epoch": 0.28746518105849583,
      "grad_norm": 1.5631170677626114,
      "learning_rate": 9.573019801980197e-07,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8702482581138611,
      "num_tokens": 53637077.0,
      "step": 1548
    },
    {
      "epoch": 0.28765088207985146,
      "grad_norm": 1.4798061021523745,
      "learning_rate": 9.579207920792078e-07,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8622060418128967,
      "num_tokens": 53672144.0,
      "step": 1549
    },
    {
      "epoch": 0.2878365831012071,
      "grad_norm": 1.4657671289976693,
      "learning_rate": 9.58539603960396e-07,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8469715714454651,
      "num_tokens": 53712269.0,
      "step": 1550
    },
    {
      "epoch": 0.2880222841225627,
      "grad_norm": 1.5553791074811114,
      "learning_rate": 9.59158415841584e-07,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8686707019805908,
      "num_tokens": 53746480.0,
      "step": 1551
    },
    {
      "epoch": 0.28820798514391827,
      "grad_norm": 1.41306076301295,
      "learning_rate": 9.597772277227723e-07,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8623086214065552,
      "num_tokens": 53786836.0,
      "step": 1552
    },
    {
      "epoch": 0.2883936861652739,
      "grad_norm": 1.5970709829714973,
      "learning_rate": 9.603960396039604e-07,
      "loss": 0.4305,
      "mean_token_accuracy": 0.856288731098175,
      "num_tokens": 53818326.0,
      "step": 1553
    },
    {
      "epoch": 0.2885793871866295,
      "grad_norm": 1.6831348406548798,
      "learning_rate": 9.610148514851485e-07,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8340927362442017,
      "num_tokens": 53850537.0,
      "step": 1554
    },
    {
      "epoch": 0.28876508820798513,
      "grad_norm": 1.5251808979375956,
      "learning_rate": 9.616336633663365e-07,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8611609935760498,
      "num_tokens": 53883621.0,
      "step": 1555
    },
    {
      "epoch": 0.28895078922934075,
      "grad_norm": 1.3205791125555137,
      "learning_rate": 9.622524752475248e-07,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8715118169784546,
      "num_tokens": 53927468.0,
      "step": 1556
    },
    {
      "epoch": 0.2891364902506964,
      "grad_norm": 1.5300364718145394,
      "learning_rate": 9.628712871287129e-07,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8566457629203796,
      "num_tokens": 53967380.0,
      "step": 1557
    },
    {
      "epoch": 0.289322191272052,
      "grad_norm": 1.7171641666449085,
      "learning_rate": 9.63490099009901e-07,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8500198125839233,
      "num_tokens": 53997978.0,
      "step": 1558
    },
    {
      "epoch": 0.2895078922934076,
      "grad_norm": 1.6140087822388969,
      "learning_rate": 9.64108910891089e-07,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8447134494781494,
      "num_tokens": 54029915.0,
      "step": 1559
    },
    {
      "epoch": 0.28969359331476324,
      "grad_norm": 1.5069825179863858,
      "learning_rate": 9.647277227722772e-07,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8603754043579102,
      "num_tokens": 54064434.0,
      "step": 1560
    },
    {
      "epoch": 0.28987929433611886,
      "grad_norm": 1.430610287466769,
      "learning_rate": 9.653465346534653e-07,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8653324842453003,
      "num_tokens": 54101016.0,
      "step": 1561
    },
    {
      "epoch": 0.2900649953574745,
      "grad_norm": 1.4907406396231682,
      "learning_rate": 9.659653465346534e-07,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8511372804641724,
      "num_tokens": 54139704.0,
      "step": 1562
    },
    {
      "epoch": 0.2902506963788301,
      "grad_norm": 1.554548125315098,
      "learning_rate": 9.665841584158414e-07,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8665767908096313,
      "num_tokens": 54174305.0,
      "step": 1563
    },
    {
      "epoch": 0.2904363974001857,
      "grad_norm": 1.5037992774491418,
      "learning_rate": 9.672029702970297e-07,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8422009944915771,
      "num_tokens": 54210480.0,
      "step": 1564
    },
    {
      "epoch": 0.2906220984215413,
      "grad_norm": 1.6412502819744068,
      "learning_rate": 9.678217821782177e-07,
      "loss": 0.4598,
      "mean_token_accuracy": 0.851740300655365,
      "num_tokens": 54244046.0,
      "step": 1565
    },
    {
      "epoch": 0.2908077994428969,
      "grad_norm": 1.4854089729632143,
      "learning_rate": 9.68440594059406e-07,
      "loss": 0.422,
      "mean_token_accuracy": 0.8609938025474548,
      "num_tokens": 54276080.0,
      "step": 1566
    },
    {
      "epoch": 0.29099350046425254,
      "grad_norm": 1.6195473347975242,
      "learning_rate": 9.69059405940594e-07,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8664470314979553,
      "num_tokens": 54301358.0,
      "step": 1567
    },
    {
      "epoch": 0.29117920148560816,
      "grad_norm": 1.4196918965204197,
      "learning_rate": 9.696782178217821e-07,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8591907024383545,
      "num_tokens": 54339851.0,
      "step": 1568
    },
    {
      "epoch": 0.2913649025069638,
      "grad_norm": 1.4944651786799743,
      "learning_rate": 9.702970297029702e-07,
      "loss": 0.3863,
      "mean_token_accuracy": 0.871600866317749,
      "num_tokens": 54373556.0,
      "step": 1569
    },
    {
      "epoch": 0.2915506035283194,
      "grad_norm": 1.4062196511843423,
      "learning_rate": 9.709158415841585e-07,
      "loss": 0.4544,
      "mean_token_accuracy": 0.849468469619751,
      "num_tokens": 54412828.0,
      "step": 1570
    },
    {
      "epoch": 0.29173630454967503,
      "grad_norm": 1.4299116809373755,
      "learning_rate": 9.715346534653465e-07,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8652646541595459,
      "num_tokens": 54449389.0,
      "step": 1571
    },
    {
      "epoch": 0.29192200557103065,
      "grad_norm": 1.770639545521736,
      "learning_rate": 9.721534653465346e-07,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8445607423782349,
      "num_tokens": 54479655.0,
      "step": 1572
    },
    {
      "epoch": 0.2921077065923863,
      "grad_norm": 1.52643796567836,
      "learning_rate": 9.727722772277226e-07,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8596633672714233,
      "num_tokens": 54515939.0,
      "step": 1573
    },
    {
      "epoch": 0.2922934076137419,
      "grad_norm": 1.4373706504266301,
      "learning_rate": 9.73391089108911e-07,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8788428902626038,
      "num_tokens": 54552322.0,
      "step": 1574
    },
    {
      "epoch": 0.2924791086350975,
      "grad_norm": 1.508127219519538,
      "learning_rate": 9.74009900990099e-07,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8482084274291992,
      "num_tokens": 54589483.0,
      "step": 1575
    },
    {
      "epoch": 0.2926648096564531,
      "grad_norm": 1.6270358421214306,
      "learning_rate": 9.74628712871287e-07,
      "loss": 0.441,
      "mean_token_accuracy": 0.8564683794975281,
      "num_tokens": 54622742.0,
      "step": 1576
    },
    {
      "epoch": 0.2928505106778087,
      "grad_norm": 1.6282252112354054,
      "learning_rate": 9.75247524752475e-07,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8517842292785645,
      "num_tokens": 54653961.0,
      "step": 1577
    },
    {
      "epoch": 0.29303621169916433,
      "grad_norm": 1.6586819749858719,
      "learning_rate": 9.758663366336633e-07,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8554067611694336,
      "num_tokens": 54684080.0,
      "step": 1578
    },
    {
      "epoch": 0.29322191272051995,
      "grad_norm": 1.5373625454467827,
      "learning_rate": 9.764851485148514e-07,
      "loss": 0.431,
      "mean_token_accuracy": 0.8602011799812317,
      "num_tokens": 54718614.0,
      "step": 1579
    },
    {
      "epoch": 0.2934076137418756,
      "grad_norm": 1.689362772921351,
      "learning_rate": 9.771039603960397e-07,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8544377684593201,
      "num_tokens": 54752751.0,
      "step": 1580
    },
    {
      "epoch": 0.2935933147632312,
      "grad_norm": 1.5528232122028154,
      "learning_rate": 9.777227722772277e-07,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8666424751281738,
      "num_tokens": 54785409.0,
      "step": 1581
    },
    {
      "epoch": 0.2937790157845868,
      "grad_norm": 1.5310101943490433,
      "learning_rate": 9.783415841584158e-07,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8654036521911621,
      "num_tokens": 54822531.0,
      "step": 1582
    },
    {
      "epoch": 0.29396471680594244,
      "grad_norm": 1.5652172479603157,
      "learning_rate": 9.789603960396039e-07,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8741985559463501,
      "num_tokens": 54853871.0,
      "step": 1583
    },
    {
      "epoch": 0.29415041782729806,
      "grad_norm": 1.377783664618076,
      "learning_rate": 9.795792079207921e-07,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8718087673187256,
      "num_tokens": 54892508.0,
      "step": 1584
    },
    {
      "epoch": 0.2943361188486537,
      "grad_norm": 1.6974977477018316,
      "learning_rate": 9.801980198019802e-07,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8453493118286133,
      "num_tokens": 54920184.0,
      "step": 1585
    },
    {
      "epoch": 0.2945218198700093,
      "grad_norm": 1.6902108695934115,
      "learning_rate": 9.808168316831682e-07,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8437483310699463,
      "num_tokens": 54949942.0,
      "step": 1586
    },
    {
      "epoch": 0.2947075208913649,
      "grad_norm": 1.5913896654804174,
      "learning_rate": 9.814356435643563e-07,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8480834364891052,
      "num_tokens": 54985105.0,
      "step": 1587
    },
    {
      "epoch": 0.2948932219127205,
      "grad_norm": 1.5584511853970662,
      "learning_rate": 9.820544554455446e-07,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8606175184249878,
      "num_tokens": 55021472.0,
      "step": 1588
    },
    {
      "epoch": 0.2950789229340761,
      "grad_norm": 1.6519120336089583,
      "learning_rate": 9.826732673267326e-07,
      "loss": 0.4334,
      "mean_token_accuracy": 0.857021689414978,
      "num_tokens": 55051328.0,
      "step": 1589
    },
    {
      "epoch": 0.29526462395543174,
      "grad_norm": 1.534822362133152,
      "learning_rate": 9.832920792079207e-07,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8676937818527222,
      "num_tokens": 55082963.0,
      "step": 1590
    },
    {
      "epoch": 0.29545032497678736,
      "grad_norm": 1.6856407475121074,
      "learning_rate": 9.83910891089109e-07,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8626550436019897,
      "num_tokens": 55110065.0,
      "step": 1591
    },
    {
      "epoch": 0.295636025998143,
      "grad_norm": 1.3481152271574992,
      "learning_rate": 9.84529702970297e-07,
      "loss": 0.3911,
      "mean_token_accuracy": 0.867979109287262,
      "num_tokens": 55151164.0,
      "step": 1592
    },
    {
      "epoch": 0.2958217270194986,
      "grad_norm": 1.639907410816997,
      "learning_rate": 9.85148514851485e-07,
      "loss": 0.3929,
      "mean_token_accuracy": 0.86637282371521,
      "num_tokens": 55185677.0,
      "step": 1593
    },
    {
      "epoch": 0.2960074280408542,
      "grad_norm": 1.5789333460777681,
      "learning_rate": 9.857673267326733e-07,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8464503884315491,
      "num_tokens": 55221326.0,
      "step": 1594
    },
    {
      "epoch": 0.29619312906220985,
      "grad_norm": 1.525993348653782,
      "learning_rate": 9.863861386138614e-07,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8587958812713623,
      "num_tokens": 55256063.0,
      "step": 1595
    },
    {
      "epoch": 0.29637883008356547,
      "grad_norm": 1.4059702706460266,
      "learning_rate": 9.870049504950495e-07,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8698660135269165,
      "num_tokens": 55291980.0,
      "step": 1596
    },
    {
      "epoch": 0.2965645311049211,
      "grad_norm": 1.5998650625898345,
      "learning_rate": 9.876237623762375e-07,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8582144379615784,
      "num_tokens": 55328273.0,
      "step": 1597
    },
    {
      "epoch": 0.2967502321262767,
      "grad_norm": 1.6399018089143604,
      "learning_rate": 9.882425742574258e-07,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8517614603042603,
      "num_tokens": 55360130.0,
      "step": 1598
    },
    {
      "epoch": 0.29693593314763234,
      "grad_norm": 1.6837896157811254,
      "learning_rate": 9.888613861386138e-07,
      "loss": 0.4039,
      "mean_token_accuracy": 0.868309736251831,
      "num_tokens": 55389445.0,
      "step": 1599
    },
    {
      "epoch": 0.2971216341689879,
      "grad_norm": 1.4842960023877965,
      "learning_rate": 9.89480198019802e-07,
      "loss": 0.401,
      "mean_token_accuracy": 0.8680185675621033,
      "num_tokens": 55424627.0,
      "step": 1600
    },
    {
      "epoch": 0.2973073351903435,
      "grad_norm": 1.6220160685573328,
      "learning_rate": 9.9009900990099e-07,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8390780687332153,
      "num_tokens": 55463496.0,
      "step": 1601
    },
    {
      "epoch": 0.29749303621169915,
      "grad_norm": 1.5762531019238255,
      "learning_rate": 9.907178217821782e-07,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8557156324386597,
      "num_tokens": 55495717.0,
      "step": 1602
    },
    {
      "epoch": 0.29767873723305477,
      "grad_norm": 1.48628543697008,
      "learning_rate": 9.913366336633663e-07,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8538755774497986,
      "num_tokens": 55535463.0,
      "step": 1603
    },
    {
      "epoch": 0.2978644382544104,
      "grad_norm": 1.6058405420763295,
      "learning_rate": 9.919554455445546e-07,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8365948796272278,
      "num_tokens": 55567847.0,
      "step": 1604
    },
    {
      "epoch": 0.298050139275766,
      "grad_norm": 1.6190817810431923,
      "learning_rate": 9.925742574257426e-07,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8440730571746826,
      "num_tokens": 55604761.0,
      "step": 1605
    },
    {
      "epoch": 0.29823584029712163,
      "grad_norm": 1.3293043887743738,
      "learning_rate": 9.931930693069307e-07,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8626724481582642,
      "num_tokens": 55648358.0,
      "step": 1606
    },
    {
      "epoch": 0.29842154131847726,
      "grad_norm": 1.4788526556397374,
      "learning_rate": 9.938118811881187e-07,
      "loss": 0.4089,
      "mean_token_accuracy": 0.863452672958374,
      "num_tokens": 55685482.0,
      "step": 1607
    },
    {
      "epoch": 0.2986072423398329,
      "grad_norm": 1.3937940898644832,
      "learning_rate": 9.94430693069307e-07,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8645181655883789,
      "num_tokens": 55726115.0,
      "step": 1608
    },
    {
      "epoch": 0.2987929433611885,
      "grad_norm": 1.6489246519290826,
      "learning_rate": 9.95049504950495e-07,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8660329580307007,
      "num_tokens": 55752330.0,
      "step": 1609
    },
    {
      "epoch": 0.2989786443825441,
      "grad_norm": 1.5536484917654327,
      "learning_rate": 9.956683168316831e-07,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8673431873321533,
      "num_tokens": 55789608.0,
      "step": 1610
    },
    {
      "epoch": 0.29916434540389975,
      "grad_norm": 1.5874242229975362,
      "learning_rate": 9.962871287128712e-07,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8553483486175537,
      "num_tokens": 55822053.0,
      "step": 1611
    },
    {
      "epoch": 0.2993500464252553,
      "grad_norm": 1.5772874317504146,
      "learning_rate": 9.969059405940595e-07,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8644309043884277,
      "num_tokens": 55857581.0,
      "step": 1612
    },
    {
      "epoch": 0.29953574744661093,
      "grad_norm": 1.590390366759092,
      "learning_rate": 9.975247524752475e-07,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8528411984443665,
      "num_tokens": 55892123.0,
      "step": 1613
    },
    {
      "epoch": 0.29972144846796656,
      "grad_norm": 1.56173510140426,
      "learning_rate": 9.981435643564356e-07,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8706243634223938,
      "num_tokens": 55926473.0,
      "step": 1614
    },
    {
      "epoch": 0.2999071494893222,
      "grad_norm": 1.6133157603358252,
      "learning_rate": 9.987623762376236e-07,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8543341159820557,
      "num_tokens": 55955793.0,
      "step": 1615
    },
    {
      "epoch": 0.3000928505106778,
      "grad_norm": 1.6193828754241575,
      "learning_rate": 9.99381188118812e-07,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8576906323432922,
      "num_tokens": 55988226.0,
      "step": 1616
    },
    {
      "epoch": 0.3002785515320334,
      "grad_norm": 1.5460734865486674,
      "learning_rate": 1e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8652065992355347,
      "num_tokens": 56018013.0,
      "step": 1617
    },
    {
      "epoch": 0.30046425255338904,
      "grad_norm": 1.46356394427953,
      "learning_rate": 1e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8430871367454529,
      "num_tokens": 56058366.0,
      "step": 1618
    },
    {
      "epoch": 0.30064995357474467,
      "grad_norm": 1.5415428795486532,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8670870065689087,
      "num_tokens": 56089141.0,
      "step": 1619
    },
    {
      "epoch": 0.3008356545961003,
      "grad_norm": 1.4507792084841968,
      "learning_rate": 1e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8487648367881775,
      "num_tokens": 56127433.0,
      "step": 1620
    },
    {
      "epoch": 0.3010213556174559,
      "grad_norm": 1.6380820214504508,
      "learning_rate": 1e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8439376354217529,
      "num_tokens": 56164126.0,
      "step": 1621
    },
    {
      "epoch": 0.30120705663881153,
      "grad_norm": 1.467404892386368,
      "learning_rate": 1e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8569790124893188,
      "num_tokens": 56201671.0,
      "step": 1622
    },
    {
      "epoch": 0.30139275766016715,
      "grad_norm": 1.5518241498544743,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8684208393096924,
      "num_tokens": 56236756.0,
      "step": 1623
    },
    {
      "epoch": 0.3015784586815227,
      "grad_norm": 1.5552603440869557,
      "learning_rate": 1e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8616358041763306,
      "num_tokens": 56271782.0,
      "step": 1624
    },
    {
      "epoch": 0.30176415970287834,
      "grad_norm": 1.502393245656906,
      "learning_rate": 1e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8542432188987732,
      "num_tokens": 56305415.0,
      "step": 1625
    },
    {
      "epoch": 0.30194986072423396,
      "grad_norm": 1.4099019420808236,
      "learning_rate": 1e-06,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8651583790779114,
      "num_tokens": 56345015.0,
      "step": 1626
    },
    {
      "epoch": 0.3021355617455896,
      "grad_norm": 1.5659833739944076,
      "learning_rate": 1e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8530951142311096,
      "num_tokens": 56380120.0,
      "step": 1627
    },
    {
      "epoch": 0.3023212627669452,
      "grad_norm": 1.4846792749581503,
      "learning_rate": 1e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8504331111907959,
      "num_tokens": 56416006.0,
      "step": 1628
    },
    {
      "epoch": 0.30250696378830083,
      "grad_norm": 1.4522305932466357,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8709071278572083,
      "num_tokens": 56449225.0,
      "step": 1629
    },
    {
      "epoch": 0.30269266480965645,
      "grad_norm": 1.5586858945971902,
      "learning_rate": 1e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8511264324188232,
      "num_tokens": 56488723.0,
      "step": 1630
    },
    {
      "epoch": 0.3028783658310121,
      "grad_norm": 1.5663389270012278,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8633700609207153,
      "num_tokens": 56519211.0,
      "step": 1631
    },
    {
      "epoch": 0.3030640668523677,
      "grad_norm": 1.4864751806470906,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8611726760864258,
      "num_tokens": 56552472.0,
      "step": 1632
    },
    {
      "epoch": 0.3032497678737233,
      "grad_norm": 1.497268632877609,
      "learning_rate": 1e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8475034236907959,
      "num_tokens": 56594155.0,
      "step": 1633
    },
    {
      "epoch": 0.30343546889507894,
      "grad_norm": 1.6330060047496076,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8653692007064819,
      "num_tokens": 56623320.0,
      "step": 1634
    },
    {
      "epoch": 0.30362116991643456,
      "grad_norm": 1.8492984401659727,
      "learning_rate": 1e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8400886654853821,
      "num_tokens": 56648737.0,
      "step": 1635
    },
    {
      "epoch": 0.3038068709377902,
      "grad_norm": 1.530575292811062,
      "learning_rate": 1e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8540325164794922,
      "num_tokens": 56685191.0,
      "step": 1636
    },
    {
      "epoch": 0.30399257195914575,
      "grad_norm": 1.4360292895953204,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.866916298866272,
      "num_tokens": 56722224.0,
      "step": 1637
    },
    {
      "epoch": 0.3041782729805014,
      "grad_norm": 1.521759406129361,
      "learning_rate": 1e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8480926752090454,
      "num_tokens": 56756400.0,
      "step": 1638
    },
    {
      "epoch": 0.304363974001857,
      "grad_norm": 1.495195308081271,
      "learning_rate": 1e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8460913300514221,
      "num_tokens": 56797535.0,
      "step": 1639
    },
    {
      "epoch": 0.3045496750232126,
      "grad_norm": 1.506124871613103,
      "learning_rate": 1e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8623368740081787,
      "num_tokens": 56830985.0,
      "step": 1640
    },
    {
      "epoch": 0.30473537604456824,
      "grad_norm": 1.4031785696218468,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8570269346237183,
      "num_tokens": 56869524.0,
      "step": 1641
    },
    {
      "epoch": 0.30492107706592386,
      "grad_norm": 1.5306543362133354,
      "learning_rate": 1e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8485794067382812,
      "num_tokens": 56904394.0,
      "step": 1642
    },
    {
      "epoch": 0.3051067780872795,
      "grad_norm": 1.6181317975115515,
      "learning_rate": 1e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8531712889671326,
      "num_tokens": 56936546.0,
      "step": 1643
    },
    {
      "epoch": 0.3052924791086351,
      "grad_norm": 1.6223662912584778,
      "learning_rate": 1e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8483788371086121,
      "num_tokens": 56969019.0,
      "step": 1644
    },
    {
      "epoch": 0.30547818012999073,
      "grad_norm": 1.5224936388940968,
      "learning_rate": 1e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8586894273757935,
      "num_tokens": 57003724.0,
      "step": 1645
    },
    {
      "epoch": 0.30566388115134635,
      "grad_norm": 1.86257880184402,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.864911675453186,
      "num_tokens": 57027858.0,
      "step": 1646
    },
    {
      "epoch": 0.305849582172702,
      "grad_norm": 1.5885605228279869,
      "learning_rate": 1e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8470951914787292,
      "num_tokens": 57062290.0,
      "step": 1647
    },
    {
      "epoch": 0.3060352831940576,
      "grad_norm": 1.6679527687293982,
      "learning_rate": 1e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.84454345703125,
      "num_tokens": 57094239.0,
      "step": 1648
    },
    {
      "epoch": 0.30622098421541316,
      "grad_norm": 1.6354551664833863,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8635281920433044,
      "num_tokens": 57126571.0,
      "step": 1649
    },
    {
      "epoch": 0.3064066852367688,
      "grad_norm": 1.4418594608581845,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.862000048160553,
      "num_tokens": 57165411.0,
      "step": 1650
    },
    {
      "epoch": 0.3065923862581244,
      "grad_norm": 1.5639737301848127,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8648136854171753,
      "num_tokens": 57197224.0,
      "step": 1651
    },
    {
      "epoch": 0.30677808727948,
      "grad_norm": 1.463061916891823,
      "learning_rate": 1e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8551162481307983,
      "num_tokens": 57236506.0,
      "step": 1652
    },
    {
      "epoch": 0.30696378830083565,
      "grad_norm": 1.5382975466482343,
      "learning_rate": 1e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8558257222175598,
      "num_tokens": 57272992.0,
      "step": 1653
    },
    {
      "epoch": 0.30714948932219127,
      "grad_norm": 1.405016693263633,
      "learning_rate": 1e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8506067991256714,
      "num_tokens": 57314499.0,
      "step": 1654
    },
    {
      "epoch": 0.3073351903435469,
      "grad_norm": 1.7545230182886697,
      "learning_rate": 1e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8510686159133911,
      "num_tokens": 57344259.0,
      "step": 1655
    },
    {
      "epoch": 0.3075208913649025,
      "grad_norm": 1.3907157428259393,
      "learning_rate": 1e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.85716712474823,
      "num_tokens": 57384862.0,
      "step": 1656
    },
    {
      "epoch": 0.30770659238625814,
      "grad_norm": 1.4832205790018909,
      "learning_rate": 1e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8542877435684204,
      "num_tokens": 57425207.0,
      "step": 1657
    },
    {
      "epoch": 0.30789229340761376,
      "grad_norm": 1.601188410985148,
      "learning_rate": 1e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8479804396629333,
      "num_tokens": 57458708.0,
      "step": 1658
    },
    {
      "epoch": 0.3080779944289694,
      "grad_norm": 1.5379187295014762,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8709402084350586,
      "num_tokens": 57490566.0,
      "step": 1659
    },
    {
      "epoch": 0.308263695450325,
      "grad_norm": 1.5427207094047586,
      "learning_rate": 1e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.842451274394989,
      "num_tokens": 57528787.0,
      "step": 1660
    },
    {
      "epoch": 0.30844939647168057,
      "grad_norm": 1.6891579616035761,
      "learning_rate": 1e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8316124677658081,
      "num_tokens": 57559675.0,
      "step": 1661
    },
    {
      "epoch": 0.3086350974930362,
      "grad_norm": 1.5781996487314642,
      "learning_rate": 1e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8565191626548767,
      "num_tokens": 57593499.0,
      "step": 1662
    },
    {
      "epoch": 0.3088207985143918,
      "grad_norm": 1.6271675990337326,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8680340051651001,
      "num_tokens": 57625211.0,
      "step": 1663
    },
    {
      "epoch": 0.30900649953574744,
      "grad_norm": 1.4682449510395714,
      "learning_rate": 1e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8611641526222229,
      "num_tokens": 57663619.0,
      "step": 1664
    },
    {
      "epoch": 0.30919220055710306,
      "grad_norm": 1.4852288868208956,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.861431360244751,
      "num_tokens": 57700653.0,
      "step": 1665
    },
    {
      "epoch": 0.3093779015784587,
      "grad_norm": 1.5283230385304984,
      "learning_rate": 1e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8537585139274597,
      "num_tokens": 57737288.0,
      "step": 1666
    },
    {
      "epoch": 0.3095636025998143,
      "grad_norm": 1.5205731228577324,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8691229224205017,
      "num_tokens": 57767982.0,
      "step": 1667
    },
    {
      "epoch": 0.3097493036211699,
      "grad_norm": 1.352824809524275,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8631271719932556,
      "num_tokens": 57809796.0,
      "step": 1668
    },
    {
      "epoch": 0.30993500464252555,
      "grad_norm": 1.3827019927795832,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8609300851821899,
      "num_tokens": 57851137.0,
      "step": 1669
    },
    {
      "epoch": 0.31012070566388117,
      "grad_norm": 1.5066169167679082,
      "learning_rate": 1e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8459450006484985,
      "num_tokens": 57886443.0,
      "step": 1670
    },
    {
      "epoch": 0.3103064066852368,
      "grad_norm": 1.845821792484292,
      "learning_rate": 1e-06,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8352432250976562,
      "num_tokens": 57917239.0,
      "step": 1671
    },
    {
      "epoch": 0.3104921077065924,
      "grad_norm": 1.5342888239831358,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8658697605133057,
      "num_tokens": 57956316.0,
      "step": 1672
    },
    {
      "epoch": 0.310677808727948,
      "grad_norm": 1.5041684808081097,
      "learning_rate": 1e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8489364385604858,
      "num_tokens": 57994065.0,
      "step": 1673
    },
    {
      "epoch": 0.3108635097493036,
      "grad_norm": 1.4726605117828935,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8605548739433289,
      "num_tokens": 58030108.0,
      "step": 1674
    },
    {
      "epoch": 0.3110492107706592,
      "grad_norm": 1.3304540107529637,
      "learning_rate": 1e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8683703541755676,
      "num_tokens": 58074630.0,
      "step": 1675
    },
    {
      "epoch": 0.31123491179201485,
      "grad_norm": 1.4358225580138175,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8577678203582764,
      "num_tokens": 58112940.0,
      "step": 1676
    },
    {
      "epoch": 0.31142061281337047,
      "grad_norm": 1.4463860564721969,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8698840737342834,
      "num_tokens": 58154383.0,
      "step": 1677
    },
    {
      "epoch": 0.3116063138347261,
      "grad_norm": 1.5141820248192301,
      "learning_rate": 1e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8494489192962646,
      "num_tokens": 58190159.0,
      "step": 1678
    },
    {
      "epoch": 0.3117920148560817,
      "grad_norm": 1.5182445310410546,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8654921054840088,
      "num_tokens": 58223470.0,
      "step": 1679
    },
    {
      "epoch": 0.31197771587743733,
      "grad_norm": 1.7110132581017754,
      "learning_rate": 1e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8462337255477905,
      "num_tokens": 58254715.0,
      "step": 1680
    },
    {
      "epoch": 0.31216341689879296,
      "grad_norm": 1.501035983742982,
      "learning_rate": 1e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8563048839569092,
      "num_tokens": 58290565.0,
      "step": 1681
    },
    {
      "epoch": 0.3123491179201486,
      "grad_norm": 1.4567294096090955,
      "learning_rate": 1e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8597830533981323,
      "num_tokens": 58328948.0,
      "step": 1682
    },
    {
      "epoch": 0.3125348189415042,
      "grad_norm": 1.69190960083609,
      "learning_rate": 1e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8642234206199646,
      "num_tokens": 58357564.0,
      "step": 1683
    },
    {
      "epoch": 0.3127205199628598,
      "grad_norm": 1.4820994206433924,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.869255006313324,
      "num_tokens": 58394351.0,
      "step": 1684
    },
    {
      "epoch": 0.3129062209842154,
      "grad_norm": 1.5516126197304652,
      "learning_rate": 1e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8294196724891663,
      "num_tokens": 58431496.0,
      "step": 1685
    },
    {
      "epoch": 0.313091922005571,
      "grad_norm": 1.4657176483403092,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8611133098602295,
      "num_tokens": 58468185.0,
      "step": 1686
    },
    {
      "epoch": 0.31327762302692663,
      "grad_norm": 1.3363130810535748,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8537344336509705,
      "num_tokens": 58511923.0,
      "step": 1687
    },
    {
      "epoch": 0.31346332404828225,
      "grad_norm": 1.4658037244147522,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8711233735084534,
      "num_tokens": 58549347.0,
      "step": 1688
    },
    {
      "epoch": 0.3136490250696379,
      "grad_norm": 1.5357595293068476,
      "learning_rate": 1e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8545816540718079,
      "num_tokens": 58584897.0,
      "step": 1689
    },
    {
      "epoch": 0.3138347260909935,
      "grad_norm": 1.5227059408686408,
      "learning_rate": 1e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8435717821121216,
      "num_tokens": 58624265.0,
      "step": 1690
    },
    {
      "epoch": 0.3140204271123491,
      "grad_norm": 1.496285551929833,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8638278245925903,
      "num_tokens": 58658477.0,
      "step": 1691
    },
    {
      "epoch": 0.31420612813370474,
      "grad_norm": 1.5520164165443981,
      "learning_rate": 1e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8637183308601379,
      "num_tokens": 58692806.0,
      "step": 1692
    },
    {
      "epoch": 0.31439182915506036,
      "grad_norm": 1.5669245696170906,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8622049689292908,
      "num_tokens": 58729150.0,
      "step": 1693
    },
    {
      "epoch": 0.314577530176416,
      "grad_norm": 1.757484742887451,
      "learning_rate": 1e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8410577178001404,
      "num_tokens": 58757111.0,
      "step": 1694
    },
    {
      "epoch": 0.3147632311977716,
      "grad_norm": 1.5727947122125776,
      "learning_rate": 1e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.852459192276001,
      "num_tokens": 58793088.0,
      "step": 1695
    },
    {
      "epoch": 0.31494893221912723,
      "grad_norm": 1.4907921379478268,
      "learning_rate": 1e-06,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8701216578483582,
      "num_tokens": 58832559.0,
      "step": 1696
    },
    {
      "epoch": 0.3151346332404828,
      "grad_norm": 1.4949875427540447,
      "learning_rate": 1e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8414640426635742,
      "num_tokens": 58871428.0,
      "step": 1697
    },
    {
      "epoch": 0.3153203342618384,
      "grad_norm": 1.6434823742252975,
      "learning_rate": 1e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8510246872901917,
      "num_tokens": 58903481.0,
      "step": 1698
    },
    {
      "epoch": 0.31550603528319404,
      "grad_norm": 1.5396451671040758,
      "learning_rate": 1e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8593026399612427,
      "num_tokens": 58937924.0,
      "step": 1699
    },
    {
      "epoch": 0.31569173630454966,
      "grad_norm": 1.7975826129695294,
      "learning_rate": 1e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8542457818984985,
      "num_tokens": 58968606.0,
      "step": 1700
    },
    {
      "epoch": 0.3158774373259053,
      "grad_norm": 1.518615578230613,
      "learning_rate": 1e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.849838376045227,
      "num_tokens": 59004567.0,
      "step": 1701
    },
    {
      "epoch": 0.3160631383472609,
      "grad_norm": 1.5870153142095291,
      "learning_rate": 1e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8561067581176758,
      "num_tokens": 59036689.0,
      "step": 1702
    },
    {
      "epoch": 0.31624883936861653,
      "grad_norm": 1.6430596726417634,
      "learning_rate": 1e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8535382747650146,
      "num_tokens": 59072120.0,
      "step": 1703
    },
    {
      "epoch": 0.31643454038997215,
      "grad_norm": 1.4076381767178625,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8645195960998535,
      "num_tokens": 59107849.0,
      "step": 1704
    },
    {
      "epoch": 0.3166202414113278,
      "grad_norm": 1.522423448397562,
      "learning_rate": 1e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8488714694976807,
      "num_tokens": 59144736.0,
      "step": 1705
    },
    {
      "epoch": 0.3168059424326834,
      "grad_norm": 1.6446314546288752,
      "learning_rate": 1e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8547555208206177,
      "num_tokens": 59174536.0,
      "step": 1706
    },
    {
      "epoch": 0.316991643454039,
      "grad_norm": 1.5991734763528271,
      "learning_rate": 1e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8625243902206421,
      "num_tokens": 59210996.0,
      "step": 1707
    },
    {
      "epoch": 0.31717734447539464,
      "grad_norm": 1.4763174796283167,
      "learning_rate": 1e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8524715304374695,
      "num_tokens": 59244911.0,
      "step": 1708
    },
    {
      "epoch": 0.3173630454967502,
      "grad_norm": 1.498246673334938,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.867963433265686,
      "num_tokens": 59279379.0,
      "step": 1709
    },
    {
      "epoch": 0.31754874651810583,
      "grad_norm": 1.5172724119916303,
      "learning_rate": 1e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8353493213653564,
      "num_tokens": 59317376.0,
      "step": 1710
    },
    {
      "epoch": 0.31773444753946145,
      "grad_norm": 1.671753790190869,
      "learning_rate": 1e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.843856155872345,
      "num_tokens": 59349842.0,
      "step": 1711
    },
    {
      "epoch": 0.3179201485608171,
      "grad_norm": 1.5048510251488718,
      "learning_rate": 1e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8482756614685059,
      "num_tokens": 59387671.0,
      "step": 1712
    },
    {
      "epoch": 0.3181058495821727,
      "grad_norm": 1.413013610230105,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8679434061050415,
      "num_tokens": 59423743.0,
      "step": 1713
    },
    {
      "epoch": 0.3182915506035283,
      "grad_norm": 1.5649114111781255,
      "learning_rate": 1e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8463733196258545,
      "num_tokens": 59458674.0,
      "step": 1714
    },
    {
      "epoch": 0.31847725162488394,
      "grad_norm": 1.5992536751451822,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8543962240219116,
      "num_tokens": 59493899.0,
      "step": 1715
    },
    {
      "epoch": 0.31866295264623956,
      "grad_norm": 1.870035746462209,
      "learning_rate": 1e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8659893274307251,
      "num_tokens": 59526339.0,
      "step": 1716
    },
    {
      "epoch": 0.3188486536675952,
      "grad_norm": 1.5859348205982535,
      "learning_rate": 1e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.859182596206665,
      "num_tokens": 59557141.0,
      "step": 1717
    },
    {
      "epoch": 0.3190343546889508,
      "grad_norm": 1.4987350889633884,
      "learning_rate": 1e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.855510413646698,
      "num_tokens": 59593493.0,
      "step": 1718
    },
    {
      "epoch": 0.3192200557103064,
      "grad_norm": 1.6307617343037455,
      "learning_rate": 1e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8439438343048096,
      "num_tokens": 59628663.0,
      "step": 1719
    },
    {
      "epoch": 0.31940575673166205,
      "grad_norm": 1.5747675827783745,
      "learning_rate": 1e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8500871658325195,
      "num_tokens": 59664773.0,
      "step": 1720
    },
    {
      "epoch": 0.3195914577530176,
      "grad_norm": 1.5951693809410195,
      "learning_rate": 1e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8648691177368164,
      "num_tokens": 59695222.0,
      "step": 1721
    },
    {
      "epoch": 0.31977715877437324,
      "grad_norm": 1.527050444662757,
      "learning_rate": 1e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.838672935962677,
      "num_tokens": 59734828.0,
      "step": 1722
    },
    {
      "epoch": 0.31996285979572886,
      "grad_norm": 1.539470534775944,
      "learning_rate": 1e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8596136569976807,
      "num_tokens": 59767538.0,
      "step": 1723
    },
    {
      "epoch": 0.3201485608170845,
      "grad_norm": 1.517048686600622,
      "learning_rate": 1e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8481153249740601,
      "num_tokens": 59804346.0,
      "step": 1724
    },
    {
      "epoch": 0.3203342618384401,
      "grad_norm": 1.5311968029940208,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8576966524124146,
      "num_tokens": 59838495.0,
      "step": 1725
    },
    {
      "epoch": 0.3205199628597957,
      "grad_norm": 1.6513297483111293,
      "learning_rate": 1e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8530781269073486,
      "num_tokens": 59869247.0,
      "step": 1726
    },
    {
      "epoch": 0.32070566388115135,
      "grad_norm": 1.665334735286949,
      "learning_rate": 1e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8550037741661072,
      "num_tokens": 59898871.0,
      "step": 1727
    },
    {
      "epoch": 0.32089136490250697,
      "grad_norm": 1.5153221825722598,
      "learning_rate": 1e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.8531752228736877,
      "num_tokens": 59937524.0,
      "step": 1728
    },
    {
      "epoch": 0.3210770659238626,
      "grad_norm": 1.4506732263445246,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8628388047218323,
      "num_tokens": 59972916.0,
      "step": 1729
    },
    {
      "epoch": 0.3212627669452182,
      "grad_norm": 1.4912310218271032,
      "learning_rate": 1e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.854386568069458,
      "num_tokens": 60009361.0,
      "step": 1730
    },
    {
      "epoch": 0.32144846796657384,
      "grad_norm": 1.5793233685527484,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8692037463188171,
      "num_tokens": 60039068.0,
      "step": 1731
    },
    {
      "epoch": 0.32163416898792946,
      "grad_norm": 1.4106254657224646,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8667815327644348,
      "num_tokens": 60078528.0,
      "step": 1732
    },
    {
      "epoch": 0.321819870009285,
      "grad_norm": 1.5441951518557244,
      "learning_rate": 1e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8495956659317017,
      "num_tokens": 60112273.0,
      "step": 1733
    },
    {
      "epoch": 0.32200557103064065,
      "grad_norm": 1.5840903158699926,
      "learning_rate": 1e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8527540564537048,
      "num_tokens": 60144098.0,
      "step": 1734
    },
    {
      "epoch": 0.32219127205199627,
      "grad_norm": 1.691869262128796,
      "learning_rate": 1e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8559667468070984,
      "num_tokens": 60171689.0,
      "step": 1735
    },
    {
      "epoch": 0.3223769730733519,
      "grad_norm": 1.5386186335441632,
      "learning_rate": 1e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8578218221664429,
      "num_tokens": 60208851.0,
      "step": 1736
    },
    {
      "epoch": 0.3225626740947075,
      "grad_norm": 1.5682027236343183,
      "learning_rate": 1e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8556807637214661,
      "num_tokens": 60241557.0,
      "step": 1737
    },
    {
      "epoch": 0.32274837511606314,
      "grad_norm": 1.7192095043207842,
      "learning_rate": 1e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8537003993988037,
      "num_tokens": 60268741.0,
      "step": 1738
    },
    {
      "epoch": 0.32293407613741876,
      "grad_norm": 1.444922212083601,
      "learning_rate": 1e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8522086143493652,
      "num_tokens": 60308208.0,
      "step": 1739
    },
    {
      "epoch": 0.3231197771587744,
      "grad_norm": 1.6177547830103975,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8725876808166504,
      "num_tokens": 60339073.0,
      "step": 1740
    },
    {
      "epoch": 0.32330547818013,
      "grad_norm": 1.3918607143149024,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8630613088607788,
      "num_tokens": 60376931.0,
      "step": 1741
    },
    {
      "epoch": 0.3234911792014856,
      "grad_norm": 1.6251722913366342,
      "learning_rate": 1e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8422373533248901,
      "num_tokens": 60409454.0,
      "step": 1742
    },
    {
      "epoch": 0.32367688022284125,
      "grad_norm": 1.462467826032611,
      "learning_rate": 1e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8559073209762573,
      "num_tokens": 60448713.0,
      "step": 1743
    },
    {
      "epoch": 0.32386258124419687,
      "grad_norm": 1.588634012201314,
      "learning_rate": 1e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8566023111343384,
      "num_tokens": 60480731.0,
      "step": 1744
    },
    {
      "epoch": 0.32404828226555243,
      "grad_norm": 1.5743444564710034,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8665814399719238,
      "num_tokens": 60511360.0,
      "step": 1745
    },
    {
      "epoch": 0.32423398328690806,
      "grad_norm": 1.5622392554942772,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8628738522529602,
      "num_tokens": 60545323.0,
      "step": 1746
    },
    {
      "epoch": 0.3244196843082637,
      "grad_norm": 1.382769566680152,
      "learning_rate": 1e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8536832928657532,
      "num_tokens": 60585091.0,
      "step": 1747
    },
    {
      "epoch": 0.3246053853296193,
      "grad_norm": 1.5526407927085184,
      "learning_rate": 1e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8482687473297119,
      "num_tokens": 60619383.0,
      "step": 1748
    },
    {
      "epoch": 0.3247910863509749,
      "grad_norm": 1.5385802325596827,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.866637110710144,
      "num_tokens": 60652575.0,
      "step": 1749
    },
    {
      "epoch": 0.32497678737233054,
      "grad_norm": 1.39527453088143,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8712599873542786,
      "num_tokens": 60689744.0,
      "step": 1750
    },
    {
      "epoch": 0.32516248839368617,
      "grad_norm": 1.4154769684668969,
      "learning_rate": 1e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8537194728851318,
      "num_tokens": 60728178.0,
      "step": 1751
    },
    {
      "epoch": 0.3253481894150418,
      "grad_norm": 1.4551008098682388,
      "learning_rate": 1e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8655731678009033,
      "num_tokens": 60765659.0,
      "step": 1752
    },
    {
      "epoch": 0.3255338904363974,
      "grad_norm": 1.5251316455761306,
      "learning_rate": 1e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8637884259223938,
      "num_tokens": 60800477.0,
      "step": 1753
    },
    {
      "epoch": 0.32571959145775303,
      "grad_norm": 1.371146760713288,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8695881366729736,
      "num_tokens": 60839322.0,
      "step": 1754
    },
    {
      "epoch": 0.32590529247910865,
      "grad_norm": 1.6677489402034562,
      "learning_rate": 1e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8592965006828308,
      "num_tokens": 60869288.0,
      "step": 1755
    },
    {
      "epoch": 0.3260909935004643,
      "grad_norm": 1.5089606231027717,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8685321807861328,
      "num_tokens": 60900896.0,
      "step": 1756
    },
    {
      "epoch": 0.32627669452181984,
      "grad_norm": 1.5475679565268758,
      "learning_rate": 1e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8603056073188782,
      "num_tokens": 60938515.0,
      "step": 1757
    },
    {
      "epoch": 0.32646239554317547,
      "grad_norm": 1.680674058592137,
      "learning_rate": 1e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8316162824630737,
      "num_tokens": 60970370.0,
      "step": 1758
    },
    {
      "epoch": 0.3266480965645311,
      "grad_norm": 1.4709129156326122,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8706055283546448,
      "num_tokens": 61008052.0,
      "step": 1759
    },
    {
      "epoch": 0.3268337975858867,
      "grad_norm": 1.5116835358157104,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8637780547142029,
      "num_tokens": 61043612.0,
      "step": 1760
    },
    {
      "epoch": 0.32701949860724233,
      "grad_norm": 1.534933678907991,
      "learning_rate": 1e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8518391847610474,
      "num_tokens": 61075317.0,
      "step": 1761
    },
    {
      "epoch": 0.32720519962859795,
      "grad_norm": 1.4631477452416495,
      "learning_rate": 1e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8578075766563416,
      "num_tokens": 61114027.0,
      "step": 1762
    },
    {
      "epoch": 0.3273909006499536,
      "grad_norm": 1.4797348007539515,
      "learning_rate": 1e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8605371713638306,
      "num_tokens": 61149839.0,
      "step": 1763
    },
    {
      "epoch": 0.3275766016713092,
      "grad_norm": 1.5044216745073,
      "learning_rate": 1e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8608366250991821,
      "num_tokens": 61185713.0,
      "step": 1764
    },
    {
      "epoch": 0.3277623026926648,
      "grad_norm": 1.4722873622626897,
      "learning_rate": 1e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8539459705352783,
      "num_tokens": 61229492.0,
      "step": 1765
    },
    {
      "epoch": 0.32794800371402044,
      "grad_norm": 1.5133643560896692,
      "learning_rate": 1e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8531932830810547,
      "num_tokens": 61265128.0,
      "step": 1766
    },
    {
      "epoch": 0.32813370473537606,
      "grad_norm": 1.5167334614967203,
      "learning_rate": 1e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8604676127433777,
      "num_tokens": 61300418.0,
      "step": 1767
    },
    {
      "epoch": 0.3283194057567317,
      "grad_norm": 1.572766057199996,
      "learning_rate": 1e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8471555709838867,
      "num_tokens": 61332284.0,
      "step": 1768
    },
    {
      "epoch": 0.32850510677808725,
      "grad_norm": 1.60374041597051,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8653621673583984,
      "num_tokens": 61364453.0,
      "step": 1769
    },
    {
      "epoch": 0.3286908077994429,
      "grad_norm": 1.5689296380885003,
      "learning_rate": 1e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8451554775238037,
      "num_tokens": 61398672.0,
      "step": 1770
    },
    {
      "epoch": 0.3288765088207985,
      "grad_norm": 1.5719334388542225,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8708016872406006,
      "num_tokens": 61428518.0,
      "step": 1771
    },
    {
      "epoch": 0.3290622098421541,
      "grad_norm": 1.5078692660161326,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8693392276763916,
      "num_tokens": 61459367.0,
      "step": 1772
    },
    {
      "epoch": 0.32924791086350974,
      "grad_norm": 1.6379706132022598,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8742743134498596,
      "num_tokens": 61488156.0,
      "step": 1773
    },
    {
      "epoch": 0.32943361188486536,
      "grad_norm": 1.6208654887193155,
      "learning_rate": 1e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8534799814224243,
      "num_tokens": 61519881.0,
      "step": 1774
    },
    {
      "epoch": 0.329619312906221,
      "grad_norm": 1.5435732296860987,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8643326163291931,
      "num_tokens": 61559777.0,
      "step": 1775
    },
    {
      "epoch": 0.3298050139275766,
      "grad_norm": 1.5855102975258637,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8578215837478638,
      "num_tokens": 61596363.0,
      "step": 1776
    },
    {
      "epoch": 0.32999071494893223,
      "grad_norm": 1.426572450783464,
      "learning_rate": 1e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8472676277160645,
      "num_tokens": 61638705.0,
      "step": 1777
    },
    {
      "epoch": 0.33017641597028785,
      "grad_norm": 1.4440598577284882,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8700757026672363,
      "num_tokens": 61676275.0,
      "step": 1778
    },
    {
      "epoch": 0.3303621169916435,
      "grad_norm": 1.3771523766256522,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8659448623657227,
      "num_tokens": 61717127.0,
      "step": 1779
    },
    {
      "epoch": 0.3305478180129991,
      "grad_norm": 1.6782623366947422,
      "learning_rate": 1e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8497812747955322,
      "num_tokens": 61744742.0,
      "step": 1780
    },
    {
      "epoch": 0.33073351903435466,
      "grad_norm": 1.6094953087802428,
      "learning_rate": 1e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8562500476837158,
      "num_tokens": 61777772.0,
      "step": 1781
    },
    {
      "epoch": 0.3309192200557103,
      "grad_norm": 1.480722468528412,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8799023032188416,
      "num_tokens": 61810659.0,
      "step": 1782
    },
    {
      "epoch": 0.3311049210770659,
      "grad_norm": 1.497024115656675,
      "learning_rate": 1e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.851799488067627,
      "num_tokens": 61847277.0,
      "step": 1783
    },
    {
      "epoch": 0.3312906220984215,
      "grad_norm": 1.3861728038005174,
      "learning_rate": 1e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8554753065109253,
      "num_tokens": 61886052.0,
      "step": 1784
    },
    {
      "epoch": 0.33147632311977715,
      "grad_norm": 1.486276151488608,
      "learning_rate": 1e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8539999723434448,
      "num_tokens": 61919581.0,
      "step": 1785
    },
    {
      "epoch": 0.33166202414113277,
      "grad_norm": 1.586347593036492,
      "learning_rate": 1e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8572384119033813,
      "num_tokens": 61955577.0,
      "step": 1786
    },
    {
      "epoch": 0.3318477251624884,
      "grad_norm": 1.4236098108761095,
      "learning_rate": 1e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8541342616081238,
      "num_tokens": 61995463.0,
      "step": 1787
    },
    {
      "epoch": 0.332033426183844,
      "grad_norm": 1.4953812802455944,
      "learning_rate": 1e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8468610048294067,
      "num_tokens": 62035795.0,
      "step": 1788
    },
    {
      "epoch": 0.33221912720519964,
      "grad_norm": 1.5518597821948796,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8592699766159058,
      "num_tokens": 62067458.0,
      "step": 1789
    },
    {
      "epoch": 0.33240482822655526,
      "grad_norm": 1.5687757026678564,
      "learning_rate": 1e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8496376276016235,
      "num_tokens": 62099883.0,
      "step": 1790
    },
    {
      "epoch": 0.3325905292479109,
      "grad_norm": 1.7012836261864925,
      "learning_rate": 1e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8553026914596558,
      "num_tokens": 62128974.0,
      "step": 1791
    },
    {
      "epoch": 0.3327762302692665,
      "grad_norm": 1.6284442883246486,
      "learning_rate": 1e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8576686382293701,
      "num_tokens": 62158494.0,
      "step": 1792
    },
    {
      "epoch": 0.33296193129062207,
      "grad_norm": 1.4674803149764248,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8643211126327515,
      "num_tokens": 62194576.0,
      "step": 1793
    },
    {
      "epoch": 0.3331476323119777,
      "grad_norm": 1.6468424040149852,
      "learning_rate": 1e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8591891527175903,
      "num_tokens": 62224789.0,
      "step": 1794
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.3839571031693312,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8611998558044434,
      "num_tokens": 62266661.0,
      "step": 1795
    },
    {
      "epoch": 0.33351903435468894,
      "grad_norm": 1.412574592425294,
      "learning_rate": 1e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8526329398155212,
      "num_tokens": 62310516.0,
      "step": 1796
    },
    {
      "epoch": 0.33370473537604456,
      "grad_norm": 1.676146021180173,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8726444840431213,
      "num_tokens": 62339606.0,
      "step": 1797
    },
    {
      "epoch": 0.3338904363974002,
      "grad_norm": 1.559453008655605,
      "learning_rate": 1e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8619877696037292,
      "num_tokens": 62372336.0,
      "step": 1798
    },
    {
      "epoch": 0.3340761374187558,
      "grad_norm": 1.5802703254020642,
      "learning_rate": 1e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8438965082168579,
      "num_tokens": 62406343.0,
      "step": 1799
    },
    {
      "epoch": 0.3342618384401114,
      "grad_norm": 1.4705979930726092,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.864630937576294,
      "num_tokens": 62441327.0,
      "step": 1800
    },
    {
      "epoch": 0.33444753946146705,
      "grad_norm": 1.5799490038961066,
      "learning_rate": 1e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8499494194984436,
      "num_tokens": 62476499.0,
      "step": 1801
    },
    {
      "epoch": 0.33463324048282267,
      "grad_norm": 1.4413541966655756,
      "learning_rate": 1e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8512061238288879,
      "num_tokens": 62517656.0,
      "step": 1802
    },
    {
      "epoch": 0.3348189415041783,
      "grad_norm": 1.4529300384779387,
      "learning_rate": 1e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8561146855354309,
      "num_tokens": 62553639.0,
      "step": 1803
    },
    {
      "epoch": 0.3350046425255339,
      "grad_norm": 1.4403261427911977,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8718229532241821,
      "num_tokens": 62585811.0,
      "step": 1804
    },
    {
      "epoch": 0.33519034354688954,
      "grad_norm": 1.4202929725614792,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8689215183258057,
      "num_tokens": 62623181.0,
      "step": 1805
    },
    {
      "epoch": 0.3353760445682451,
      "grad_norm": 1.5029344485921863,
      "learning_rate": 1e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8412448167800903,
      "num_tokens": 62664778.0,
      "step": 1806
    },
    {
      "epoch": 0.3355617455896007,
      "grad_norm": 1.5472817823269247,
      "learning_rate": 1e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8450530767440796,
      "num_tokens": 62698440.0,
      "step": 1807
    },
    {
      "epoch": 0.33574744661095635,
      "grad_norm": 1.4162040070203037,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8698253035545349,
      "num_tokens": 62733577.0,
      "step": 1808
    },
    {
      "epoch": 0.33593314763231197,
      "grad_norm": 1.4363323431355954,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8691810369491577,
      "num_tokens": 62766441.0,
      "step": 1809
    },
    {
      "epoch": 0.3361188486536676,
      "grad_norm": 1.6413745530028274,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8688057661056519,
      "num_tokens": 62795290.0,
      "step": 1810
    },
    {
      "epoch": 0.3363045496750232,
      "grad_norm": 1.4468357228830513,
      "learning_rate": 1e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8607486486434937,
      "num_tokens": 62834149.0,
      "step": 1811
    },
    {
      "epoch": 0.33649025069637883,
      "grad_norm": 1.5726357744130663,
      "learning_rate": 1e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8604519963264465,
      "num_tokens": 62867655.0,
      "step": 1812
    },
    {
      "epoch": 0.33667595171773446,
      "grad_norm": 1.5655221250922764,
      "learning_rate": 1e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8632485866546631,
      "num_tokens": 62899008.0,
      "step": 1813
    },
    {
      "epoch": 0.3368616527390901,
      "grad_norm": 1.512862643721964,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8589059114456177,
      "num_tokens": 62934123.0,
      "step": 1814
    },
    {
      "epoch": 0.3370473537604457,
      "grad_norm": 1.4972705742585544,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8634209632873535,
      "num_tokens": 62966894.0,
      "step": 1815
    },
    {
      "epoch": 0.3372330547818013,
      "grad_norm": 1.5253900822773927,
      "learning_rate": 1e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8520104289054871,
      "num_tokens": 63000427.0,
      "step": 1816
    },
    {
      "epoch": 0.33741875580315694,
      "grad_norm": 1.5093291313314872,
      "learning_rate": 1e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8554364442825317,
      "num_tokens": 63036418.0,
      "step": 1817
    },
    {
      "epoch": 0.3376044568245125,
      "grad_norm": 1.4898127629321691,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8665525913238525,
      "num_tokens": 63075244.0,
      "step": 1818
    },
    {
      "epoch": 0.33779015784586813,
      "grad_norm": 1.5633157764357608,
      "learning_rate": 1e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8561228513717651,
      "num_tokens": 63109889.0,
      "step": 1819
    },
    {
      "epoch": 0.33797585886722376,
      "grad_norm": 1.6946301145876848,
      "learning_rate": 1e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8545499444007874,
      "num_tokens": 63136239.0,
      "step": 1820
    },
    {
      "epoch": 0.3381615598885794,
      "grad_norm": 1.6897541030866041,
      "learning_rate": 1e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8494831919670105,
      "num_tokens": 63165857.0,
      "step": 1821
    },
    {
      "epoch": 0.338347260909935,
      "grad_norm": 1.487878221495871,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8755590915679932,
      "num_tokens": 63196467.0,
      "step": 1822
    },
    {
      "epoch": 0.3385329619312906,
      "grad_norm": 1.6471767446630716,
      "learning_rate": 1e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8361078500747681,
      "num_tokens": 63227845.0,
      "step": 1823
    },
    {
      "epoch": 0.33871866295264624,
      "grad_norm": 1.477444954603375,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8626662492752075,
      "num_tokens": 63262502.0,
      "step": 1824
    },
    {
      "epoch": 0.33890436397400187,
      "grad_norm": 1.506610877926118,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8686971068382263,
      "num_tokens": 63296376.0,
      "step": 1825
    },
    {
      "epoch": 0.3390900649953575,
      "grad_norm": 1.535044106575925,
      "learning_rate": 1e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8617571592330933,
      "num_tokens": 63328915.0,
      "step": 1826
    },
    {
      "epoch": 0.3392757660167131,
      "grad_norm": 1.3681004596959916,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8663958311080933,
      "num_tokens": 63367522.0,
      "step": 1827
    },
    {
      "epoch": 0.33946146703806873,
      "grad_norm": 1.6055579027193827,
      "learning_rate": 1e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8548024892807007,
      "num_tokens": 63400867.0,
      "step": 1828
    },
    {
      "epoch": 0.33964716805942435,
      "grad_norm": 1.5900733274633185,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.873932957649231,
      "num_tokens": 63429937.0,
      "step": 1829
    },
    {
      "epoch": 0.3398328690807799,
      "grad_norm": 1.5511653649167925,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8611655235290527,
      "num_tokens": 63463640.0,
      "step": 1830
    },
    {
      "epoch": 0.34001857010213554,
      "grad_norm": 1.6270515042550284,
      "learning_rate": 1e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8603183031082153,
      "num_tokens": 63494431.0,
      "step": 1831
    },
    {
      "epoch": 0.34020427112349116,
      "grad_norm": 1.5354205438264668,
      "learning_rate": 1e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8540538549423218,
      "num_tokens": 63529485.0,
      "step": 1832
    },
    {
      "epoch": 0.3403899721448468,
      "grad_norm": 1.5633576187219371,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.868623673915863,
      "num_tokens": 63559757.0,
      "step": 1833
    },
    {
      "epoch": 0.3405756731662024,
      "grad_norm": 1.4672782437856877,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8602478504180908,
      "num_tokens": 63594970.0,
      "step": 1834
    },
    {
      "epoch": 0.34076137418755803,
      "grad_norm": 1.4856768067736505,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8619219660758972,
      "num_tokens": 63630624.0,
      "step": 1835
    },
    {
      "epoch": 0.34094707520891365,
      "grad_norm": 1.5606070908411243,
      "learning_rate": 1e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8609090447425842,
      "num_tokens": 63662919.0,
      "step": 1836
    },
    {
      "epoch": 0.3411327762302693,
      "grad_norm": 1.5827150580130385,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8693372011184692,
      "num_tokens": 63695736.0,
      "step": 1837
    },
    {
      "epoch": 0.3413184772516249,
      "grad_norm": 1.538173532377983,
      "learning_rate": 1e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8528634309768677,
      "num_tokens": 63733568.0,
      "step": 1838
    },
    {
      "epoch": 0.3415041782729805,
      "grad_norm": 1.462860594011132,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8714030981063843,
      "num_tokens": 63767095.0,
      "step": 1839
    },
    {
      "epoch": 0.34168987929433614,
      "grad_norm": 1.4915006568571383,
      "learning_rate": 1e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8512690663337708,
      "num_tokens": 63804287.0,
      "step": 1840
    },
    {
      "epoch": 0.34187558031569176,
      "grad_norm": 1.3472504332841357,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8656405210494995,
      "num_tokens": 63847533.0,
      "step": 1841
    },
    {
      "epoch": 0.34206128133704733,
      "grad_norm": 1.5795231261299292,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8785300254821777,
      "num_tokens": 63875980.0,
      "step": 1842
    },
    {
      "epoch": 0.34224698235840295,
      "grad_norm": 1.719006722975097,
      "learning_rate": 1e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8485653400421143,
      "num_tokens": 63909948.0,
      "step": 1843
    },
    {
      "epoch": 0.3424326833797586,
      "grad_norm": 1.5792610040651742,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8761098980903625,
      "num_tokens": 63937105.0,
      "step": 1844
    },
    {
      "epoch": 0.3426183844011142,
      "grad_norm": 1.6669242217589808,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8613342046737671,
      "num_tokens": 63965053.0,
      "step": 1845
    },
    {
      "epoch": 0.3428040854224698,
      "grad_norm": 1.5252263578172915,
      "learning_rate": 1e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.857556939125061,
      "num_tokens": 63999671.0,
      "step": 1846
    },
    {
      "epoch": 0.34298978644382544,
      "grad_norm": 1.6131935493291294,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8720046877861023,
      "num_tokens": 64029403.0,
      "step": 1847
    },
    {
      "epoch": 0.34317548746518106,
      "grad_norm": 1.5143885806533608,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8668398261070251,
      "num_tokens": 64062529.0,
      "step": 1848
    },
    {
      "epoch": 0.3433611884865367,
      "grad_norm": 1.4947699718823106,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8670788407325745,
      "num_tokens": 64095852.0,
      "step": 1849
    },
    {
      "epoch": 0.3435468895078923,
      "grad_norm": 1.4374138920660782,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8735170960426331,
      "num_tokens": 64129310.0,
      "step": 1850
    },
    {
      "epoch": 0.34373259052924793,
      "grad_norm": 1.6675719086735965,
      "learning_rate": 1e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8534989356994629,
      "num_tokens": 64160548.0,
      "step": 1851
    },
    {
      "epoch": 0.34391829155060355,
      "grad_norm": 1.5212008717438747,
      "learning_rate": 1e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.83829265832901,
      "num_tokens": 64197954.0,
      "step": 1852
    },
    {
      "epoch": 0.34410399257195917,
      "grad_norm": 1.488262784126476,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8638328909873962,
      "num_tokens": 64231582.0,
      "step": 1853
    },
    {
      "epoch": 0.34428969359331474,
      "grad_norm": 1.4237259457095846,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8628188371658325,
      "num_tokens": 64272019.0,
      "step": 1854
    },
    {
      "epoch": 0.34447539461467036,
      "grad_norm": 1.4974036021304218,
      "learning_rate": 1e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8589785099029541,
      "num_tokens": 64305132.0,
      "step": 1855
    },
    {
      "epoch": 0.344661095636026,
      "grad_norm": 1.5238044891578428,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8567896485328674,
      "num_tokens": 64340322.0,
      "step": 1856
    },
    {
      "epoch": 0.3448467966573816,
      "grad_norm": 1.4734280480605706,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8728227019309998,
      "num_tokens": 64376576.0,
      "step": 1857
    },
    {
      "epoch": 0.3450324976787372,
      "grad_norm": 1.4034579363427293,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8651474714279175,
      "num_tokens": 64417344.0,
      "step": 1858
    },
    {
      "epoch": 0.34521819870009285,
      "grad_norm": 1.5108246270688375,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.864690363407135,
      "num_tokens": 64452598.0,
      "step": 1859
    },
    {
      "epoch": 0.34540389972144847,
      "grad_norm": 1.5770283768435556,
      "learning_rate": 1e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8458939790725708,
      "num_tokens": 64486746.0,
      "step": 1860
    },
    {
      "epoch": 0.3455896007428041,
      "grad_norm": 1.53618311352884,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.859952986240387,
      "num_tokens": 64522037.0,
      "step": 1861
    },
    {
      "epoch": 0.3457753017641597,
      "grad_norm": 1.5069436248081582,
      "learning_rate": 1e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8556258082389832,
      "num_tokens": 64559113.0,
      "step": 1862
    },
    {
      "epoch": 0.34596100278551534,
      "grad_norm": 1.5228517474448415,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8584913015365601,
      "num_tokens": 64591665.0,
      "step": 1863
    },
    {
      "epoch": 0.34614670380687096,
      "grad_norm": 1.406115715868148,
      "learning_rate": 1e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8627187013626099,
      "num_tokens": 64627126.0,
      "step": 1864
    },
    {
      "epoch": 0.3463324048282266,
      "grad_norm": 1.6856427881241436,
      "learning_rate": 1e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.849029541015625,
      "num_tokens": 64654924.0,
      "step": 1865
    },
    {
      "epoch": 0.34651810584958215,
      "grad_norm": 1.752296455802253,
      "learning_rate": 1e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8401058912277222,
      "num_tokens": 64683768.0,
      "step": 1866
    },
    {
      "epoch": 0.34670380687093777,
      "grad_norm": 1.4876313880743064,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.873029351234436,
      "num_tokens": 64722630.0,
      "step": 1867
    },
    {
      "epoch": 0.3468895078922934,
      "grad_norm": 1.536892874474312,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8636020421981812,
      "num_tokens": 64756301.0,
      "step": 1868
    },
    {
      "epoch": 0.347075208913649,
      "grad_norm": 1.4865478106332481,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8679338097572327,
      "num_tokens": 64789654.0,
      "step": 1869
    },
    {
      "epoch": 0.34726090993500464,
      "grad_norm": 1.5698105663836326,
      "learning_rate": 1e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8608981370925903,
      "num_tokens": 64821405.0,
      "step": 1870
    },
    {
      "epoch": 0.34744661095636026,
      "grad_norm": 1.5062223878310579,
      "learning_rate": 1e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8670027852058411,
      "num_tokens": 64854959.0,
      "step": 1871
    },
    {
      "epoch": 0.3476323119777159,
      "grad_norm": 1.501693764820395,
      "learning_rate": 1e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8590764999389648,
      "num_tokens": 64893838.0,
      "step": 1872
    },
    {
      "epoch": 0.3478180129990715,
      "grad_norm": 1.504416207911299,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8598209619522095,
      "num_tokens": 64931664.0,
      "step": 1873
    },
    {
      "epoch": 0.3480037140204271,
      "grad_norm": 1.5263350654599663,
      "learning_rate": 1e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8420110940933228,
      "num_tokens": 64968856.0,
      "step": 1874
    },
    {
      "epoch": 0.34818941504178275,
      "grad_norm": 1.5687806906758266,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.863471269607544,
      "num_tokens": 64999313.0,
      "step": 1875
    },
    {
      "epoch": 0.34837511606313837,
      "grad_norm": 1.6052513320504866,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.858413815498352,
      "num_tokens": 65032872.0,
      "step": 1876
    },
    {
      "epoch": 0.348560817084494,
      "grad_norm": 1.58260167904773,
      "learning_rate": 1e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8609799146652222,
      "num_tokens": 65064488.0,
      "step": 1877
    },
    {
      "epoch": 0.34874651810584956,
      "grad_norm": 1.5876488430677813,
      "learning_rate": 1e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8452761173248291,
      "num_tokens": 65098924.0,
      "step": 1878
    },
    {
      "epoch": 0.3489322191272052,
      "grad_norm": 1.4528262091739284,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8606854677200317,
      "num_tokens": 65136183.0,
      "step": 1879
    },
    {
      "epoch": 0.3491179201485608,
      "grad_norm": 1.5161895568279762,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.870123028755188,
      "num_tokens": 65174868.0,
      "step": 1880
    },
    {
      "epoch": 0.3493036211699164,
      "grad_norm": 1.702901220925831,
      "learning_rate": 1e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.863071858882904,
      "num_tokens": 65203337.0,
      "step": 1881
    },
    {
      "epoch": 0.34948932219127204,
      "grad_norm": 1.4651000480214094,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8694436550140381,
      "num_tokens": 65236887.0,
      "step": 1882
    },
    {
      "epoch": 0.34967502321262767,
      "grad_norm": 1.5510996401076016,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8646535873413086,
      "num_tokens": 65269194.0,
      "step": 1883
    },
    {
      "epoch": 0.3498607242339833,
      "grad_norm": 1.4682319022986936,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8611850738525391,
      "num_tokens": 65302515.0,
      "step": 1884
    },
    {
      "epoch": 0.3500464252553389,
      "grad_norm": 1.6830049258192796,
      "learning_rate": 1e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8376344442367554,
      "num_tokens": 65336344.0,
      "step": 1885
    },
    {
      "epoch": 0.35023212627669453,
      "grad_norm": 1.5023286174647896,
      "learning_rate": 1e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8557727336883545,
      "num_tokens": 65370462.0,
      "step": 1886
    },
    {
      "epoch": 0.35041782729805016,
      "grad_norm": 1.5087383713447302,
      "learning_rate": 1e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8589392900466919,
      "num_tokens": 65406107.0,
      "step": 1887
    },
    {
      "epoch": 0.3506035283194058,
      "grad_norm": 1.5498660783122418,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8753592371940613,
      "num_tokens": 65436826.0,
      "step": 1888
    },
    {
      "epoch": 0.3507892293407614,
      "grad_norm": 1.645242410408465,
      "learning_rate": 1e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8458642959594727,
      "num_tokens": 65471519.0,
      "step": 1889
    },
    {
      "epoch": 0.35097493036211697,
      "grad_norm": 1.445605346721591,
      "learning_rate": 1e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8569011688232422,
      "num_tokens": 65509658.0,
      "step": 1890
    },
    {
      "epoch": 0.3511606313834726,
      "grad_norm": 1.4395373277658894,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8684258460998535,
      "num_tokens": 65546899.0,
      "step": 1891
    },
    {
      "epoch": 0.3513463324048282,
      "grad_norm": 1.4426173883987419,
      "learning_rate": 1e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8548211455345154,
      "num_tokens": 65583897.0,
      "step": 1892
    },
    {
      "epoch": 0.35153203342618383,
      "grad_norm": 1.5636668225802677,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8639602661132812,
      "num_tokens": 65617102.0,
      "step": 1893
    },
    {
      "epoch": 0.35171773444753945,
      "grad_norm": 1.6181726434974888,
      "learning_rate": 1e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8609175682067871,
      "num_tokens": 65650661.0,
      "step": 1894
    },
    {
      "epoch": 0.3519034354688951,
      "grad_norm": 1.4741803247135714,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8642686605453491,
      "num_tokens": 65686019.0,
      "step": 1895
    },
    {
      "epoch": 0.3520891364902507,
      "grad_norm": 1.5614302490625815,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8671882152557373,
      "num_tokens": 65718185.0,
      "step": 1896
    },
    {
      "epoch": 0.3522748375116063,
      "grad_norm": 1.5994925068611587,
      "learning_rate": 1e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8454656004905701,
      "num_tokens": 65749388.0,
      "step": 1897
    },
    {
      "epoch": 0.35246053853296194,
      "grad_norm": 1.4355078118178466,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8607968091964722,
      "num_tokens": 65786302.0,
      "step": 1898
    },
    {
      "epoch": 0.35264623955431756,
      "grad_norm": 1.672116110726292,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8611421585083008,
      "num_tokens": 65816835.0,
      "step": 1899
    },
    {
      "epoch": 0.3528319405756732,
      "grad_norm": 1.533929889072848,
      "learning_rate": 1e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8459467887878418,
      "num_tokens": 65851408.0,
      "step": 1900
    },
    {
      "epoch": 0.3530176415970288,
      "grad_norm": 1.5218654955983202,
      "learning_rate": 1e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8511584997177124,
      "num_tokens": 65888883.0,
      "step": 1901
    },
    {
      "epoch": 0.3532033426183844,
      "grad_norm": 1.4589555939281913,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8706613779067993,
      "num_tokens": 65926322.0,
      "step": 1902
    },
    {
      "epoch": 0.35338904363974,
      "grad_norm": 1.4670984186555085,
      "learning_rate": 1e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8459640741348267,
      "num_tokens": 65964754.0,
      "step": 1903
    },
    {
      "epoch": 0.3535747446610956,
      "grad_norm": 1.5216983216104993,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8827341794967651,
      "num_tokens": 65997170.0,
      "step": 1904
    },
    {
      "epoch": 0.35376044568245124,
      "grad_norm": 1.6523893901936593,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.858406126499176,
      "num_tokens": 66028316.0,
      "step": 1905
    },
    {
      "epoch": 0.35394614670380686,
      "grad_norm": 1.450224517575967,
      "learning_rate": 1e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.866620659828186,
      "num_tokens": 66066209.0,
      "step": 1906
    },
    {
      "epoch": 0.3541318477251625,
      "grad_norm": 1.4991408739828391,
      "learning_rate": 1e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8516289591789246,
      "num_tokens": 66099689.0,
      "step": 1907
    },
    {
      "epoch": 0.3543175487465181,
      "grad_norm": 1.3590579176668134,
      "learning_rate": 1e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8534443378448486,
      "num_tokens": 66144478.0,
      "step": 1908
    },
    {
      "epoch": 0.35450324976787373,
      "grad_norm": 1.4290469380517494,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8609378933906555,
      "num_tokens": 66181117.0,
      "step": 1909
    },
    {
      "epoch": 0.35468895078922935,
      "grad_norm": 1.3860212727878427,
      "learning_rate": 1e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8479882478713989,
      "num_tokens": 66221580.0,
      "step": 1910
    },
    {
      "epoch": 0.354874651810585,
      "grad_norm": 1.4813660257737598,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8741436004638672,
      "num_tokens": 66254545.0,
      "step": 1911
    },
    {
      "epoch": 0.3550603528319406,
      "grad_norm": 1.4901566786975766,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8664693832397461,
      "num_tokens": 66288696.0,
      "step": 1912
    },
    {
      "epoch": 0.3552460538532962,
      "grad_norm": 1.5117770797174945,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8755765557289124,
      "num_tokens": 66322598.0,
      "step": 1913
    },
    {
      "epoch": 0.3554317548746518,
      "grad_norm": 1.508636659402876,
      "learning_rate": 1e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8548866510391235,
      "num_tokens": 66359366.0,
      "step": 1914
    },
    {
      "epoch": 0.3556174558960074,
      "grad_norm": 1.4328127493102043,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8560423851013184,
      "num_tokens": 66397072.0,
      "step": 1915
    },
    {
      "epoch": 0.35580315691736303,
      "grad_norm": 1.5646072750544218,
      "learning_rate": 1e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8607049584388733,
      "num_tokens": 66432067.0,
      "step": 1916
    },
    {
      "epoch": 0.35598885793871865,
      "grad_norm": 1.6720730801369443,
      "learning_rate": 1e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8507245182991028,
      "num_tokens": 66461245.0,
      "step": 1917
    },
    {
      "epoch": 0.3561745589600743,
      "grad_norm": 1.6460421936715741,
      "learning_rate": 1e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8506067991256714,
      "num_tokens": 66499024.0,
      "step": 1918
    },
    {
      "epoch": 0.3563602599814299,
      "grad_norm": 1.6170230997228667,
      "learning_rate": 1e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8479368686676025,
      "num_tokens": 66535361.0,
      "step": 1919
    },
    {
      "epoch": 0.3565459610027855,
      "grad_norm": 1.4409089631835945,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8690575361251831,
      "num_tokens": 66570597.0,
      "step": 1920
    },
    {
      "epoch": 0.35673166202414114,
      "grad_norm": 1.5339901139519618,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.860503613948822,
      "num_tokens": 66602357.0,
      "step": 1921
    },
    {
      "epoch": 0.35691736304549676,
      "grad_norm": 1.5591193664923526,
      "learning_rate": 1e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8500638604164124,
      "num_tokens": 66636412.0,
      "step": 1922
    },
    {
      "epoch": 0.3571030640668524,
      "grad_norm": 1.4503204237865337,
      "learning_rate": 1e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8596457242965698,
      "num_tokens": 66676998.0,
      "step": 1923
    },
    {
      "epoch": 0.357288765088208,
      "grad_norm": 1.5206407258667551,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8674516677856445,
      "num_tokens": 66708655.0,
      "step": 1924
    },
    {
      "epoch": 0.3574744661095636,
      "grad_norm": 1.6077584116397172,
      "learning_rate": 1e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8493467569351196,
      "num_tokens": 66738378.0,
      "step": 1925
    },
    {
      "epoch": 0.3576601671309192,
      "grad_norm": 1.5344693154596662,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8598315715789795,
      "num_tokens": 66770284.0,
      "step": 1926
    },
    {
      "epoch": 0.3578458681522748,
      "grad_norm": 1.54191692456389,
      "learning_rate": 1e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8573063015937805,
      "num_tokens": 66808613.0,
      "step": 1927
    },
    {
      "epoch": 0.35803156917363044,
      "grad_norm": 1.508519843223824,
      "learning_rate": 1e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8528707027435303,
      "num_tokens": 66841472.0,
      "step": 1928
    },
    {
      "epoch": 0.35821727019498606,
      "grad_norm": 1.534049374082454,
      "learning_rate": 1e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8577075004577637,
      "num_tokens": 66876749.0,
      "step": 1929
    },
    {
      "epoch": 0.3584029712163417,
      "grad_norm": 1.5120559645709877,
      "learning_rate": 1e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8635390400886536,
      "num_tokens": 66911635.0,
      "step": 1930
    },
    {
      "epoch": 0.3585886722376973,
      "grad_norm": 1.4216514915376068,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8657311201095581,
      "num_tokens": 66946842.0,
      "step": 1931
    },
    {
      "epoch": 0.3587743732590529,
      "grad_norm": 1.5125269835868362,
      "learning_rate": 1e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8613051772117615,
      "num_tokens": 66980165.0,
      "step": 1932
    },
    {
      "epoch": 0.35896007428040855,
      "grad_norm": 1.5097683922033474,
      "learning_rate": 1e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8627424240112305,
      "num_tokens": 67017556.0,
      "step": 1933
    },
    {
      "epoch": 0.35914577530176417,
      "grad_norm": 1.5250157064343315,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8617303967475891,
      "num_tokens": 67051766.0,
      "step": 1934
    },
    {
      "epoch": 0.3593314763231198,
      "grad_norm": 1.545332797004706,
      "learning_rate": 1e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8543362617492676,
      "num_tokens": 67085664.0,
      "step": 1935
    },
    {
      "epoch": 0.3595171773444754,
      "grad_norm": 1.6267778761529874,
      "learning_rate": 1e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8546101450920105,
      "num_tokens": 67117773.0,
      "step": 1936
    },
    {
      "epoch": 0.35970287836583104,
      "grad_norm": 1.5594549369445259,
      "learning_rate": 1e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8507431149482727,
      "num_tokens": 67152629.0,
      "step": 1937
    },
    {
      "epoch": 0.3598885793871866,
      "grad_norm": 1.442603686989229,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8597894906997681,
      "num_tokens": 67189118.0,
      "step": 1938
    },
    {
      "epoch": 0.3600742804085422,
      "grad_norm": 1.4790603877586666,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8615620136260986,
      "num_tokens": 67223833.0,
      "step": 1939
    },
    {
      "epoch": 0.36025998142989785,
      "grad_norm": 1.55243242336477,
      "learning_rate": 1e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8576771020889282,
      "num_tokens": 67258704.0,
      "step": 1940
    },
    {
      "epoch": 0.36044568245125347,
      "grad_norm": 1.5768316719537074,
      "learning_rate": 1e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8595570921897888,
      "num_tokens": 67296996.0,
      "step": 1941
    },
    {
      "epoch": 0.3606313834726091,
      "grad_norm": 1.4742333372077066,
      "learning_rate": 1e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8542954921722412,
      "num_tokens": 67332274.0,
      "step": 1942
    },
    {
      "epoch": 0.3608170844939647,
      "grad_norm": 1.5547189455061776,
      "learning_rate": 1e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.854432225227356,
      "num_tokens": 67362848.0,
      "step": 1943
    },
    {
      "epoch": 0.36100278551532033,
      "grad_norm": 1.5667103532681401,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8738542795181274,
      "num_tokens": 67394513.0,
      "step": 1944
    },
    {
      "epoch": 0.36118848653667596,
      "grad_norm": 1.4667439663186495,
      "learning_rate": 1e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8486608266830444,
      "num_tokens": 67432735.0,
      "step": 1945
    },
    {
      "epoch": 0.3613741875580316,
      "grad_norm": 1.5108770192803582,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8674870133399963,
      "num_tokens": 67465806.0,
      "step": 1946
    },
    {
      "epoch": 0.3615598885793872,
      "grad_norm": 1.468975326562348,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8587355613708496,
      "num_tokens": 67503538.0,
      "step": 1947
    },
    {
      "epoch": 0.3617455896007428,
      "grad_norm": 1.4913379952482855,
      "learning_rate": 1e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8657535314559937,
      "num_tokens": 67542164.0,
      "step": 1948
    },
    {
      "epoch": 0.36193129062209844,
      "grad_norm": 1.642095340584203,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8678771257400513,
      "num_tokens": 67568140.0,
      "step": 1949
    },
    {
      "epoch": 0.362116991643454,
      "grad_norm": 1.5129318210799036,
      "learning_rate": 1e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8480233550071716,
      "num_tokens": 67604545.0,
      "step": 1950
    },
    {
      "epoch": 0.36230269266480963,
      "grad_norm": 1.4784183160837758,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.867009162902832,
      "num_tokens": 67636553.0,
      "step": 1951
    },
    {
      "epoch": 0.36248839368616526,
      "grad_norm": 1.4774416586822594,
      "learning_rate": 1e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8591336607933044,
      "num_tokens": 67672250.0,
      "step": 1952
    },
    {
      "epoch": 0.3626740947075209,
      "grad_norm": 1.5720606037430456,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8703839182853699,
      "num_tokens": 67711764.0,
      "step": 1953
    },
    {
      "epoch": 0.3628597957288765,
      "grad_norm": 1.4528242195559857,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8685448169708252,
      "num_tokens": 67745876.0,
      "step": 1954
    },
    {
      "epoch": 0.3630454967502321,
      "grad_norm": 1.4349567536092522,
      "learning_rate": 1e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8630783557891846,
      "num_tokens": 67779574.0,
      "step": 1955
    },
    {
      "epoch": 0.36323119777158774,
      "grad_norm": 1.432834257921753,
      "learning_rate": 1e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8610080480575562,
      "num_tokens": 67819233.0,
      "step": 1956
    },
    {
      "epoch": 0.36341689879294337,
      "grad_norm": 1.4144020545664695,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8678647875785828,
      "num_tokens": 67854859.0,
      "step": 1957
    },
    {
      "epoch": 0.363602599814299,
      "grad_norm": 1.546400016999082,
      "learning_rate": 1e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8553446531295776,
      "num_tokens": 67890900.0,
      "step": 1958
    },
    {
      "epoch": 0.3637883008356546,
      "grad_norm": 1.587346063299365,
      "learning_rate": 1e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.860655665397644,
      "num_tokens": 67924589.0,
      "step": 1959
    },
    {
      "epoch": 0.36397400185701023,
      "grad_norm": 1.4713777784083142,
      "learning_rate": 1e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8602368235588074,
      "num_tokens": 67962186.0,
      "step": 1960
    },
    {
      "epoch": 0.36415970287836585,
      "grad_norm": 1.6488592037739813,
      "learning_rate": 1e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8513364791870117,
      "num_tokens": 67993569.0,
      "step": 1961
    },
    {
      "epoch": 0.3643454038997215,
      "grad_norm": 1.3412884045765296,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8695284724235535,
      "num_tokens": 68033693.0,
      "step": 1962
    },
    {
      "epoch": 0.36453110492107704,
      "grad_norm": 1.4805292213315209,
      "learning_rate": 1e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8572432994842529,
      "num_tokens": 68069115.0,
      "step": 1963
    },
    {
      "epoch": 0.36471680594243266,
      "grad_norm": 1.6415325654369681,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8614627718925476,
      "num_tokens": 68099287.0,
      "step": 1964
    },
    {
      "epoch": 0.3649025069637883,
      "grad_norm": 1.3647619837258251,
      "learning_rate": 1e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8641815185546875,
      "num_tokens": 68139558.0,
      "step": 1965
    },
    {
      "epoch": 0.3650882079851439,
      "grad_norm": 1.5273288328372687,
      "learning_rate": 1e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.858208954334259,
      "num_tokens": 68172494.0,
      "step": 1966
    },
    {
      "epoch": 0.36527390900649953,
      "grad_norm": 1.5605506735729289,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8608896136283875,
      "num_tokens": 68205056.0,
      "step": 1967
    },
    {
      "epoch": 0.36545961002785515,
      "grad_norm": 1.4851768344400427,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8643618226051331,
      "num_tokens": 68239878.0,
      "step": 1968
    },
    {
      "epoch": 0.3656453110492108,
      "grad_norm": 1.5555293200131761,
      "learning_rate": 1e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8522218465805054,
      "num_tokens": 68274815.0,
      "step": 1969
    },
    {
      "epoch": 0.3658310120705664,
      "grad_norm": 1.5233043536848274,
      "learning_rate": 1e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8509641885757446,
      "num_tokens": 68311761.0,
      "step": 1970
    },
    {
      "epoch": 0.366016713091922,
      "grad_norm": 1.51807804809745,
      "learning_rate": 1e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8490865230560303,
      "num_tokens": 68350031.0,
      "step": 1971
    },
    {
      "epoch": 0.36620241411327764,
      "grad_norm": 1.5799162579052104,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8642723560333252,
      "num_tokens": 68380858.0,
      "step": 1972
    },
    {
      "epoch": 0.36638811513463326,
      "grad_norm": 1.5100189191309312,
      "learning_rate": 1e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8525334000587463,
      "num_tokens": 68415353.0,
      "step": 1973
    },
    {
      "epoch": 0.3665738161559889,
      "grad_norm": 1.4734456794797874,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8652558326721191,
      "num_tokens": 68448968.0,
      "step": 1974
    },
    {
      "epoch": 0.36675951717734445,
      "grad_norm": 1.4370156946897301,
      "learning_rate": 1e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.862595796585083,
      "num_tokens": 68485717.0,
      "step": 1975
    },
    {
      "epoch": 0.3669452181987001,
      "grad_norm": 1.4733724782288014,
      "learning_rate": 1e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8606289029121399,
      "num_tokens": 68519216.0,
      "step": 1976
    },
    {
      "epoch": 0.3671309192200557,
      "grad_norm": 1.460693534167245,
      "learning_rate": 1e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8521738648414612,
      "num_tokens": 68559437.0,
      "step": 1977
    },
    {
      "epoch": 0.3673166202414113,
      "grad_norm": 1.3896903555806694,
      "learning_rate": 1e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8510003089904785,
      "num_tokens": 68601784.0,
      "step": 1978
    },
    {
      "epoch": 0.36750232126276694,
      "grad_norm": 1.5275283866675444,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8646669387817383,
      "num_tokens": 68635025.0,
      "step": 1979
    },
    {
      "epoch": 0.36768802228412256,
      "grad_norm": 1.4887037278021371,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8693980574607849,
      "num_tokens": 68670799.0,
      "step": 1980
    },
    {
      "epoch": 0.3678737233054782,
      "grad_norm": 1.5393985895654123,
      "learning_rate": 1e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8601716160774231,
      "num_tokens": 68702197.0,
      "step": 1981
    },
    {
      "epoch": 0.3680594243268338,
      "grad_norm": 1.6173071534873784,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8572320938110352,
      "num_tokens": 68734675.0,
      "step": 1982
    },
    {
      "epoch": 0.36824512534818943,
      "grad_norm": 1.448655849184574,
      "learning_rate": 1e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8574709892272949,
      "num_tokens": 68770286.0,
      "step": 1983
    },
    {
      "epoch": 0.36843082636954505,
      "grad_norm": 1.6303680238430194,
      "learning_rate": 1e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8489061594009399,
      "num_tokens": 68801942.0,
      "step": 1984
    },
    {
      "epoch": 0.3686165273909007,
      "grad_norm": 1.5239419854644058,
      "learning_rate": 1e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8556252717971802,
      "num_tokens": 68837440.0,
      "step": 1985
    },
    {
      "epoch": 0.3688022284122563,
      "grad_norm": 1.5666452064899805,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8745930790901184,
      "num_tokens": 68867898.0,
      "step": 1986
    },
    {
      "epoch": 0.36898792943361186,
      "grad_norm": 1.646794434651515,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8612595200538635,
      "num_tokens": 68899611.0,
      "step": 1987
    },
    {
      "epoch": 0.3691736304549675,
      "grad_norm": 1.4107310166968245,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.863642692565918,
      "num_tokens": 68938774.0,
      "step": 1988
    },
    {
      "epoch": 0.3693593314763231,
      "grad_norm": 1.5245109476474228,
      "learning_rate": 1e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8518821001052856,
      "num_tokens": 68975615.0,
      "step": 1989
    },
    {
      "epoch": 0.3695450324976787,
      "grad_norm": 1.3360149716195426,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8650370836257935,
      "num_tokens": 69018134.0,
      "step": 1990
    },
    {
      "epoch": 0.36973073351903435,
      "grad_norm": 1.605260097841614,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8676824569702148,
      "num_tokens": 69051693.0,
      "step": 1991
    },
    {
      "epoch": 0.36991643454038997,
      "grad_norm": 1.5497906857862747,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8619892001152039,
      "num_tokens": 69086459.0,
      "step": 1992
    },
    {
      "epoch": 0.3701021355617456,
      "grad_norm": 1.4255891495559914,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8768459558486938,
      "num_tokens": 69120556.0,
      "step": 1993
    },
    {
      "epoch": 0.3702878365831012,
      "grad_norm": 1.45030144618173,
      "learning_rate": 1e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8615844249725342,
      "num_tokens": 69155860.0,
      "step": 1994
    },
    {
      "epoch": 0.37047353760445684,
      "grad_norm": 1.5160044645566513,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.859633207321167,
      "num_tokens": 69192910.0,
      "step": 1995
    },
    {
      "epoch": 0.37065923862581246,
      "grad_norm": 1.555152535824148,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8637605905532837,
      "num_tokens": 69226558.0,
      "step": 1996
    },
    {
      "epoch": 0.3708449396471681,
      "grad_norm": 1.4511474174475265,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8709398508071899,
      "num_tokens": 69262529.0,
      "step": 1997
    },
    {
      "epoch": 0.3710306406685237,
      "grad_norm": 1.4978735719899818,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8596773743629456,
      "num_tokens": 69299410.0,
      "step": 1998
    },
    {
      "epoch": 0.37121634168987927,
      "grad_norm": 1.5329904094800133,
      "learning_rate": 1e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8547014594078064,
      "num_tokens": 69332200.0,
      "step": 1999
    },
    {
      "epoch": 0.3714020427112349,
      "grad_norm": 1.4005379048009858,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8690449595451355,
      "num_tokens": 69369751.0,
      "step": 2000
    },
    {
      "epoch": 0.3715877437325905,
      "grad_norm": 1.7095312773675277,
      "learning_rate": 1e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.858971357345581,
      "num_tokens": 69395928.0,
      "step": 2001
    },
    {
      "epoch": 0.37177344475394614,
      "grad_norm": 1.4345544290544354,
      "learning_rate": 1e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8493624329566956,
      "num_tokens": 69437908.0,
      "step": 2002
    },
    {
      "epoch": 0.37195914577530176,
      "grad_norm": 1.4739174432306479,
      "learning_rate": 1e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8587502241134644,
      "num_tokens": 69477511.0,
      "step": 2003
    },
    {
      "epoch": 0.3721448467966574,
      "grad_norm": 1.438232274564262,
      "learning_rate": 1e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8559175729751587,
      "num_tokens": 69512367.0,
      "step": 2004
    },
    {
      "epoch": 0.372330547818013,
      "grad_norm": 1.541753538423453,
      "learning_rate": 1e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.860893964767456,
      "num_tokens": 69540763.0,
      "step": 2005
    },
    {
      "epoch": 0.3725162488393686,
      "grad_norm": 1.7839606709738678,
      "learning_rate": 1e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8642531037330627,
      "num_tokens": 69565075.0,
      "step": 2006
    },
    {
      "epoch": 0.37270194986072425,
      "grad_norm": 1.5775695801281688,
      "learning_rate": 1e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.853643536567688,
      "num_tokens": 69599296.0,
      "step": 2007
    },
    {
      "epoch": 0.37288765088207987,
      "grad_norm": 1.5110701776539557,
      "learning_rate": 1e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8520143032073975,
      "num_tokens": 69637277.0,
      "step": 2008
    },
    {
      "epoch": 0.3730733519034355,
      "grad_norm": 1.4354948322256358,
      "learning_rate": 1e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8524101376533508,
      "num_tokens": 69677082.0,
      "step": 2009
    },
    {
      "epoch": 0.3732590529247911,
      "grad_norm": 1.7519936926575437,
      "learning_rate": 1e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8428770899772644,
      "num_tokens": 69705522.0,
      "step": 2010
    },
    {
      "epoch": 0.3734447539461467,
      "grad_norm": 1.685668289978633,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8679508566856384,
      "num_tokens": 69737098.0,
      "step": 2011
    },
    {
      "epoch": 0.3736304549675023,
      "grad_norm": 1.3476816886518403,
      "learning_rate": 1e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8516567945480347,
      "num_tokens": 69781976.0,
      "step": 2012
    },
    {
      "epoch": 0.3738161559888579,
      "grad_norm": 1.556779627768275,
      "learning_rate": 1e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8609662055969238,
      "num_tokens": 69817609.0,
      "step": 2013
    },
    {
      "epoch": 0.37400185701021355,
      "grad_norm": 1.5731911801540432,
      "learning_rate": 1e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8622894287109375,
      "num_tokens": 69850934.0,
      "step": 2014
    },
    {
      "epoch": 0.37418755803156917,
      "grad_norm": 1.5161176461752528,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8622868061065674,
      "num_tokens": 69883068.0,
      "step": 2015
    },
    {
      "epoch": 0.3743732590529248,
      "grad_norm": 1.445405338655555,
      "learning_rate": 1e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8563199639320374,
      "num_tokens": 69921986.0,
      "step": 2016
    },
    {
      "epoch": 0.3745589600742804,
      "grad_norm": 1.517614658068515,
      "learning_rate": 1e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8326046466827393,
      "num_tokens": 69960028.0,
      "step": 2017
    },
    {
      "epoch": 0.37474466109563603,
      "grad_norm": 1.497385165936578,
      "learning_rate": 1e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8485281467437744,
      "num_tokens": 69995637.0,
      "step": 2018
    },
    {
      "epoch": 0.37493036211699166,
      "grad_norm": 1.6065292839977225,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8669864535331726,
      "num_tokens": 70025820.0,
      "step": 2019
    },
    {
      "epoch": 0.3751160631383473,
      "grad_norm": 1.650778697880877,
      "learning_rate": 1e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8448936343193054,
      "num_tokens": 70058626.0,
      "step": 2020
    },
    {
      "epoch": 0.3753017641597029,
      "grad_norm": 1.4774159835985168,
      "learning_rate": 1e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8459995985031128,
      "num_tokens": 70094587.0,
      "step": 2021
    },
    {
      "epoch": 0.3754874651810585,
      "grad_norm": 1.556236190680057,
      "learning_rate": 1e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8590713143348694,
      "num_tokens": 70132376.0,
      "step": 2022
    },
    {
      "epoch": 0.3756731662024141,
      "grad_norm": 1.5116008903901303,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8669487237930298,
      "num_tokens": 70165815.0,
      "step": 2023
    },
    {
      "epoch": 0.3758588672237697,
      "grad_norm": 1.5542498073797895,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8626301288604736,
      "num_tokens": 70198474.0,
      "step": 2024
    },
    {
      "epoch": 0.37604456824512533,
      "grad_norm": 1.7595784867765931,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8593508005142212,
      "num_tokens": 70224673.0,
      "step": 2025
    },
    {
      "epoch": 0.37623026926648095,
      "grad_norm": 1.53335235811408,
      "learning_rate": 1e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8540221452713013,
      "num_tokens": 70259227.0,
      "step": 2026
    },
    {
      "epoch": 0.3764159702878366,
      "grad_norm": 1.5074743412662812,
      "learning_rate": 1e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8622738122940063,
      "num_tokens": 70291722.0,
      "step": 2027
    },
    {
      "epoch": 0.3766016713091922,
      "grad_norm": 1.4149775890306449,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8596547245979309,
      "num_tokens": 70328023.0,
      "step": 2028
    },
    {
      "epoch": 0.3767873723305478,
      "grad_norm": 1.5876308191351198,
      "learning_rate": 1e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8634160757064819,
      "num_tokens": 70364901.0,
      "step": 2029
    },
    {
      "epoch": 0.37697307335190344,
      "grad_norm": 1.4799134787013146,
      "learning_rate": 1e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8512575626373291,
      "num_tokens": 70402627.0,
      "step": 2030
    },
    {
      "epoch": 0.37715877437325906,
      "grad_norm": 1.5026945769645972,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.859723687171936,
      "num_tokens": 70440360.0,
      "step": 2031
    },
    {
      "epoch": 0.3773444753946147,
      "grad_norm": 1.4508768518897406,
      "learning_rate": 1e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8463541865348816,
      "num_tokens": 70481310.0,
      "step": 2032
    },
    {
      "epoch": 0.3775301764159703,
      "grad_norm": 1.4505244654215927,
      "learning_rate": 1e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8504592776298523,
      "num_tokens": 70516784.0,
      "step": 2033
    },
    {
      "epoch": 0.37771587743732593,
      "grad_norm": 1.5455441932943783,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8599622249603271,
      "num_tokens": 70551174.0,
      "step": 2034
    },
    {
      "epoch": 0.3779015784586815,
      "grad_norm": 1.3829694589996044,
      "learning_rate": 1e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.848962128162384,
      "num_tokens": 70594417.0,
      "step": 2035
    },
    {
      "epoch": 0.3780872794800371,
      "grad_norm": 1.4626175395177032,
      "learning_rate": 1e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8628008961677551,
      "num_tokens": 70629296.0,
      "step": 2036
    },
    {
      "epoch": 0.37827298050139274,
      "grad_norm": 1.4111524489357479,
      "learning_rate": 1e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8570605516433716,
      "num_tokens": 70667819.0,
      "step": 2037
    },
    {
      "epoch": 0.37845868152274836,
      "grad_norm": 1.6255863511201294,
      "learning_rate": 1e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8530609607696533,
      "num_tokens": 70699695.0,
      "step": 2038
    },
    {
      "epoch": 0.378644382544104,
      "grad_norm": 1.509683482353571,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8614813089370728,
      "num_tokens": 70733913.0,
      "step": 2039
    },
    {
      "epoch": 0.3788300835654596,
      "grad_norm": 1.4641228695221362,
      "learning_rate": 1e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8533217906951904,
      "num_tokens": 70768993.0,
      "step": 2040
    },
    {
      "epoch": 0.37901578458681523,
      "grad_norm": 1.5856022719504024,
      "learning_rate": 1e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8698316812515259,
      "num_tokens": 70802730.0,
      "step": 2041
    },
    {
      "epoch": 0.37920148560817085,
      "grad_norm": 1.5481674683368154,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8792680501937866,
      "num_tokens": 70834054.0,
      "step": 2042
    },
    {
      "epoch": 0.3793871866295265,
      "grad_norm": 1.4268754493997902,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8715473413467407,
      "num_tokens": 70870707.0,
      "step": 2043
    },
    {
      "epoch": 0.3795728876508821,
      "grad_norm": 1.5128005993387512,
      "learning_rate": 1e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8555088043212891,
      "num_tokens": 70908616.0,
      "step": 2044
    },
    {
      "epoch": 0.3797585886722377,
      "grad_norm": 1.5885648500459473,
      "learning_rate": 1e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8500330448150635,
      "num_tokens": 70943237.0,
      "step": 2045
    },
    {
      "epoch": 0.37994428969359334,
      "grad_norm": 1.5667449649345984,
      "learning_rate": 1e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8586376309394836,
      "num_tokens": 70980083.0,
      "step": 2046
    },
    {
      "epoch": 0.3801299907149489,
      "grad_norm": 1.5806539566026077,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.858057975769043,
      "num_tokens": 71011518.0,
      "step": 2047
    },
    {
      "epoch": 0.38031569173630453,
      "grad_norm": 1.4953583816575413,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8639364242553711,
      "num_tokens": 71046698.0,
      "step": 2048
    },
    {
      "epoch": 0.38050139275766015,
      "grad_norm": 1.4962858526982643,
      "learning_rate": 1e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8458989262580872,
      "num_tokens": 71086729.0,
      "step": 2049
    },
    {
      "epoch": 0.3806870937790158,
      "grad_norm": 1.3893090166231707,
      "learning_rate": 1e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8610232472419739,
      "num_tokens": 71127880.0,
      "step": 2050
    },
    {
      "epoch": 0.3808727948003714,
      "grad_norm": 1.6940730852510086,
      "learning_rate": 1e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8463491201400757,
      "num_tokens": 71161242.0,
      "step": 2051
    },
    {
      "epoch": 0.381058495821727,
      "grad_norm": 1.5835341530252394,
      "learning_rate": 1e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8528894186019897,
      "num_tokens": 71199014.0,
      "step": 2052
    },
    {
      "epoch": 0.38124419684308264,
      "grad_norm": 1.517084303031573,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8685739040374756,
      "num_tokens": 71233018.0,
      "step": 2053
    },
    {
      "epoch": 0.38142989786443826,
      "grad_norm": 1.4669902180806174,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8614490032196045,
      "num_tokens": 71271803.0,
      "step": 2054
    },
    {
      "epoch": 0.3816155988857939,
      "grad_norm": 1.4594308859977567,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8747477531433105,
      "num_tokens": 71301500.0,
      "step": 2055
    },
    {
      "epoch": 0.3818012999071495,
      "grad_norm": 1.5410092078570086,
      "learning_rate": 1e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.844043493270874,
      "num_tokens": 71333535.0,
      "step": 2056
    },
    {
      "epoch": 0.3819870009285051,
      "grad_norm": 1.421738926730309,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8614802360534668,
      "num_tokens": 71368402.0,
      "step": 2057
    },
    {
      "epoch": 0.38217270194986075,
      "grad_norm": 1.6629387373142432,
      "learning_rate": 1e-06,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8608582615852356,
      "num_tokens": 71401065.0,
      "step": 2058
    },
    {
      "epoch": 0.3823584029712163,
      "grad_norm": 1.4747527847818531,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8552675843238831,
      "num_tokens": 71437651.0,
      "step": 2059
    },
    {
      "epoch": 0.38254410399257194,
      "grad_norm": 1.6029710338063037,
      "learning_rate": 1e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8564184904098511,
      "num_tokens": 71467086.0,
      "step": 2060
    },
    {
      "epoch": 0.38272980501392756,
      "grad_norm": 1.5406365101801476,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8781372308731079,
      "num_tokens": 71495937.0,
      "step": 2061
    },
    {
      "epoch": 0.3829155060352832,
      "grad_norm": 1.4884737868876579,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8667967319488525,
      "num_tokens": 71532905.0,
      "step": 2062
    },
    {
      "epoch": 0.3831012070566388,
      "grad_norm": 1.3802431167851898,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8710910081863403,
      "num_tokens": 71573342.0,
      "step": 2063
    },
    {
      "epoch": 0.3832869080779944,
      "grad_norm": 1.4729627856044747,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8681535720825195,
      "num_tokens": 71608178.0,
      "step": 2064
    },
    {
      "epoch": 0.38347260909935005,
      "grad_norm": 1.5066421224245843,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8694461584091187,
      "num_tokens": 71644694.0,
      "step": 2065
    },
    {
      "epoch": 0.38365831012070567,
      "grad_norm": 1.5475165558221184,
      "learning_rate": 1e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8511400818824768,
      "num_tokens": 71679290.0,
      "step": 2066
    },
    {
      "epoch": 0.3838440111420613,
      "grad_norm": 1.4513641578310152,
      "learning_rate": 1e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8557670712471008,
      "num_tokens": 71716138.0,
      "step": 2067
    },
    {
      "epoch": 0.3840297121634169,
      "grad_norm": 1.6028003997743436,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.871610164642334,
      "num_tokens": 71751282.0,
      "step": 2068
    },
    {
      "epoch": 0.38421541318477254,
      "grad_norm": 1.444543423933209,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8663800954818726,
      "num_tokens": 71787737.0,
      "step": 2069
    },
    {
      "epoch": 0.38440111420612816,
      "grad_norm": 1.4720567979094799,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8623565435409546,
      "num_tokens": 71824096.0,
      "step": 2070
    },
    {
      "epoch": 0.3845868152274837,
      "grad_norm": 1.5061725934666736,
      "learning_rate": 1e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8508209586143494,
      "num_tokens": 71859987.0,
      "step": 2071
    },
    {
      "epoch": 0.38477251624883935,
      "grad_norm": 1.5249743228490513,
      "learning_rate": 1e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8561644554138184,
      "num_tokens": 71893056.0,
      "step": 2072
    },
    {
      "epoch": 0.38495821727019497,
      "grad_norm": 1.528056311947346,
      "learning_rate": 1e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8534399271011353,
      "num_tokens": 71925940.0,
      "step": 2073
    },
    {
      "epoch": 0.3851439182915506,
      "grad_norm": 1.3835197669218637,
      "learning_rate": 1e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.858370304107666,
      "num_tokens": 71965778.0,
      "step": 2074
    },
    {
      "epoch": 0.3853296193129062,
      "grad_norm": 1.4422883620178197,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8688592910766602,
      "num_tokens": 72001749.0,
      "step": 2075
    },
    {
      "epoch": 0.38551532033426184,
      "grad_norm": 1.560311500258051,
      "learning_rate": 1e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.8666499257087708,
      "num_tokens": 72030249.0,
      "step": 2076
    },
    {
      "epoch": 0.38570102135561746,
      "grad_norm": 1.4604735843893035,
      "learning_rate": 1e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8616881370544434,
      "num_tokens": 72065503.0,
      "step": 2077
    },
    {
      "epoch": 0.3858867223769731,
      "grad_norm": 1.3482092009095823,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8658517599105835,
      "num_tokens": 72109469.0,
      "step": 2078
    },
    {
      "epoch": 0.3860724233983287,
      "grad_norm": 1.5302931031442752,
      "learning_rate": 1e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8402848839759827,
      "num_tokens": 72145657.0,
      "step": 2079
    },
    {
      "epoch": 0.3862581244196843,
      "grad_norm": 1.5185279084302779,
      "learning_rate": 1e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8587055206298828,
      "num_tokens": 72178398.0,
      "step": 2080
    },
    {
      "epoch": 0.38644382544103995,
      "grad_norm": 1.5552483202568186,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8763555288314819,
      "num_tokens": 72210397.0,
      "step": 2081
    },
    {
      "epoch": 0.38662952646239557,
      "grad_norm": 1.4660220692619488,
      "learning_rate": 1e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8628993034362793,
      "num_tokens": 72246953.0,
      "step": 2082
    },
    {
      "epoch": 0.38681522748375113,
      "grad_norm": 1.4592889074786493,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8630112409591675,
      "num_tokens": 72283447.0,
      "step": 2083
    },
    {
      "epoch": 0.38700092850510676,
      "grad_norm": 1.6878935509475137,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8664594292640686,
      "num_tokens": 72315853.0,
      "step": 2084
    },
    {
      "epoch": 0.3871866295264624,
      "grad_norm": 1.4357444073732142,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8636605739593506,
      "num_tokens": 72351099.0,
      "step": 2085
    },
    {
      "epoch": 0.387372330547818,
      "grad_norm": 1.3926116031330573,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8760141730308533,
      "num_tokens": 72383415.0,
      "step": 2086
    },
    {
      "epoch": 0.3875580315691736,
      "grad_norm": 1.435921517182951,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8620220422744751,
      "num_tokens": 72419821.0,
      "step": 2087
    },
    {
      "epoch": 0.38774373259052924,
      "grad_norm": 1.5132197998601062,
      "learning_rate": 1e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8580952882766724,
      "num_tokens": 72459111.0,
      "step": 2088
    },
    {
      "epoch": 0.38792943361188487,
      "grad_norm": 1.531530532886763,
      "learning_rate": 1e-06,
      "loss": 0.3871,
      "mean_token_accuracy": 0.8691678047180176,
      "num_tokens": 72490062.0,
      "step": 2089
    },
    {
      "epoch": 0.3881151346332405,
      "grad_norm": 1.5677745277480886,
      "learning_rate": 1e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8407381772994995,
      "num_tokens": 72526220.0,
      "step": 2090
    },
    {
      "epoch": 0.3883008356545961,
      "grad_norm": 1.5073921565562955,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8701749444007874,
      "num_tokens": 72561055.0,
      "step": 2091
    },
    {
      "epoch": 0.38848653667595173,
      "grad_norm": 1.4779325994240191,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8588665127754211,
      "num_tokens": 72596682.0,
      "step": 2092
    },
    {
      "epoch": 0.38867223769730735,
      "grad_norm": 1.4080315775013479,
      "learning_rate": 1e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8551884889602661,
      "num_tokens": 72636499.0,
      "step": 2093
    },
    {
      "epoch": 0.388857938718663,
      "grad_norm": 1.6781769398685396,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8616305589675903,
      "num_tokens": 72664810.0,
      "step": 2094
    },
    {
      "epoch": 0.38904363974001854,
      "grad_norm": 1.5807727508667835,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8592065572738647,
      "num_tokens": 72694983.0,
      "step": 2095
    },
    {
      "epoch": 0.38922934076137417,
      "grad_norm": 1.5372187694979724,
      "learning_rate": 1e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8499789834022522,
      "num_tokens": 72726948.0,
      "step": 2096
    },
    {
      "epoch": 0.3894150417827298,
      "grad_norm": 1.4179330737801457,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8672894835472107,
      "num_tokens": 72768074.0,
      "step": 2097
    },
    {
      "epoch": 0.3896007428040854,
      "grad_norm": 1.600149674125095,
      "learning_rate": 1e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8541212677955627,
      "num_tokens": 72799787.0,
      "step": 2098
    },
    {
      "epoch": 0.38978644382544103,
      "grad_norm": 1.5137401367393342,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.865157961845398,
      "num_tokens": 72833498.0,
      "step": 2099
    },
    {
      "epoch": 0.38997214484679665,
      "grad_norm": 1.6693157419680165,
      "learning_rate": 1e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8591502904891968,
      "num_tokens": 72863667.0,
      "step": 2100
    },
    {
      "epoch": 0.3901578458681523,
      "grad_norm": 1.532483773724434,
      "learning_rate": 1e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8494835495948792,
      "num_tokens": 72900709.0,
      "step": 2101
    },
    {
      "epoch": 0.3903435468895079,
      "grad_norm": 1.4305909065730051,
      "learning_rate": 1e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8512662053108215,
      "num_tokens": 72939623.0,
      "step": 2102
    },
    {
      "epoch": 0.3905292479108635,
      "grad_norm": 1.4585734956510819,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.869347333908081,
      "num_tokens": 72974423.0,
      "step": 2103
    },
    {
      "epoch": 0.39071494893221914,
      "grad_norm": 1.3285780656175443,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8750135898590088,
      "num_tokens": 73015479.0,
      "step": 2104
    },
    {
      "epoch": 0.39090064995357476,
      "grad_norm": 1.4813806687455273,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8750452995300293,
      "num_tokens": 73049700.0,
      "step": 2105
    },
    {
      "epoch": 0.3910863509749304,
      "grad_norm": 1.5163877613757117,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8738715648651123,
      "num_tokens": 73083440.0,
      "step": 2106
    },
    {
      "epoch": 0.39127205199628595,
      "grad_norm": 1.5016324181687943,
      "learning_rate": 1e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8655838370323181,
      "num_tokens": 73112842.0,
      "step": 2107
    },
    {
      "epoch": 0.3914577530176416,
      "grad_norm": 1.3157650441296393,
      "learning_rate": 1e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8628548383712769,
      "num_tokens": 73157317.0,
      "step": 2108
    },
    {
      "epoch": 0.3916434540389972,
      "grad_norm": 1.4898891121270716,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8652327060699463,
      "num_tokens": 73196003.0,
      "step": 2109
    },
    {
      "epoch": 0.3918291550603528,
      "grad_norm": 1.6309551366923187,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.860466718673706,
      "num_tokens": 73226749.0,
      "step": 2110
    },
    {
      "epoch": 0.39201485608170844,
      "grad_norm": 1.4692970765554378,
      "learning_rate": 1e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8640799522399902,
      "num_tokens": 73263489.0,
      "step": 2111
    },
    {
      "epoch": 0.39220055710306406,
      "grad_norm": 1.550814903852766,
      "learning_rate": 1e-06,
      "loss": 0.446,
      "mean_token_accuracy": 0.8533127307891846,
      "num_tokens": 73297272.0,
      "step": 2112
    },
    {
      "epoch": 0.3923862581244197,
      "grad_norm": 1.5322684470930472,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.859699547290802,
      "num_tokens": 73333307.0,
      "step": 2113
    },
    {
      "epoch": 0.3925719591457753,
      "grad_norm": 1.59147654188662,
      "learning_rate": 1e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8521312475204468,
      "num_tokens": 73363090.0,
      "step": 2114
    },
    {
      "epoch": 0.39275766016713093,
      "grad_norm": 1.4592650935931315,
      "learning_rate": 1e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8567245006561279,
      "num_tokens": 73400322.0,
      "step": 2115
    },
    {
      "epoch": 0.39294336118848655,
      "grad_norm": 1.4256124239883123,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.870327353477478,
      "num_tokens": 73435542.0,
      "step": 2116
    },
    {
      "epoch": 0.3931290622098422,
      "grad_norm": 1.5526895798135927,
      "learning_rate": 1e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8549947142601013,
      "num_tokens": 73467433.0,
      "step": 2117
    },
    {
      "epoch": 0.3933147632311978,
      "grad_norm": 1.6045096486121524,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8605537414550781,
      "num_tokens": 73506967.0,
      "step": 2118
    },
    {
      "epoch": 0.39350046425255336,
      "grad_norm": 1.4356504125286358,
      "learning_rate": 1e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8571053743362427,
      "num_tokens": 73541887.0,
      "step": 2119
    },
    {
      "epoch": 0.393686165273909,
      "grad_norm": 1.6954655272764683,
      "learning_rate": 1e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8399609923362732,
      "num_tokens": 73569523.0,
      "step": 2120
    },
    {
      "epoch": 0.3938718662952646,
      "grad_norm": 1.3306006046222618,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8756812810897827,
      "num_tokens": 73610138.0,
      "step": 2121
    },
    {
      "epoch": 0.3940575673166202,
      "grad_norm": 1.5583616879050488,
      "learning_rate": 1e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8561320304870605,
      "num_tokens": 73643018.0,
      "step": 2122
    },
    {
      "epoch": 0.39424326833797585,
      "grad_norm": 1.6361202199006002,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8710250854492188,
      "num_tokens": 73668770.0,
      "step": 2123
    },
    {
      "epoch": 0.39442896935933147,
      "grad_norm": 1.6775839151984395,
      "learning_rate": 1e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8491501212120056,
      "num_tokens": 73694983.0,
      "step": 2124
    },
    {
      "epoch": 0.3946146703806871,
      "grad_norm": 1.3441093482770157,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8713220357894897,
      "num_tokens": 73730625.0,
      "step": 2125
    },
    {
      "epoch": 0.3948003714020427,
      "grad_norm": 1.449064343818745,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.873291015625,
      "num_tokens": 73763226.0,
      "step": 2126
    },
    {
      "epoch": 0.39498607242339834,
      "grad_norm": 1.541630678938628,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8688525557518005,
      "num_tokens": 73795780.0,
      "step": 2127
    },
    {
      "epoch": 0.39517177344475396,
      "grad_norm": 1.7112969483121878,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8615984916687012,
      "num_tokens": 73826300.0,
      "step": 2128
    },
    {
      "epoch": 0.3953574744661096,
      "grad_norm": 1.5181855332830951,
      "learning_rate": 1e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8611714243888855,
      "num_tokens": 73861132.0,
      "step": 2129
    },
    {
      "epoch": 0.3955431754874652,
      "grad_norm": 1.6108519287932592,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8721590042114258,
      "num_tokens": 73885593.0,
      "step": 2130
    },
    {
      "epoch": 0.3957288765088208,
      "grad_norm": 1.5161035305149697,
      "learning_rate": 1e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8622671961784363,
      "num_tokens": 73920688.0,
      "step": 2131
    },
    {
      "epoch": 0.3959145775301764,
      "grad_norm": 1.3808393055612402,
      "learning_rate": 1e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8529006838798523,
      "num_tokens": 73962627.0,
      "step": 2132
    },
    {
      "epoch": 0.396100278551532,
      "grad_norm": 1.5936009187150373,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8618860840797424,
      "num_tokens": 73997580.0,
      "step": 2133
    },
    {
      "epoch": 0.39628597957288764,
      "grad_norm": 1.5179692687831041,
      "learning_rate": 1e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8589956760406494,
      "num_tokens": 74032462.0,
      "step": 2134
    },
    {
      "epoch": 0.39647168059424326,
      "grad_norm": 1.604034357688145,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8674418926239014,
      "num_tokens": 74064305.0,
      "step": 2135
    },
    {
      "epoch": 0.3966573816155989,
      "grad_norm": 1.4318956755004153,
      "learning_rate": 1e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8605620861053467,
      "num_tokens": 74101083.0,
      "step": 2136
    },
    {
      "epoch": 0.3968430826369545,
      "grad_norm": 1.3914076373757625,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8645532131195068,
      "num_tokens": 74142696.0,
      "step": 2137
    },
    {
      "epoch": 0.3970287836583101,
      "grad_norm": 1.5569670370852708,
      "learning_rate": 1e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.848035454750061,
      "num_tokens": 74175693.0,
      "step": 2138
    },
    {
      "epoch": 0.39721448467966575,
      "grad_norm": 1.5273545046535546,
      "learning_rate": 1e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8533107042312622,
      "num_tokens": 74211285.0,
      "step": 2139
    },
    {
      "epoch": 0.39740018570102137,
      "grad_norm": 1.4114396373890243,
      "learning_rate": 1e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8547247648239136,
      "num_tokens": 74250922.0,
      "step": 2140
    },
    {
      "epoch": 0.397585886722377,
      "grad_norm": 1.5976749253926357,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8754231929779053,
      "num_tokens": 74283256.0,
      "step": 2141
    },
    {
      "epoch": 0.3977715877437326,
      "grad_norm": 1.405249774898946,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8677330017089844,
      "num_tokens": 74322318.0,
      "step": 2142
    },
    {
      "epoch": 0.39795728876508824,
      "grad_norm": 1.336772541105593,
      "learning_rate": 1e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8500170707702637,
      "num_tokens": 74369433.0,
      "step": 2143
    },
    {
      "epoch": 0.3981429897864438,
      "grad_norm": 1.3797458435505277,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8736428618431091,
      "num_tokens": 74406002.0,
      "step": 2144
    },
    {
      "epoch": 0.3983286908077994,
      "grad_norm": 1.5795976760861357,
      "learning_rate": 1e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8556495308876038,
      "num_tokens": 74439638.0,
      "step": 2145
    },
    {
      "epoch": 0.39851439182915505,
      "grad_norm": 1.4694565987881032,
      "learning_rate": 1e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.852308988571167,
      "num_tokens": 74476871.0,
      "step": 2146
    },
    {
      "epoch": 0.39870009285051067,
      "grad_norm": 1.6498403718941679,
      "learning_rate": 1e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8596926927566528,
      "num_tokens": 74508858.0,
      "step": 2147
    },
    {
      "epoch": 0.3988857938718663,
      "grad_norm": 1.5727835149696658,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8748977780342102,
      "num_tokens": 74541822.0,
      "step": 2148
    },
    {
      "epoch": 0.3990714948932219,
      "grad_norm": 1.6346686324483508,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.877220630645752,
      "num_tokens": 74570884.0,
      "step": 2149
    },
    {
      "epoch": 0.39925719591457753,
      "grad_norm": 1.592041184574718,
      "learning_rate": 1e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8605207204818726,
      "num_tokens": 74602128.0,
      "step": 2150
    },
    {
      "epoch": 0.39944289693593316,
      "grad_norm": 1.4162265723639378,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8711594343185425,
      "num_tokens": 74637471.0,
      "step": 2151
    },
    {
      "epoch": 0.3996285979572888,
      "grad_norm": 1.404586442846134,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8597097992897034,
      "num_tokens": 74677161.0,
      "step": 2152
    },
    {
      "epoch": 0.3998142989786444,
      "grad_norm": 1.480697011351984,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8700253963470459,
      "num_tokens": 74708026.0,
      "step": 2153
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6212104675272592,
      "learning_rate": 1e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8608448505401611,
      "num_tokens": 74736519.0,
      "step": 2154
    },
    {
      "epoch": 0.40018570102135564,
      "grad_norm": 1.3400500680764393,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8705202341079712,
      "num_tokens": 74778123.0,
      "step": 2155
    },
    {
      "epoch": 0.4003714020427112,
      "grad_norm": 1.5348608804155468,
      "learning_rate": 1e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8524091243743896,
      "num_tokens": 74814072.0,
      "step": 2156
    },
    {
      "epoch": 0.40055710306406683,
      "grad_norm": 1.5939775765409048,
      "learning_rate": 1e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8521822690963745,
      "num_tokens": 74849678.0,
      "step": 2157
    },
    {
      "epoch": 0.40074280408542245,
      "grad_norm": 1.7188933676520408,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.860119104385376,
      "num_tokens": 74878253.0,
      "step": 2158
    },
    {
      "epoch": 0.4009285051067781,
      "grad_norm": 1.5405288562278057,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8692190051078796,
      "num_tokens": 74910811.0,
      "step": 2159
    },
    {
      "epoch": 0.4011142061281337,
      "grad_norm": 1.4882101303871158,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8692901730537415,
      "num_tokens": 74946593.0,
      "step": 2160
    },
    {
      "epoch": 0.4012999071494893,
      "grad_norm": 1.4188003751172842,
      "learning_rate": 1e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.863967776298523,
      "num_tokens": 74987185.0,
      "step": 2161
    },
    {
      "epoch": 0.40148560817084494,
      "grad_norm": 1.5903967158492254,
      "learning_rate": 1e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8452446460723877,
      "num_tokens": 75023503.0,
      "step": 2162
    },
    {
      "epoch": 0.40167130919220057,
      "grad_norm": 1.592289142646996,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.876917839050293,
      "num_tokens": 75052953.0,
      "step": 2163
    },
    {
      "epoch": 0.4018570102135562,
      "grad_norm": 1.5093478744440358,
      "learning_rate": 1e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8620449304580688,
      "num_tokens": 75088542.0,
      "step": 2164
    },
    {
      "epoch": 0.4020427112349118,
      "grad_norm": 1.4978459678044826,
      "learning_rate": 1e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8562384247779846,
      "num_tokens": 75121040.0,
      "step": 2165
    },
    {
      "epoch": 0.40222841225626743,
      "grad_norm": 1.4474543323342433,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8702971935272217,
      "num_tokens": 75157980.0,
      "step": 2166
    },
    {
      "epoch": 0.40241411327762305,
      "grad_norm": 1.4454103404951335,
      "learning_rate": 1e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8436517715454102,
      "num_tokens": 75199986.0,
      "step": 2167
    },
    {
      "epoch": 0.4025998142989786,
      "grad_norm": 1.5429057928748013,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.868719220161438,
      "num_tokens": 75233568.0,
      "step": 2168
    },
    {
      "epoch": 0.40278551532033424,
      "grad_norm": 1.5534297458245185,
      "learning_rate": 1e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8630636930465698,
      "num_tokens": 75268269.0,
      "step": 2169
    },
    {
      "epoch": 0.40297121634168986,
      "grad_norm": 1.416977485050321,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8668900728225708,
      "num_tokens": 75306188.0,
      "step": 2170
    },
    {
      "epoch": 0.4031569173630455,
      "grad_norm": 1.5085145202730517,
      "learning_rate": 1e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8537420630455017,
      "num_tokens": 75341121.0,
      "step": 2171
    },
    {
      "epoch": 0.4033426183844011,
      "grad_norm": 1.5299963803411616,
      "learning_rate": 1e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8428403735160828,
      "num_tokens": 75379276.0,
      "step": 2172
    },
    {
      "epoch": 0.40352831940575673,
      "grad_norm": 1.5192531828630347,
      "learning_rate": 1e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8633631467819214,
      "num_tokens": 75412824.0,
      "step": 2173
    },
    {
      "epoch": 0.40371402042711235,
      "grad_norm": 1.702130695098396,
      "learning_rate": 1e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8474820852279663,
      "num_tokens": 75439913.0,
      "step": 2174
    },
    {
      "epoch": 0.403899721448468,
      "grad_norm": 1.5560545551149012,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8618242144584656,
      "num_tokens": 75472675.0,
      "step": 2175
    },
    {
      "epoch": 0.4040854224698236,
      "grad_norm": 1.678294758766401,
      "learning_rate": 1e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8518999814987183,
      "num_tokens": 75503885.0,
      "step": 2176
    },
    {
      "epoch": 0.4042711234911792,
      "grad_norm": 1.531931548935808,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.852116584777832,
      "num_tokens": 75541433.0,
      "step": 2177
    },
    {
      "epoch": 0.40445682451253484,
      "grad_norm": 1.4950616989204693,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8639270663261414,
      "num_tokens": 75577603.0,
      "step": 2178
    },
    {
      "epoch": 0.40464252553389046,
      "grad_norm": 1.3498248203389578,
      "learning_rate": 1e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8482940196990967,
      "num_tokens": 75622046.0,
      "step": 2179
    },
    {
      "epoch": 0.40482822655524603,
      "grad_norm": 1.5253074029197122,
      "learning_rate": 1e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8618776798248291,
      "num_tokens": 75658447.0,
      "step": 2180
    },
    {
      "epoch": 0.40501392757660165,
      "grad_norm": 1.3977079242159853,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8681448101997375,
      "num_tokens": 75694668.0,
      "step": 2181
    },
    {
      "epoch": 0.4051996285979573,
      "grad_norm": 1.4377300661740429,
      "learning_rate": 1e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8552202582359314,
      "num_tokens": 75736237.0,
      "step": 2182
    },
    {
      "epoch": 0.4053853296193129,
      "grad_norm": 1.6137634539576295,
      "learning_rate": 1e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8448483943939209,
      "num_tokens": 75769605.0,
      "step": 2183
    },
    {
      "epoch": 0.4055710306406685,
      "grad_norm": 1.5011537489368838,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8706591129302979,
      "num_tokens": 75801610.0,
      "step": 2184
    },
    {
      "epoch": 0.40575673166202414,
      "grad_norm": 1.4893409082368843,
      "learning_rate": 1e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.859180212020874,
      "num_tokens": 75838297.0,
      "step": 2185
    },
    {
      "epoch": 0.40594243268337976,
      "grad_norm": 1.4487201707108406,
      "learning_rate": 1e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8602046370506287,
      "num_tokens": 75875761.0,
      "step": 2186
    },
    {
      "epoch": 0.4061281337047354,
      "grad_norm": 1.4700421157031087,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8628146648406982,
      "num_tokens": 75908236.0,
      "step": 2187
    },
    {
      "epoch": 0.406313834726091,
      "grad_norm": 1.5418231654997456,
      "learning_rate": 1e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8494922518730164,
      "num_tokens": 75942982.0,
      "step": 2188
    },
    {
      "epoch": 0.4064995357474466,
      "grad_norm": 1.388380311895085,
      "learning_rate": 1e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8635108470916748,
      "num_tokens": 75980766.0,
      "step": 2189
    },
    {
      "epoch": 0.40668523676880225,
      "grad_norm": 1.4157968099448766,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8694245219230652,
      "num_tokens": 76019105.0,
      "step": 2190
    },
    {
      "epoch": 0.40687093779015787,
      "grad_norm": 1.4916652874592764,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8685347437858582,
      "num_tokens": 76055107.0,
      "step": 2191
    },
    {
      "epoch": 0.40705663881151344,
      "grad_norm": 1.4640890772414838,
      "learning_rate": 1e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8483364582061768,
      "num_tokens": 76092088.0,
      "step": 2192
    },
    {
      "epoch": 0.40724233983286906,
      "grad_norm": 1.547199305870445,
      "learning_rate": 1e-06,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8703500032424927,
      "num_tokens": 76128171.0,
      "step": 2193
    },
    {
      "epoch": 0.4074280408542247,
      "grad_norm": 1.4988536101135599,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8585749864578247,
      "num_tokens": 76165985.0,
      "step": 2194
    },
    {
      "epoch": 0.4076137418755803,
      "grad_norm": 1.500342262403888,
      "learning_rate": 1e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8578122854232788,
      "num_tokens": 76198942.0,
      "step": 2195
    },
    {
      "epoch": 0.4077994428969359,
      "grad_norm": 1.7013909558482998,
      "learning_rate": 1e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8475319147109985,
      "num_tokens": 76227849.0,
      "step": 2196
    },
    {
      "epoch": 0.40798514391829155,
      "grad_norm": 1.3394206755375577,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8729931712150574,
      "num_tokens": 76268470.0,
      "step": 2197
    },
    {
      "epoch": 0.40817084493964717,
      "grad_norm": 1.5045645306651452,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8727006316184998,
      "num_tokens": 76302983.0,
      "step": 2198
    },
    {
      "epoch": 0.4083565459610028,
      "grad_norm": 1.4529191092234552,
      "learning_rate": 1e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8483829498291016,
      "num_tokens": 76342432.0,
      "step": 2199
    },
    {
      "epoch": 0.4085422469823584,
      "grad_norm": 1.542574621052026,
      "learning_rate": 1e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8494713306427002,
      "num_tokens": 76375351.0,
      "step": 2200
    },
    {
      "epoch": 0.40872794800371404,
      "grad_norm": 1.6094505231089533,
      "learning_rate": 1e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8581300973892212,
      "num_tokens": 76403929.0,
      "step": 2201
    },
    {
      "epoch": 0.40891364902506966,
      "grad_norm": 1.6065198743685356,
      "learning_rate": 1e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8657637238502502,
      "num_tokens": 76432048.0,
      "step": 2202
    },
    {
      "epoch": 0.4090993500464253,
      "grad_norm": 1.5124368312982772,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8730271458625793,
      "num_tokens": 76463994.0,
      "step": 2203
    },
    {
      "epoch": 0.40928505106778085,
      "grad_norm": 1.6402483677508048,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8593199849128723,
      "num_tokens": 76494776.0,
      "step": 2204
    },
    {
      "epoch": 0.40947075208913647,
      "grad_norm": 1.4595021314315137,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8744110465049744,
      "num_tokens": 76530292.0,
      "step": 2205
    },
    {
      "epoch": 0.4096564531104921,
      "grad_norm": 1.486325561130645,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8653264045715332,
      "num_tokens": 76566082.0,
      "step": 2206
    },
    {
      "epoch": 0.4098421541318477,
      "grad_norm": 1.609887480878166,
      "learning_rate": 1e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8557668328285217,
      "num_tokens": 76598773.0,
      "step": 2207
    },
    {
      "epoch": 0.41002785515320334,
      "grad_norm": 1.6072065727674256,
      "learning_rate": 1e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8477591276168823,
      "num_tokens": 76629376.0,
      "step": 2208
    },
    {
      "epoch": 0.41021355617455896,
      "grad_norm": 1.5082469952441007,
      "learning_rate": 1e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8627672791481018,
      "num_tokens": 76663497.0,
      "step": 2209
    },
    {
      "epoch": 0.4103992571959146,
      "grad_norm": 1.6186658723007556,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8681949377059937,
      "num_tokens": 76696229.0,
      "step": 2210
    },
    {
      "epoch": 0.4105849582172702,
      "grad_norm": 1.4668263047953822,
      "learning_rate": 1e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8634141683578491,
      "num_tokens": 76731061.0,
      "step": 2211
    },
    {
      "epoch": 0.4107706592386258,
      "grad_norm": 1.5544021719733794,
      "learning_rate": 1e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8520695567131042,
      "num_tokens": 76768770.0,
      "step": 2212
    },
    {
      "epoch": 0.41095636025998145,
      "grad_norm": 1.4804076532237174,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.858576774597168,
      "num_tokens": 76803221.0,
      "step": 2213
    },
    {
      "epoch": 0.41114206128133707,
      "grad_norm": 1.49792957207672,
      "learning_rate": 1e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8378444910049438,
      "num_tokens": 76839017.0,
      "step": 2214
    },
    {
      "epoch": 0.4113277623026927,
      "grad_norm": 1.450847168410163,
      "learning_rate": 1e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8634832501411438,
      "num_tokens": 76872841.0,
      "step": 2215
    },
    {
      "epoch": 0.41151346332404826,
      "grad_norm": 1.477112779634383,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8657373189926147,
      "num_tokens": 76904972.0,
      "step": 2216
    },
    {
      "epoch": 0.4116991643454039,
      "grad_norm": 1.4625970869749227,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8705381155014038,
      "num_tokens": 76940426.0,
      "step": 2217
    },
    {
      "epoch": 0.4118848653667595,
      "grad_norm": 1.393785239549549,
      "learning_rate": 1e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8574564456939697,
      "num_tokens": 76979841.0,
      "step": 2218
    },
    {
      "epoch": 0.4120705663881151,
      "grad_norm": 1.6231341096800243,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8621913194656372,
      "num_tokens": 77010982.0,
      "step": 2219
    },
    {
      "epoch": 0.41225626740947074,
      "grad_norm": 1.536149291435251,
      "learning_rate": 1e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8544789552688599,
      "num_tokens": 77043858.0,
      "step": 2220
    },
    {
      "epoch": 0.41244196843082637,
      "grad_norm": 1.562473799927853,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8623925447463989,
      "num_tokens": 77076481.0,
      "step": 2221
    },
    {
      "epoch": 0.412627669452182,
      "grad_norm": 1.6792299180839676,
      "learning_rate": 1e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8422421216964722,
      "num_tokens": 77106856.0,
      "step": 2222
    },
    {
      "epoch": 0.4128133704735376,
      "grad_norm": 1.4845396692608415,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8528257608413696,
      "num_tokens": 77144780.0,
      "step": 2223
    },
    {
      "epoch": 0.41299907149489323,
      "grad_norm": 1.548258442594737,
      "learning_rate": 1e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8613545298576355,
      "num_tokens": 77178039.0,
      "step": 2224
    },
    {
      "epoch": 0.41318477251624885,
      "grad_norm": 1.5958565984091864,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8588362336158752,
      "num_tokens": 77208674.0,
      "step": 2225
    },
    {
      "epoch": 0.4133704735376045,
      "grad_norm": 1.471060726335119,
      "learning_rate": 1e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8557097911834717,
      "num_tokens": 77245219.0,
      "step": 2226
    },
    {
      "epoch": 0.4135561745589601,
      "grad_norm": 1.6270125301351288,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8670113682746887,
      "num_tokens": 77277299.0,
      "step": 2227
    },
    {
      "epoch": 0.41374187558031567,
      "grad_norm": 1.5328011479117025,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8706220984458923,
      "num_tokens": 77308901.0,
      "step": 2228
    },
    {
      "epoch": 0.4139275766016713,
      "grad_norm": 1.5540478268159574,
      "learning_rate": 1e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8444801568984985,
      "num_tokens": 77342546.0,
      "step": 2229
    },
    {
      "epoch": 0.4141132776230269,
      "grad_norm": 1.480612478988531,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8694843649864197,
      "num_tokens": 77376051.0,
      "step": 2230
    },
    {
      "epoch": 0.41429897864438253,
      "grad_norm": 1.452872875305871,
      "learning_rate": 1e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8488340377807617,
      "num_tokens": 77416072.0,
      "step": 2231
    },
    {
      "epoch": 0.41448467966573815,
      "grad_norm": 1.616639610505551,
      "learning_rate": 1e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8540474772453308,
      "num_tokens": 77446348.0,
      "step": 2232
    },
    {
      "epoch": 0.4146703806870938,
      "grad_norm": 1.4474400233124187,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.856308102607727,
      "num_tokens": 77481576.0,
      "step": 2233
    },
    {
      "epoch": 0.4148560817084494,
      "grad_norm": 1.4232095285149435,
      "learning_rate": 1e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.860600471496582,
      "num_tokens": 77517412.0,
      "step": 2234
    },
    {
      "epoch": 0.415041782729805,
      "grad_norm": 1.5223927860473778,
      "learning_rate": 1e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8520500063896179,
      "num_tokens": 77550100.0,
      "step": 2235
    },
    {
      "epoch": 0.41522748375116064,
      "grad_norm": 1.5184872467524537,
      "learning_rate": 1e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.859346866607666,
      "num_tokens": 77582523.0,
      "step": 2236
    },
    {
      "epoch": 0.41541318477251626,
      "grad_norm": 1.4858803181947304,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8547732830047607,
      "num_tokens": 77622299.0,
      "step": 2237
    },
    {
      "epoch": 0.4155988857938719,
      "grad_norm": 1.4183652493814523,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8706884384155273,
      "num_tokens": 77661492.0,
      "step": 2238
    },
    {
      "epoch": 0.4157845868152275,
      "grad_norm": 1.6534012701367224,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8638101816177368,
      "num_tokens": 77689921.0,
      "step": 2239
    },
    {
      "epoch": 0.4159702878365831,
      "grad_norm": 1.4739715146801404,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8774896264076233,
      "num_tokens": 77727180.0,
      "step": 2240
    },
    {
      "epoch": 0.4161559888579387,
      "grad_norm": 1.3682759945736684,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8693052530288696,
      "num_tokens": 77764527.0,
      "step": 2241
    },
    {
      "epoch": 0.4163416898792943,
      "grad_norm": 1.4208417729845488,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8656911253929138,
      "num_tokens": 77802622.0,
      "step": 2242
    },
    {
      "epoch": 0.41652739090064994,
      "grad_norm": 1.3697655067337005,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8735079765319824,
      "num_tokens": 77838950.0,
      "step": 2243
    },
    {
      "epoch": 0.41671309192200556,
      "grad_norm": 1.5890833382894205,
      "learning_rate": 1e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8589287996292114,
      "num_tokens": 77871218.0,
      "step": 2244
    },
    {
      "epoch": 0.4168987929433612,
      "grad_norm": 1.5017847189999611,
      "learning_rate": 1e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8565496206283569,
      "num_tokens": 77908080.0,
      "step": 2245
    },
    {
      "epoch": 0.4170844939647168,
      "grad_norm": 1.5438562909368192,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8617232441902161,
      "num_tokens": 77943300.0,
      "step": 2246
    },
    {
      "epoch": 0.41727019498607243,
      "grad_norm": 1.5224300233589534,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8688799142837524,
      "num_tokens": 77978315.0,
      "step": 2247
    },
    {
      "epoch": 0.41745589600742805,
      "grad_norm": 1.7845799342189779,
      "learning_rate": 1e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8413243293762207,
      "num_tokens": 78007413.0,
      "step": 2248
    },
    {
      "epoch": 0.4176415970287837,
      "grad_norm": 1.458823159819231,
      "learning_rate": 1e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8569765090942383,
      "num_tokens": 78042661.0,
      "step": 2249
    },
    {
      "epoch": 0.4178272980501393,
      "grad_norm": 1.536790508342537,
      "learning_rate": 1e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8510279655456543,
      "num_tokens": 78079121.0,
      "step": 2250
    },
    {
      "epoch": 0.4180129990714949,
      "grad_norm": 1.6788838278851188,
      "learning_rate": 1e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8672874569892883,
      "num_tokens": 78105429.0,
      "step": 2251
    },
    {
      "epoch": 0.4181987000928505,
      "grad_norm": 1.5005938999279356,
      "learning_rate": 1e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8364262580871582,
      "num_tokens": 78145161.0,
      "step": 2252
    },
    {
      "epoch": 0.4183844011142061,
      "grad_norm": 1.4069370548262627,
      "learning_rate": 1e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8539716005325317,
      "num_tokens": 78186594.0,
      "step": 2253
    },
    {
      "epoch": 0.41857010213556173,
      "grad_norm": 1.5887153448944416,
      "learning_rate": 1e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8601826429367065,
      "num_tokens": 78220470.0,
      "step": 2254
    },
    {
      "epoch": 0.41875580315691735,
      "grad_norm": 1.8010499161929237,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8693956136703491,
      "num_tokens": 78245967.0,
      "step": 2255
    },
    {
      "epoch": 0.41894150417827297,
      "grad_norm": 1.536817521615407,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8602220416069031,
      "num_tokens": 78279003.0,
      "step": 2256
    },
    {
      "epoch": 0.4191272051996286,
      "grad_norm": 1.5316101804257303,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8641639947891235,
      "num_tokens": 78317996.0,
      "step": 2257
    },
    {
      "epoch": 0.4193129062209842,
      "grad_norm": 1.431666332365413,
      "learning_rate": 1e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8621330261230469,
      "num_tokens": 78355024.0,
      "step": 2258
    },
    {
      "epoch": 0.41949860724233984,
      "grad_norm": 1.555516199420637,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8862855434417725,
      "num_tokens": 78382379.0,
      "step": 2259
    },
    {
      "epoch": 0.41968430826369546,
      "grad_norm": 1.5425751382749626,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8586422204971313,
      "num_tokens": 78416560.0,
      "step": 2260
    },
    {
      "epoch": 0.4198700092850511,
      "grad_norm": 1.6606663963011805,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.873353123664856,
      "num_tokens": 78452177.0,
      "step": 2261
    },
    {
      "epoch": 0.4200557103064067,
      "grad_norm": 1.4813792470371034,
      "learning_rate": 1e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8498038053512573,
      "num_tokens": 78491322.0,
      "step": 2262
    },
    {
      "epoch": 0.4202414113277623,
      "grad_norm": 1.4130572916475819,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.859959065914154,
      "num_tokens": 78532399.0,
      "step": 2263
    },
    {
      "epoch": 0.4204271123491179,
      "grad_norm": 1.4775554572827474,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8681368827819824,
      "num_tokens": 78565592.0,
      "step": 2264
    },
    {
      "epoch": 0.4206128133704735,
      "grad_norm": 1.6755037788137015,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8627311587333679,
      "num_tokens": 78596644.0,
      "step": 2265
    },
    {
      "epoch": 0.42079851439182914,
      "grad_norm": 1.6402649014724913,
      "learning_rate": 1e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8661863803863525,
      "num_tokens": 78626089.0,
      "step": 2266
    },
    {
      "epoch": 0.42098421541318476,
      "grad_norm": 1.5097757746591616,
      "learning_rate": 1e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8512747287750244,
      "num_tokens": 78662137.0,
      "step": 2267
    },
    {
      "epoch": 0.4211699164345404,
      "grad_norm": 1.446092726932415,
      "learning_rate": 1e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8477512001991272,
      "num_tokens": 78700602.0,
      "step": 2268
    },
    {
      "epoch": 0.421355617455896,
      "grad_norm": 1.4468290602174145,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8611677289009094,
      "num_tokens": 78737346.0,
      "step": 2269
    },
    {
      "epoch": 0.4215413184772516,
      "grad_norm": 1.4549891760157145,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8658980131149292,
      "num_tokens": 78777304.0,
      "step": 2270
    },
    {
      "epoch": 0.42172701949860725,
      "grad_norm": 1.438057912208794,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8657417297363281,
      "num_tokens": 78814369.0,
      "step": 2271
    },
    {
      "epoch": 0.42191272051996287,
      "grad_norm": 1.4895255221376518,
      "learning_rate": 1e-06,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8589962720870972,
      "num_tokens": 78851414.0,
      "step": 2272
    },
    {
      "epoch": 0.4220984215413185,
      "grad_norm": 1.5094096069913507,
      "learning_rate": 1e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8610062599182129,
      "num_tokens": 78885173.0,
      "step": 2273
    },
    {
      "epoch": 0.4222841225626741,
      "grad_norm": 1.462150844476487,
      "learning_rate": 1e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.858022153377533,
      "num_tokens": 78926276.0,
      "step": 2274
    },
    {
      "epoch": 0.42246982358402974,
      "grad_norm": 1.4506657815603867,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8728289604187012,
      "num_tokens": 78958231.0,
      "step": 2275
    },
    {
      "epoch": 0.4226555246053853,
      "grad_norm": 1.4637783576184276,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8568351864814758,
      "num_tokens": 78996126.0,
      "step": 2276
    },
    {
      "epoch": 0.4228412256267409,
      "grad_norm": 1.4364674229721182,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8569128513336182,
      "num_tokens": 79031975.0,
      "step": 2277
    },
    {
      "epoch": 0.42302692664809655,
      "grad_norm": 1.4776243232232198,
      "learning_rate": 1e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8493691086769104,
      "num_tokens": 79073251.0,
      "step": 2278
    },
    {
      "epoch": 0.42321262766945217,
      "grad_norm": 1.706217674104008,
      "learning_rate": 1e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.831630527973175,
      "num_tokens": 79107084.0,
      "step": 2279
    },
    {
      "epoch": 0.4233983286908078,
      "grad_norm": 1.5354920816773532,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8604425191879272,
      "num_tokens": 79143026.0,
      "step": 2280
    },
    {
      "epoch": 0.4235840297121634,
      "grad_norm": 1.4808260085558138,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8722240924835205,
      "num_tokens": 79175053.0,
      "step": 2281
    },
    {
      "epoch": 0.42376973073351903,
      "grad_norm": 1.3854986501202422,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8746867179870605,
      "num_tokens": 79215237.0,
      "step": 2282
    },
    {
      "epoch": 0.42395543175487466,
      "grad_norm": 1.3635402924796278,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8740991353988647,
      "num_tokens": 79254273.0,
      "step": 2283
    },
    {
      "epoch": 0.4241411327762303,
      "grad_norm": 1.5104299446042213,
      "learning_rate": 1e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8486511707305908,
      "num_tokens": 79288316.0,
      "step": 2284
    },
    {
      "epoch": 0.4243268337975859,
      "grad_norm": 1.5185106437936238,
      "learning_rate": 1e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8526186347007751,
      "num_tokens": 79323033.0,
      "step": 2285
    },
    {
      "epoch": 0.4245125348189415,
      "grad_norm": 1.4723479929216634,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8660473227500916,
      "num_tokens": 79358570.0,
      "step": 2286
    },
    {
      "epoch": 0.42469823584029714,
      "grad_norm": 1.4596273785630227,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8725574016571045,
      "num_tokens": 79392888.0,
      "step": 2287
    },
    {
      "epoch": 0.42488393686165277,
      "grad_norm": 1.4677944595784058,
      "learning_rate": 1e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8568732738494873,
      "num_tokens": 79428438.0,
      "step": 2288
    },
    {
      "epoch": 0.42506963788300833,
      "grad_norm": 1.3704756812591639,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8755836486816406,
      "num_tokens": 79465400.0,
      "step": 2289
    },
    {
      "epoch": 0.42525533890436396,
      "grad_norm": 1.486222194764676,
      "learning_rate": 1e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8510200381278992,
      "num_tokens": 79507218.0,
      "step": 2290
    },
    {
      "epoch": 0.4254410399257196,
      "grad_norm": 1.5341535472670742,
      "learning_rate": 1e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8521915674209595,
      "num_tokens": 79545172.0,
      "step": 2291
    },
    {
      "epoch": 0.4256267409470752,
      "grad_norm": 1.486723743542986,
      "learning_rate": 1e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8583415746688843,
      "num_tokens": 79580463.0,
      "step": 2292
    },
    {
      "epoch": 0.4258124419684308,
      "grad_norm": 1.5786409216200392,
      "learning_rate": 1e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8525609970092773,
      "num_tokens": 79615030.0,
      "step": 2293
    },
    {
      "epoch": 0.42599814298978644,
      "grad_norm": 1.7158674980319226,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8658081293106079,
      "num_tokens": 79646709.0,
      "step": 2294
    },
    {
      "epoch": 0.42618384401114207,
      "grad_norm": 1.531393338850271,
      "learning_rate": 1e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8484877347946167,
      "num_tokens": 79683122.0,
      "step": 2295
    },
    {
      "epoch": 0.4263695450324977,
      "grad_norm": 1.5382051319256236,
      "learning_rate": 1e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8549149036407471,
      "num_tokens": 79719471.0,
      "step": 2296
    },
    {
      "epoch": 0.4265552460538533,
      "grad_norm": 1.3360445888349555,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8655278086662292,
      "num_tokens": 79757773.0,
      "step": 2297
    },
    {
      "epoch": 0.42674094707520893,
      "grad_norm": 1.4304054515382847,
      "learning_rate": 1e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.863761842250824,
      "num_tokens": 79793536.0,
      "step": 2298
    },
    {
      "epoch": 0.42692664809656455,
      "grad_norm": 1.472188515067193,
      "learning_rate": 1e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8455477952957153,
      "num_tokens": 79834192.0,
      "step": 2299
    },
    {
      "epoch": 0.4271123491179202,
      "grad_norm": 1.6291507300297718,
      "learning_rate": 1e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8432608842849731,
      "num_tokens": 79869399.0,
      "step": 2300
    },
    {
      "epoch": 0.42729805013927574,
      "grad_norm": 1.4349505520481058,
      "learning_rate": 1e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.847250759601593,
      "num_tokens": 79911784.0,
      "step": 2301
    },
    {
      "epoch": 0.42748375116063136,
      "grad_norm": 1.5699974472903944,
      "learning_rate": 1e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8574880957603455,
      "num_tokens": 79942939.0,
      "step": 2302
    },
    {
      "epoch": 0.427669452181987,
      "grad_norm": 1.601898389459405,
      "learning_rate": 1e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8464220762252808,
      "num_tokens": 79972869.0,
      "step": 2303
    },
    {
      "epoch": 0.4278551532033426,
      "grad_norm": 1.5178511667338541,
      "learning_rate": 1e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.8584756851196289,
      "num_tokens": 80006591.0,
      "step": 2304
    },
    {
      "epoch": 0.42804085422469823,
      "grad_norm": 1.4303517552164025,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8656225204467773,
      "num_tokens": 80043119.0,
      "step": 2305
    },
    {
      "epoch": 0.42822655524605385,
      "grad_norm": 1.523274760746846,
      "learning_rate": 1e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8558973073959351,
      "num_tokens": 80076751.0,
      "step": 2306
    },
    {
      "epoch": 0.4284122562674095,
      "grad_norm": 1.5120106737873356,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8591595888137817,
      "num_tokens": 80109356.0,
      "step": 2307
    },
    {
      "epoch": 0.4285979572887651,
      "grad_norm": 1.360306280574886,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8605243563652039,
      "num_tokens": 80151792.0,
      "step": 2308
    },
    {
      "epoch": 0.4287836583101207,
      "grad_norm": 1.407389369640314,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8666136264801025,
      "num_tokens": 80192371.0,
      "step": 2309
    },
    {
      "epoch": 0.42896935933147634,
      "grad_norm": 1.5538969777490503,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8592998385429382,
      "num_tokens": 80226022.0,
      "step": 2310
    },
    {
      "epoch": 0.42915506035283196,
      "grad_norm": 1.5521297396670297,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8618929982185364,
      "num_tokens": 80257760.0,
      "step": 2311
    },
    {
      "epoch": 0.4293407613741876,
      "grad_norm": 1.4357090676708364,
      "learning_rate": 1e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8559011220932007,
      "num_tokens": 80296264.0,
      "step": 2312
    },
    {
      "epoch": 0.42952646239554315,
      "grad_norm": 1.4357156879627675,
      "learning_rate": 1e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8556232452392578,
      "num_tokens": 80336518.0,
      "step": 2313
    },
    {
      "epoch": 0.4297121634168988,
      "grad_norm": 1.390658975704075,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8715778589248657,
      "num_tokens": 80373051.0,
      "step": 2314
    },
    {
      "epoch": 0.4298978644382544,
      "grad_norm": 1.5299327921066548,
      "learning_rate": 1e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8623397350311279,
      "num_tokens": 80405669.0,
      "step": 2315
    },
    {
      "epoch": 0.43008356545961,
      "grad_norm": 1.412662072763242,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8665060997009277,
      "num_tokens": 80442386.0,
      "step": 2316
    },
    {
      "epoch": 0.43026926648096564,
      "grad_norm": 1.4693083076330387,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8651904463768005,
      "num_tokens": 80475013.0,
      "step": 2317
    },
    {
      "epoch": 0.43045496750232126,
      "grad_norm": 1.4287722805213168,
      "learning_rate": 1e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8534599542617798,
      "num_tokens": 80515621.0,
      "step": 2318
    },
    {
      "epoch": 0.4306406685236769,
      "grad_norm": 1.4787787875091505,
      "learning_rate": 1e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8578972220420837,
      "num_tokens": 80549087.0,
      "step": 2319
    },
    {
      "epoch": 0.4308263695450325,
      "grad_norm": 1.5259963068509357,
      "learning_rate": 1e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8538442850112915,
      "num_tokens": 80582465.0,
      "step": 2320
    },
    {
      "epoch": 0.43101207056638813,
      "grad_norm": 1.5288090008383657,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8704679608345032,
      "num_tokens": 80615551.0,
      "step": 2321
    },
    {
      "epoch": 0.43119777158774375,
      "grad_norm": 1.569984730601373,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8726828098297119,
      "num_tokens": 80642753.0,
      "step": 2322
    },
    {
      "epoch": 0.43138347260909937,
      "grad_norm": 1.6545425457051233,
      "learning_rate": 1e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8510053157806396,
      "num_tokens": 80673043.0,
      "step": 2323
    },
    {
      "epoch": 0.431569173630455,
      "grad_norm": 1.4838888130060326,
      "learning_rate": 1e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8559755682945251,
      "num_tokens": 80708789.0,
      "step": 2324
    },
    {
      "epoch": 0.43175487465181056,
      "grad_norm": 1.4084421673771128,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8689358234405518,
      "num_tokens": 80742890.0,
      "step": 2325
    },
    {
      "epoch": 0.4319405756731662,
      "grad_norm": 1.6125016375564414,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8703945875167847,
      "num_tokens": 80776440.0,
      "step": 2326
    },
    {
      "epoch": 0.4321262766945218,
      "grad_norm": 1.6212705750412175,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8630422949790955,
      "num_tokens": 80809223.0,
      "step": 2327
    },
    {
      "epoch": 0.4323119777158774,
      "grad_norm": 1.5240329333329774,
      "learning_rate": 1e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8332990407943726,
      "num_tokens": 80848943.0,
      "step": 2328
    },
    {
      "epoch": 0.43249767873723305,
      "grad_norm": 1.6687901483482115,
      "learning_rate": 1e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8404780626296997,
      "num_tokens": 80880816.0,
      "step": 2329
    },
    {
      "epoch": 0.43268337975858867,
      "grad_norm": 1.56153426545346,
      "learning_rate": 1e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.861869215965271,
      "num_tokens": 80912678.0,
      "step": 2330
    },
    {
      "epoch": 0.4328690807799443,
      "grad_norm": 1.5473052591020962,
      "learning_rate": 1e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8547817468643188,
      "num_tokens": 80948784.0,
      "step": 2331
    },
    {
      "epoch": 0.4330547818012999,
      "grad_norm": 1.4137391784382436,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8777136206626892,
      "num_tokens": 80987244.0,
      "step": 2332
    },
    {
      "epoch": 0.43324048282265554,
      "grad_norm": 1.4383032744494546,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.867835521697998,
      "num_tokens": 81023318.0,
      "step": 2333
    },
    {
      "epoch": 0.43342618384401116,
      "grad_norm": 1.4959028392205334,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8597210049629211,
      "num_tokens": 81058203.0,
      "step": 2334
    },
    {
      "epoch": 0.4336118848653668,
      "grad_norm": 1.5095491584434764,
      "learning_rate": 1e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8574098348617554,
      "num_tokens": 81091354.0,
      "step": 2335
    },
    {
      "epoch": 0.4337975858867224,
      "grad_norm": 1.526594384615793,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8630340099334717,
      "num_tokens": 81124548.0,
      "step": 2336
    },
    {
      "epoch": 0.43398328690807797,
      "grad_norm": 1.5725841859187994,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8649923205375671,
      "num_tokens": 81156824.0,
      "step": 2337
    },
    {
      "epoch": 0.4341689879294336,
      "grad_norm": 1.5616660140563277,
      "learning_rate": 1e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8538299798965454,
      "num_tokens": 81189078.0,
      "step": 2338
    },
    {
      "epoch": 0.4343546889507892,
      "grad_norm": 1.5243384533911524,
      "learning_rate": 1e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8474977016448975,
      "num_tokens": 81227187.0,
      "step": 2339
    },
    {
      "epoch": 0.43454038997214484,
      "grad_norm": 1.5540884831435156,
      "learning_rate": 1e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8463646173477173,
      "num_tokens": 81261074.0,
      "step": 2340
    },
    {
      "epoch": 0.43472609099350046,
      "grad_norm": 1.6008344939544012,
      "learning_rate": 1e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8530848026275635,
      "num_tokens": 81290549.0,
      "step": 2341
    },
    {
      "epoch": 0.4349117920148561,
      "grad_norm": 1.4601975646764118,
      "learning_rate": 1e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.8499513268470764,
      "num_tokens": 81330605.0,
      "step": 2342
    },
    {
      "epoch": 0.4350974930362117,
      "grad_norm": 1.459164815208732,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8677460551261902,
      "num_tokens": 81365825.0,
      "step": 2343
    },
    {
      "epoch": 0.4352831940575673,
      "grad_norm": 1.4456414484883056,
      "learning_rate": 1e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.853097140789032,
      "num_tokens": 81407105.0,
      "step": 2344
    },
    {
      "epoch": 0.43546889507892295,
      "grad_norm": 1.6572223443341896,
      "learning_rate": 1e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8445883989334106,
      "num_tokens": 81437456.0,
      "step": 2345
    },
    {
      "epoch": 0.43565459610027857,
      "grad_norm": 1.516190555351185,
      "learning_rate": 1e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8630049228668213,
      "num_tokens": 81470597.0,
      "step": 2346
    },
    {
      "epoch": 0.4358402971216342,
      "grad_norm": 1.4776197117624048,
      "learning_rate": 1e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8412061333656311,
      "num_tokens": 81508249.0,
      "step": 2347
    },
    {
      "epoch": 0.4360259981429898,
      "grad_norm": 1.563461844526257,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8632361888885498,
      "num_tokens": 81538948.0,
      "step": 2348
    },
    {
      "epoch": 0.4362116991643454,
      "grad_norm": 1.5395205354432935,
      "learning_rate": 1e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8635945320129395,
      "num_tokens": 81572301.0,
      "step": 2349
    },
    {
      "epoch": 0.436397400185701,
      "grad_norm": 1.471372557142393,
      "learning_rate": 1e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8578802347183228,
      "num_tokens": 81607925.0,
      "step": 2350
    },
    {
      "epoch": 0.4365831012070566,
      "grad_norm": 1.6203265934498041,
      "learning_rate": 1e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8450247645378113,
      "num_tokens": 81637847.0,
      "step": 2351
    },
    {
      "epoch": 0.43676880222841225,
      "grad_norm": 1.5206134695410158,
      "learning_rate": 1e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8468664884567261,
      "num_tokens": 81672153.0,
      "step": 2352
    },
    {
      "epoch": 0.43695450324976787,
      "grad_norm": 1.4258850248455692,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8584259748458862,
      "num_tokens": 81708556.0,
      "step": 2353
    },
    {
      "epoch": 0.4371402042711235,
      "grad_norm": 1.3781740688805186,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.875048041343689,
      "num_tokens": 81748338.0,
      "step": 2354
    },
    {
      "epoch": 0.4373259052924791,
      "grad_norm": 1.5882901522364334,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.871142566204071,
      "num_tokens": 81776535.0,
      "step": 2355
    },
    {
      "epoch": 0.43751160631383473,
      "grad_norm": 1.5785374037315243,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8641701936721802,
      "num_tokens": 81810692.0,
      "step": 2356
    },
    {
      "epoch": 0.43769730733519036,
      "grad_norm": 1.492031834542965,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8563143014907837,
      "num_tokens": 81846187.0,
      "step": 2357
    },
    {
      "epoch": 0.437883008356546,
      "grad_norm": 1.4110280966063364,
      "learning_rate": 1e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8652307987213135,
      "num_tokens": 81884498.0,
      "step": 2358
    },
    {
      "epoch": 0.4380687093779016,
      "grad_norm": 1.4861891810661485,
      "learning_rate": 1e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8587262034416199,
      "num_tokens": 81921932.0,
      "step": 2359
    },
    {
      "epoch": 0.4382544103992572,
      "grad_norm": 1.4781726212512745,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8684132099151611,
      "num_tokens": 81960530.0,
      "step": 2360
    },
    {
      "epoch": 0.4384401114206128,
      "grad_norm": 1.4544670814196798,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8617493510246277,
      "num_tokens": 81996567.0,
      "step": 2361
    },
    {
      "epoch": 0.4386258124419684,
      "grad_norm": 1.5283248373513945,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8650545477867126,
      "num_tokens": 82030227.0,
      "step": 2362
    },
    {
      "epoch": 0.43881151346332403,
      "grad_norm": 1.4656167694202225,
      "learning_rate": 1e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8586056232452393,
      "num_tokens": 82068792.0,
      "step": 2363
    },
    {
      "epoch": 0.43899721448467965,
      "grad_norm": 1.545537700112552,
      "learning_rate": 1e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8543493747711182,
      "num_tokens": 82103463.0,
      "step": 2364
    },
    {
      "epoch": 0.4391829155060353,
      "grad_norm": 1.4751039483810464,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8627803921699524,
      "num_tokens": 82138749.0,
      "step": 2365
    },
    {
      "epoch": 0.4393686165273909,
      "grad_norm": 1.414569447617983,
      "learning_rate": 1e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8644740581512451,
      "num_tokens": 82174705.0,
      "step": 2366
    },
    {
      "epoch": 0.4395543175487465,
      "grad_norm": 1.4722208966914032,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8556015491485596,
      "num_tokens": 82210183.0,
      "step": 2367
    },
    {
      "epoch": 0.43974001857010214,
      "grad_norm": 1.6250560788344999,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8646914958953857,
      "num_tokens": 82242874.0,
      "step": 2368
    },
    {
      "epoch": 0.43992571959145776,
      "grad_norm": 1.4022534727022036,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8699123859405518,
      "num_tokens": 82282830.0,
      "step": 2369
    },
    {
      "epoch": 0.4401114206128134,
      "grad_norm": 1.3023667360687086,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8594194054603577,
      "num_tokens": 82325486.0,
      "step": 2370
    },
    {
      "epoch": 0.440297121634169,
      "grad_norm": 1.5070341942814107,
      "learning_rate": 1e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8598855137825012,
      "num_tokens": 82359063.0,
      "step": 2371
    },
    {
      "epoch": 0.44048282265552463,
      "grad_norm": 1.4250860500157763,
      "learning_rate": 1e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8596233129501343,
      "num_tokens": 82400600.0,
      "step": 2372
    },
    {
      "epoch": 0.4406685236768802,
      "grad_norm": 1.5307862466590842,
      "learning_rate": 1e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8585370779037476,
      "num_tokens": 82437913.0,
      "step": 2373
    },
    {
      "epoch": 0.4408542246982358,
      "grad_norm": 1.4936380735339287,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8678299188613892,
      "num_tokens": 82473293.0,
      "step": 2374
    },
    {
      "epoch": 0.44103992571959144,
      "grad_norm": 1.5919751641724258,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8658573627471924,
      "num_tokens": 82507100.0,
      "step": 2375
    },
    {
      "epoch": 0.44122562674094706,
      "grad_norm": 1.5917963916957885,
      "learning_rate": 1e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8667043447494507,
      "num_tokens": 82536240.0,
      "step": 2376
    },
    {
      "epoch": 0.4414113277623027,
      "grad_norm": 1.491644156299597,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8665394186973572,
      "num_tokens": 82571674.0,
      "step": 2377
    },
    {
      "epoch": 0.4415970287836583,
      "grad_norm": 1.5422401478764738,
      "learning_rate": 1e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8650404214859009,
      "num_tokens": 82607320.0,
      "step": 2378
    },
    {
      "epoch": 0.44178272980501393,
      "grad_norm": 1.6250302108320716,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8599322438240051,
      "num_tokens": 82636045.0,
      "step": 2379
    },
    {
      "epoch": 0.44196843082636955,
      "grad_norm": 1.488199977176625,
      "learning_rate": 1e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8460464477539062,
      "num_tokens": 82677288.0,
      "step": 2380
    },
    {
      "epoch": 0.4421541318477252,
      "grad_norm": 1.6016617794156465,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8716485500335693,
      "num_tokens": 82706382.0,
      "step": 2381
    },
    {
      "epoch": 0.4423398328690808,
      "grad_norm": 1.4906730538143151,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8619623184204102,
      "num_tokens": 82740276.0,
      "step": 2382
    },
    {
      "epoch": 0.4425255338904364,
      "grad_norm": 1.4410114656868735,
      "learning_rate": 1e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8530055284500122,
      "num_tokens": 82778580.0,
      "step": 2383
    },
    {
      "epoch": 0.44271123491179204,
      "grad_norm": 1.3740270931929945,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8691650032997131,
      "num_tokens": 82819834.0,
      "step": 2384
    },
    {
      "epoch": 0.4428969359331476,
      "grad_norm": 1.516430422457254,
      "learning_rate": 1e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8493658304214478,
      "num_tokens": 82857646.0,
      "step": 2385
    },
    {
      "epoch": 0.44308263695450323,
      "grad_norm": 1.5177463986804098,
      "learning_rate": 1e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8505808115005493,
      "num_tokens": 82892368.0,
      "step": 2386
    },
    {
      "epoch": 0.44326833797585885,
      "grad_norm": 1.4489530741330277,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8602178692817688,
      "num_tokens": 82928617.0,
      "step": 2387
    },
    {
      "epoch": 0.4434540389972145,
      "grad_norm": 1.425938009681025,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8542685508728027,
      "num_tokens": 82969175.0,
      "step": 2388
    },
    {
      "epoch": 0.4436397400185701,
      "grad_norm": 1.3608897368030706,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8685262799263,
      "num_tokens": 83005789.0,
      "step": 2389
    },
    {
      "epoch": 0.4438254410399257,
      "grad_norm": 1.5178665988927018,
      "learning_rate": 1e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8661348819732666,
      "num_tokens": 83039627.0,
      "step": 2390
    },
    {
      "epoch": 0.44401114206128134,
      "grad_norm": 1.6145728183229848,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8685206174850464,
      "num_tokens": 83068142.0,
      "step": 2391
    },
    {
      "epoch": 0.44419684308263696,
      "grad_norm": 1.570736534503171,
      "learning_rate": 1e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8553346395492554,
      "num_tokens": 83102839.0,
      "step": 2392
    },
    {
      "epoch": 0.4443825441039926,
      "grad_norm": 1.5188270420453707,
      "learning_rate": 1e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.855556845664978,
      "num_tokens": 83137228.0,
      "step": 2393
    },
    {
      "epoch": 0.4445682451253482,
      "grad_norm": 1.5340779165685663,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8574926257133484,
      "num_tokens": 83168115.0,
      "step": 2394
    },
    {
      "epoch": 0.4447539461467038,
      "grad_norm": 1.553866491820242,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8713610768318176,
      "num_tokens": 83198393.0,
      "step": 2395
    },
    {
      "epoch": 0.44493964716805945,
      "grad_norm": 1.471741224762646,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8719065189361572,
      "num_tokens": 83233906.0,
      "step": 2396
    },
    {
      "epoch": 0.445125348189415,
      "grad_norm": 1.5286534894428783,
      "learning_rate": 1e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8534731268882751,
      "num_tokens": 83271413.0,
      "step": 2397
    },
    {
      "epoch": 0.44531104921077064,
      "grad_norm": 1.55344333750497,
      "learning_rate": 1e-06,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8517454862594604,
      "num_tokens": 83305989.0,
      "step": 2398
    },
    {
      "epoch": 0.44549675023212626,
      "grad_norm": 1.4366556747889878,
      "learning_rate": 1e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.847830057144165,
      "num_tokens": 83350794.0,
      "step": 2399
    },
    {
      "epoch": 0.4456824512534819,
      "grad_norm": 1.5302936090319297,
      "learning_rate": 1e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8582032322883606,
      "num_tokens": 83385398.0,
      "step": 2400
    },
    {
      "epoch": 0.4458681522748375,
      "grad_norm": 1.4306378302309057,
      "learning_rate": 1e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8628529906272888,
      "num_tokens": 83421286.0,
      "step": 2401
    },
    {
      "epoch": 0.4460538532961931,
      "grad_norm": 1.4259720719708275,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8744399547576904,
      "num_tokens": 83457503.0,
      "step": 2402
    },
    {
      "epoch": 0.44623955431754875,
      "grad_norm": 1.5565621977395907,
      "learning_rate": 1e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8428962826728821,
      "num_tokens": 83491064.0,
      "step": 2403
    },
    {
      "epoch": 0.44642525533890437,
      "grad_norm": 1.4011046993659355,
      "learning_rate": 1e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8512422442436218,
      "num_tokens": 83528668.0,
      "step": 2404
    },
    {
      "epoch": 0.44661095636026,
      "grad_norm": 1.4610544614755139,
      "learning_rate": 1e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8547447323799133,
      "num_tokens": 83566340.0,
      "step": 2405
    },
    {
      "epoch": 0.4467966573816156,
      "grad_norm": 1.377868217699112,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8659764528274536,
      "num_tokens": 83606775.0,
      "step": 2406
    },
    {
      "epoch": 0.44698235840297124,
      "grad_norm": 1.5500038281881738,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8682976961135864,
      "num_tokens": 83641369.0,
      "step": 2407
    },
    {
      "epoch": 0.44716805942432686,
      "grad_norm": 1.721995781993472,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.847470760345459,
      "num_tokens": 83668883.0,
      "step": 2408
    },
    {
      "epoch": 0.4473537604456824,
      "grad_norm": 1.4611486114034167,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8738563060760498,
      "num_tokens": 83702897.0,
      "step": 2409
    },
    {
      "epoch": 0.44753946146703805,
      "grad_norm": 1.449864157778422,
      "learning_rate": 1e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8607712984085083,
      "num_tokens": 83739615.0,
      "step": 2410
    },
    {
      "epoch": 0.44772516248839367,
      "grad_norm": 1.3902711742566995,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8605151176452637,
      "num_tokens": 83777075.0,
      "step": 2411
    },
    {
      "epoch": 0.4479108635097493,
      "grad_norm": 1.436223422099499,
      "learning_rate": 1e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.854272186756134,
      "num_tokens": 83813115.0,
      "step": 2412
    },
    {
      "epoch": 0.4480965645311049,
      "grad_norm": 1.550922308643031,
      "learning_rate": 1e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8597532510757446,
      "num_tokens": 83844791.0,
      "step": 2413
    },
    {
      "epoch": 0.44828226555246053,
      "grad_norm": 1.470277333942769,
      "learning_rate": 1e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8546684384346008,
      "num_tokens": 83881645.0,
      "step": 2414
    },
    {
      "epoch": 0.44846796657381616,
      "grad_norm": 1.4387320556297276,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8725839853286743,
      "num_tokens": 83918642.0,
      "step": 2415
    },
    {
      "epoch": 0.4486536675951718,
      "grad_norm": 1.582698795220128,
      "learning_rate": 1e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8526350259780884,
      "num_tokens": 83954423.0,
      "step": 2416
    },
    {
      "epoch": 0.4488393686165274,
      "grad_norm": 1.5806511624502002,
      "learning_rate": 1e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8523461222648621,
      "num_tokens": 83987156.0,
      "step": 2417
    },
    {
      "epoch": 0.449025069637883,
      "grad_norm": 1.462171122721557,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8698694705963135,
      "num_tokens": 84020697.0,
      "step": 2418
    },
    {
      "epoch": 0.44921077065923865,
      "grad_norm": 1.4206804251360647,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8709107637405396,
      "num_tokens": 84057046.0,
      "step": 2419
    },
    {
      "epoch": 0.44939647168059427,
      "grad_norm": 1.5490052527042562,
      "learning_rate": 1e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8571087121963501,
      "num_tokens": 84088151.0,
      "step": 2420
    },
    {
      "epoch": 0.44958217270194983,
      "grad_norm": 1.5617611554639261,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8564088940620422,
      "num_tokens": 84124788.0,
      "step": 2421
    },
    {
      "epoch": 0.44976787372330546,
      "grad_norm": 1.4339264409614538,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8727940320968628,
      "num_tokens": 84157185.0,
      "step": 2422
    },
    {
      "epoch": 0.4499535747446611,
      "grad_norm": 1.4598004992250435,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8545130491256714,
      "num_tokens": 84194277.0,
      "step": 2423
    },
    {
      "epoch": 0.4501392757660167,
      "grad_norm": 1.4445489056676057,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8771897554397583,
      "num_tokens": 84228240.0,
      "step": 2424
    },
    {
      "epoch": 0.4503249767873723,
      "grad_norm": 1.365253291325308,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8774451017379761,
      "num_tokens": 84260616.0,
      "step": 2425
    },
    {
      "epoch": 0.45051067780872794,
      "grad_norm": 1.5477839048408442,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8657835721969604,
      "num_tokens": 84291928.0,
      "step": 2426
    },
    {
      "epoch": 0.45069637883008357,
      "grad_norm": 1.5245569098839566,
      "learning_rate": 1e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8508906364440918,
      "num_tokens": 84327147.0,
      "step": 2427
    },
    {
      "epoch": 0.4508820798514392,
      "grad_norm": 1.4440432085185255,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8600913286209106,
      "num_tokens": 84362750.0,
      "step": 2428
    },
    {
      "epoch": 0.4510677808727948,
      "grad_norm": 1.4610036232609773,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8650171160697937,
      "num_tokens": 84398467.0,
      "step": 2429
    },
    {
      "epoch": 0.45125348189415043,
      "grad_norm": 1.3402026017480713,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8656243085861206,
      "num_tokens": 84443378.0,
      "step": 2430
    },
    {
      "epoch": 0.45143918291550605,
      "grad_norm": 1.5059452426201496,
      "learning_rate": 1e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8646207451820374,
      "num_tokens": 84476886.0,
      "step": 2431
    },
    {
      "epoch": 0.4516248839368617,
      "grad_norm": 1.425968983223347,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8645361065864563,
      "num_tokens": 84515254.0,
      "step": 2432
    },
    {
      "epoch": 0.45181058495821724,
      "grad_norm": 1.5953064618117638,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8694910407066345,
      "num_tokens": 84543290.0,
      "step": 2433
    },
    {
      "epoch": 0.45199628597957286,
      "grad_norm": 1.6038282530894414,
      "learning_rate": 1e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8416696190834045,
      "num_tokens": 84577876.0,
      "step": 2434
    },
    {
      "epoch": 0.4521819870009285,
      "grad_norm": 1.3810390011175124,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8650912642478943,
      "num_tokens": 84619385.0,
      "step": 2435
    },
    {
      "epoch": 0.4523676880222841,
      "grad_norm": 1.6075770804589504,
      "learning_rate": 1e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8484313488006592,
      "num_tokens": 84653353.0,
      "step": 2436
    },
    {
      "epoch": 0.45255338904363973,
      "grad_norm": 1.4776818954785784,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8584388494491577,
      "num_tokens": 84690466.0,
      "step": 2437
    },
    {
      "epoch": 0.45273909006499535,
      "grad_norm": 1.362350732324496,
      "learning_rate": 1e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8580453395843506,
      "num_tokens": 84730760.0,
      "step": 2438
    },
    {
      "epoch": 0.452924791086351,
      "grad_norm": 1.42099911044926,
      "learning_rate": 1e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.855067253112793,
      "num_tokens": 84770782.0,
      "step": 2439
    },
    {
      "epoch": 0.4531104921077066,
      "grad_norm": 1.3552834561157667,
      "learning_rate": 1e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8620176315307617,
      "num_tokens": 84810625.0,
      "step": 2440
    },
    {
      "epoch": 0.4532961931290622,
      "grad_norm": 1.5794771343685914,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8694778680801392,
      "num_tokens": 84840366.0,
      "step": 2441
    },
    {
      "epoch": 0.45348189415041784,
      "grad_norm": 1.5462037288229356,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8639078736305237,
      "num_tokens": 84869796.0,
      "step": 2442
    },
    {
      "epoch": 0.45366759517177346,
      "grad_norm": 1.5189912477872387,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8581048250198364,
      "num_tokens": 84905969.0,
      "step": 2443
    },
    {
      "epoch": 0.4538532961931291,
      "grad_norm": 1.4821456488077611,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8586875796318054,
      "num_tokens": 84942745.0,
      "step": 2444
    },
    {
      "epoch": 0.45403899721448465,
      "grad_norm": 1.4701575858585998,
      "learning_rate": 1e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8448127508163452,
      "num_tokens": 84980006.0,
      "step": 2445
    },
    {
      "epoch": 0.4542246982358403,
      "grad_norm": 1.3785152734908979,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8684154152870178,
      "num_tokens": 85018609.0,
      "step": 2446
    },
    {
      "epoch": 0.4544103992571959,
      "grad_norm": 1.4806216843791244,
      "learning_rate": 1e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8551012277603149,
      "num_tokens": 85055036.0,
      "step": 2447
    },
    {
      "epoch": 0.4545961002785515,
      "grad_norm": 1.5135400994391963,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.86483234167099,
      "num_tokens": 85087518.0,
      "step": 2448
    },
    {
      "epoch": 0.45478180129990714,
      "grad_norm": 1.4626804102022488,
      "learning_rate": 1e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8482259511947632,
      "num_tokens": 85126754.0,
      "step": 2449
    },
    {
      "epoch": 0.45496750232126276,
      "grad_norm": 1.48158891782079,
      "learning_rate": 1e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8598620295524597,
      "num_tokens": 85165497.0,
      "step": 2450
    },
    {
      "epoch": 0.4551532033426184,
      "grad_norm": 1.4535350759104106,
      "learning_rate": 1e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8636742234230042,
      "num_tokens": 85203200.0,
      "step": 2451
    },
    {
      "epoch": 0.455338904363974,
      "grad_norm": 1.4858047178821614,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8647260069847107,
      "num_tokens": 85238784.0,
      "step": 2452
    },
    {
      "epoch": 0.45552460538532963,
      "grad_norm": 1.9117215493724806,
      "learning_rate": 1e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8545464277267456,
      "num_tokens": 85262787.0,
      "step": 2453
    },
    {
      "epoch": 0.45571030640668525,
      "grad_norm": 1.5141960606832587,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8597227931022644,
      "num_tokens": 85296920.0,
      "step": 2454
    },
    {
      "epoch": 0.4558960074280409,
      "grad_norm": 1.4823524565074226,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8632588386535645,
      "num_tokens": 85332495.0,
      "step": 2455
    },
    {
      "epoch": 0.4560817084493965,
      "grad_norm": 1.4851506796404628,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8681579232215881,
      "num_tokens": 85366594.0,
      "step": 2456
    },
    {
      "epoch": 0.4562674094707521,
      "grad_norm": 1.4420692017945098,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8866878747940063,
      "num_tokens": 85399139.0,
      "step": 2457
    },
    {
      "epoch": 0.4564531104921077,
      "grad_norm": 1.594209813806763,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8629128932952881,
      "num_tokens": 85431933.0,
      "step": 2458
    },
    {
      "epoch": 0.4566388115134633,
      "grad_norm": 1.5246071783565003,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8634209632873535,
      "num_tokens": 85465239.0,
      "step": 2459
    },
    {
      "epoch": 0.4568245125348189,
      "grad_norm": 1.4639157935812896,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8672546148300171,
      "num_tokens": 85504976.0,
      "step": 2460
    },
    {
      "epoch": 0.45701021355617455,
      "grad_norm": 1.5257749949676955,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8637496829032898,
      "num_tokens": 85537920.0,
      "step": 2461
    },
    {
      "epoch": 0.45719591457753017,
      "grad_norm": 1.5963639170505415,
      "learning_rate": 1e-06,
      "loss": 0.41,
      "mean_token_accuracy": 0.8627279996871948,
      "num_tokens": 85568524.0,
      "step": 2462
    },
    {
      "epoch": 0.4573816155988858,
      "grad_norm": 1.4487710839686427,
      "learning_rate": 1e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8578687906265259,
      "num_tokens": 85604035.0,
      "step": 2463
    },
    {
      "epoch": 0.4575673166202414,
      "grad_norm": 1.485950367022051,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8604142665863037,
      "num_tokens": 85638677.0,
      "step": 2464
    },
    {
      "epoch": 0.45775301764159704,
      "grad_norm": 1.4227960065321679,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8594633340835571,
      "num_tokens": 85676384.0,
      "step": 2465
    },
    {
      "epoch": 0.45793871866295266,
      "grad_norm": 1.5259431842081093,
      "learning_rate": 1e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8519587516784668,
      "num_tokens": 85710300.0,
      "step": 2466
    },
    {
      "epoch": 0.4581244196843083,
      "grad_norm": 1.5635246060207322,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8743394017219543,
      "num_tokens": 85742512.0,
      "step": 2467
    },
    {
      "epoch": 0.4583101207056639,
      "grad_norm": 1.6025131229912717,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8745912909507751,
      "num_tokens": 85769951.0,
      "step": 2468
    },
    {
      "epoch": 0.4584958217270195,
      "grad_norm": 1.4331232242054441,
      "learning_rate": 1e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8604668974876404,
      "num_tokens": 85809249.0,
      "step": 2469
    },
    {
      "epoch": 0.4586815227483751,
      "grad_norm": 1.441354156046377,
      "learning_rate": 1e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8600295782089233,
      "num_tokens": 85844457.0,
      "step": 2470
    },
    {
      "epoch": 0.4588672237697307,
      "grad_norm": 1.3583834755573665,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8667706251144409,
      "num_tokens": 85885233.0,
      "step": 2471
    },
    {
      "epoch": 0.45905292479108634,
      "grad_norm": 1.419545847486547,
      "learning_rate": 1e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8664299249649048,
      "num_tokens": 85921409.0,
      "step": 2472
    },
    {
      "epoch": 0.45923862581244196,
      "grad_norm": 1.5135814519531428,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8733474016189575,
      "num_tokens": 85955843.0,
      "step": 2473
    },
    {
      "epoch": 0.4594243268337976,
      "grad_norm": 1.6502310235887276,
      "learning_rate": 1e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8420697450637817,
      "num_tokens": 85989720.0,
      "step": 2474
    },
    {
      "epoch": 0.4596100278551532,
      "grad_norm": 1.5832412990124793,
      "learning_rate": 1e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.849163293838501,
      "num_tokens": 86020518.0,
      "step": 2475
    },
    {
      "epoch": 0.4597957288765088,
      "grad_norm": 1.455810938515791,
      "learning_rate": 1e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8620474338531494,
      "num_tokens": 86057893.0,
      "step": 2476
    },
    {
      "epoch": 0.45998142989786445,
      "grad_norm": 1.553491224030847,
      "learning_rate": 1e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8437814712524414,
      "num_tokens": 86091537.0,
      "step": 2477
    },
    {
      "epoch": 0.46016713091922007,
      "grad_norm": 1.431818012041923,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8703604936599731,
      "num_tokens": 86123492.0,
      "step": 2478
    },
    {
      "epoch": 0.4603528319405757,
      "grad_norm": 1.4780630720625085,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8536795377731323,
      "num_tokens": 86160324.0,
      "step": 2479
    },
    {
      "epoch": 0.4605385329619313,
      "grad_norm": 1.533995157262253,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8673454523086548,
      "num_tokens": 86190544.0,
      "step": 2480
    },
    {
      "epoch": 0.46072423398328693,
      "grad_norm": 1.3945337421123996,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8778132200241089,
      "num_tokens": 86227317.0,
      "step": 2481
    },
    {
      "epoch": 0.4609099350046425,
      "grad_norm": 1.3542629146674812,
      "learning_rate": 1e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.866313099861145,
      "num_tokens": 86265358.0,
      "step": 2482
    },
    {
      "epoch": 0.4610956360259981,
      "grad_norm": 1.565417580530056,
      "learning_rate": 1e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8515198230743408,
      "num_tokens": 86299903.0,
      "step": 2483
    },
    {
      "epoch": 0.46128133704735375,
      "grad_norm": 1.49336222755073,
      "learning_rate": 1e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8473951816558838,
      "num_tokens": 86338585.0,
      "step": 2484
    },
    {
      "epoch": 0.46146703806870937,
      "grad_norm": 1.5446655381187744,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8578715920448303,
      "num_tokens": 86374957.0,
      "step": 2485
    },
    {
      "epoch": 0.461652739090065,
      "grad_norm": 1.3720389061550948,
      "learning_rate": 1e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.860403299331665,
      "num_tokens": 86415810.0,
      "step": 2486
    },
    {
      "epoch": 0.4618384401114206,
      "grad_norm": 1.732758628686103,
      "learning_rate": 1e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8483306169509888,
      "num_tokens": 86444738.0,
      "step": 2487
    },
    {
      "epoch": 0.46202414113277623,
      "grad_norm": 1.6138309332930236,
      "learning_rate": 1e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8564858436584473,
      "num_tokens": 86477704.0,
      "step": 2488
    },
    {
      "epoch": 0.46220984215413186,
      "grad_norm": 1.584738386852436,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8786989450454712,
      "num_tokens": 86506636.0,
      "step": 2489
    },
    {
      "epoch": 0.4623955431754875,
      "grad_norm": 1.4587346271158697,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8648675680160522,
      "num_tokens": 86542909.0,
      "step": 2490
    },
    {
      "epoch": 0.4625812441968431,
      "grad_norm": 1.5649876181899693,
      "learning_rate": 1e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.854710578918457,
      "num_tokens": 86575940.0,
      "step": 2491
    },
    {
      "epoch": 0.4627669452181987,
      "grad_norm": 1.3600085672470534,
      "learning_rate": 1e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8444321155548096,
      "num_tokens": 86619631.0,
      "step": 2492
    },
    {
      "epoch": 0.46295264623955434,
      "grad_norm": 1.613150105305489,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8682142496109009,
      "num_tokens": 86647772.0,
      "step": 2493
    },
    {
      "epoch": 0.4631383472609099,
      "grad_norm": 1.4760413889486075,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8585246801376343,
      "num_tokens": 86682230.0,
      "step": 2494
    },
    {
      "epoch": 0.46332404828226553,
      "grad_norm": 1.6043692634257258,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8596270680427551,
      "num_tokens": 86713916.0,
      "step": 2495
    },
    {
      "epoch": 0.46350974930362115,
      "grad_norm": 1.5633557992979774,
      "learning_rate": 1e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8613964915275574,
      "num_tokens": 86744815.0,
      "step": 2496
    },
    {
      "epoch": 0.4636954503249768,
      "grad_norm": 1.594841914503663,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.868000864982605,
      "num_tokens": 86779060.0,
      "step": 2497
    },
    {
      "epoch": 0.4638811513463324,
      "grad_norm": 1.5051479175195093,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8594794273376465,
      "num_tokens": 86811335.0,
      "step": 2498
    },
    {
      "epoch": 0.464066852367688,
      "grad_norm": 1.571428382455872,
      "learning_rate": 1e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8459563255310059,
      "num_tokens": 86842496.0,
      "step": 2499
    },
    {
      "epoch": 0.46425255338904364,
      "grad_norm": 1.5152399187869419,
      "learning_rate": 1e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8622211217880249,
      "num_tokens": 86877680.0,
      "step": 2500
    },
    {
      "epoch": 0.46443825441039926,
      "grad_norm": 1.4817671373127577,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.87050461769104,
      "num_tokens": 86916094.0,
      "step": 2501
    },
    {
      "epoch": 0.4646239554317549,
      "grad_norm": 1.3958571873451648,
      "learning_rate": 1e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8586663603782654,
      "num_tokens": 86955852.0,
      "step": 2502
    },
    {
      "epoch": 0.4648096564531105,
      "grad_norm": 1.5547222014215811,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8663923740386963,
      "num_tokens": 86989539.0,
      "step": 2503
    },
    {
      "epoch": 0.46499535747446613,
      "grad_norm": 1.6085043030427069,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8713070154190063,
      "num_tokens": 87019657.0,
      "step": 2504
    },
    {
      "epoch": 0.46518105849582175,
      "grad_norm": 1.4138875600747225,
      "learning_rate": 1e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8488062620162964,
      "num_tokens": 87058650.0,
      "step": 2505
    },
    {
      "epoch": 0.4653667595171773,
      "grad_norm": 1.4441423095753734,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.872090220451355,
      "num_tokens": 87095207.0,
      "step": 2506
    },
    {
      "epoch": 0.46555246053853294,
      "grad_norm": 1.4333153213471805,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.872552216053009,
      "num_tokens": 87136562.0,
      "step": 2507
    },
    {
      "epoch": 0.46573816155988856,
      "grad_norm": 1.5898020765097185,
      "learning_rate": 1e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.838456392288208,
      "num_tokens": 87168356.0,
      "step": 2508
    },
    {
      "epoch": 0.4659238625812442,
      "grad_norm": 1.5738333347719895,
      "learning_rate": 1e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.847753643989563,
      "num_tokens": 87204049.0,
      "step": 2509
    },
    {
      "epoch": 0.4661095636025998,
      "grad_norm": 1.5264082379186223,
      "learning_rate": 1e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.852132260799408,
      "num_tokens": 87241184.0,
      "step": 2510
    },
    {
      "epoch": 0.46629526462395543,
      "grad_norm": 1.5320472823605111,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8744208812713623,
      "num_tokens": 87274426.0,
      "step": 2511
    },
    {
      "epoch": 0.46648096564531105,
      "grad_norm": 1.3849923329025724,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8704288601875305,
      "num_tokens": 87313847.0,
      "step": 2512
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 1.5686480829072411,
      "learning_rate": 1e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8556408286094666,
      "num_tokens": 87349807.0,
      "step": 2513
    },
    {
      "epoch": 0.4668523676880223,
      "grad_norm": 1.474152075527582,
      "learning_rate": 1e-06,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8560962677001953,
      "num_tokens": 87385360.0,
      "step": 2514
    },
    {
      "epoch": 0.4670380687093779,
      "grad_norm": 1.4766267236465516,
      "learning_rate": 1e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.850063681602478,
      "num_tokens": 87423991.0,
      "step": 2515
    },
    {
      "epoch": 0.46722376973073354,
      "grad_norm": 1.5214070942968434,
      "learning_rate": 1e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8539857268333435,
      "num_tokens": 87459914.0,
      "step": 2516
    },
    {
      "epoch": 0.46740947075208916,
      "grad_norm": 1.3438373797622563,
      "learning_rate": 1e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8625637292861938,
      "num_tokens": 87500190.0,
      "step": 2517
    },
    {
      "epoch": 0.46759517177344473,
      "grad_norm": 1.4710045519723278,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8711457252502441,
      "num_tokens": 87537472.0,
      "step": 2518
    },
    {
      "epoch": 0.46778087279480035,
      "grad_norm": 1.335888370734282,
      "learning_rate": 1e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8614774942398071,
      "num_tokens": 87576024.0,
      "step": 2519
    },
    {
      "epoch": 0.467966573816156,
      "grad_norm": 1.3998565995274277,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8693770170211792,
      "num_tokens": 87613570.0,
      "step": 2520
    },
    {
      "epoch": 0.4681522748375116,
      "grad_norm": 1.585512814386193,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8749442100524902,
      "num_tokens": 87641042.0,
      "step": 2521
    },
    {
      "epoch": 0.4683379758588672,
      "grad_norm": 1.3907919811880745,
      "learning_rate": 1e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8585556149482727,
      "num_tokens": 87680229.0,
      "step": 2522
    },
    {
      "epoch": 0.46852367688022284,
      "grad_norm": 1.604833725478006,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.868243396282196,
      "num_tokens": 87713672.0,
      "step": 2523
    },
    {
      "epoch": 0.46870937790157846,
      "grad_norm": 1.4024380059112143,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8719788789749146,
      "num_tokens": 87749602.0,
      "step": 2524
    },
    {
      "epoch": 0.4688950789229341,
      "grad_norm": 1.3239461398261516,
      "learning_rate": 1e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.848161518573761,
      "num_tokens": 87793454.0,
      "step": 2525
    },
    {
      "epoch": 0.4690807799442897,
      "grad_norm": 1.5443253223271687,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8661092519760132,
      "num_tokens": 87825718.0,
      "step": 2526
    },
    {
      "epoch": 0.4692664809656453,
      "grad_norm": 1.4675602174639966,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8610155582427979,
      "num_tokens": 87863013.0,
      "step": 2527
    },
    {
      "epoch": 0.46945218198700095,
      "grad_norm": 1.7152844285428068,
      "learning_rate": 1e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8460239171981812,
      "num_tokens": 87892184.0,
      "step": 2528
    },
    {
      "epoch": 0.46963788300835657,
      "grad_norm": 1.4402298032016805,
      "learning_rate": 1e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.850655198097229,
      "num_tokens": 87932926.0,
      "step": 2529
    },
    {
      "epoch": 0.46982358402971214,
      "grad_norm": 1.5080579272989596,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8550472855567932,
      "num_tokens": 87968920.0,
      "step": 2530
    },
    {
      "epoch": 0.47000928505106776,
      "grad_norm": 1.256694309977001,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8804553747177124,
      "num_tokens": 88011044.0,
      "step": 2531
    },
    {
      "epoch": 0.4701949860724234,
      "grad_norm": 1.3975581592337163,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8549602627754211,
      "num_tokens": 88052700.0,
      "step": 2532
    },
    {
      "epoch": 0.470380687093779,
      "grad_norm": 1.5402399859768119,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.85748291015625,
      "num_tokens": 88085525.0,
      "step": 2533
    },
    {
      "epoch": 0.4705663881151346,
      "grad_norm": 1.5269929021614261,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8772234320640564,
      "num_tokens": 88119424.0,
      "step": 2534
    },
    {
      "epoch": 0.47075208913649025,
      "grad_norm": 1.4629682764349072,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8707452416419983,
      "num_tokens": 88155840.0,
      "step": 2535
    },
    {
      "epoch": 0.47093779015784587,
      "grad_norm": 1.4905532144378755,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8686906099319458,
      "num_tokens": 88190261.0,
      "step": 2536
    },
    {
      "epoch": 0.4711234911792015,
      "grad_norm": 1.4675948048752858,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8710008859634399,
      "num_tokens": 88225531.0,
      "step": 2537
    },
    {
      "epoch": 0.4713091922005571,
      "grad_norm": 1.6707709372930688,
      "learning_rate": 1e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8519579172134399,
      "num_tokens": 88251612.0,
      "step": 2538
    },
    {
      "epoch": 0.47149489322191274,
      "grad_norm": 1.4797452462289573,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8708850145339966,
      "num_tokens": 88290026.0,
      "step": 2539
    },
    {
      "epoch": 0.47168059424326836,
      "grad_norm": 1.5061297053155749,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8705103397369385,
      "num_tokens": 88324850.0,
      "step": 2540
    },
    {
      "epoch": 0.471866295264624,
      "grad_norm": 1.6916627847910628,
      "learning_rate": 1e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8510860800743103,
      "num_tokens": 88355581.0,
      "step": 2541
    },
    {
      "epoch": 0.47205199628597955,
      "grad_norm": 1.502353607073337,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8609448671340942,
      "num_tokens": 88389228.0,
      "step": 2542
    },
    {
      "epoch": 0.47223769730733517,
      "grad_norm": 1.4446568362965189,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8656755685806274,
      "num_tokens": 88426688.0,
      "step": 2543
    },
    {
      "epoch": 0.4724233983286908,
      "grad_norm": 1.4547716718914327,
      "learning_rate": 1e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8518187999725342,
      "num_tokens": 88470379.0,
      "step": 2544
    },
    {
      "epoch": 0.4726090993500464,
      "grad_norm": 1.5334483193550639,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8674353957176208,
      "num_tokens": 88503887.0,
      "step": 2545
    },
    {
      "epoch": 0.47279480037140204,
      "grad_norm": 1.5745998733632902,
      "learning_rate": 1e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8509669303894043,
      "num_tokens": 88536052.0,
      "step": 2546
    },
    {
      "epoch": 0.47298050139275766,
      "grad_norm": 1.6072552052718636,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8591402173042297,
      "num_tokens": 88570668.0,
      "step": 2547
    },
    {
      "epoch": 0.4731662024141133,
      "grad_norm": 1.4529120586984738,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8600901365280151,
      "num_tokens": 88610598.0,
      "step": 2548
    },
    {
      "epoch": 0.4733519034354689,
      "grad_norm": 1.46175802648421,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8671947717666626,
      "num_tokens": 88643336.0,
      "step": 2549
    },
    {
      "epoch": 0.4735376044568245,
      "grad_norm": 1.5833694298633814,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8650041818618774,
      "num_tokens": 88676919.0,
      "step": 2550
    },
    {
      "epoch": 0.47372330547818015,
      "grad_norm": 1.4452902564890893,
      "learning_rate": 1e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8526183366775513,
      "num_tokens": 88715683.0,
      "step": 2551
    },
    {
      "epoch": 0.47390900649953577,
      "grad_norm": 1.36999990608907,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8543791770935059,
      "num_tokens": 88753472.0,
      "step": 2552
    },
    {
      "epoch": 0.4740947075208914,
      "grad_norm": 1.4180985032284044,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.873075008392334,
      "num_tokens": 88788698.0,
      "step": 2553
    },
    {
      "epoch": 0.47428040854224696,
      "grad_norm": 1.5523206208639573,
      "learning_rate": 1e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8504480719566345,
      "num_tokens": 88821960.0,
      "step": 2554
    },
    {
      "epoch": 0.4744661095636026,
      "grad_norm": 1.5167787269877404,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8698241114616394,
      "num_tokens": 88854300.0,
      "step": 2555
    },
    {
      "epoch": 0.4746518105849582,
      "grad_norm": 1.5152495539424504,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.853691041469574,
      "num_tokens": 88890289.0,
      "step": 2556
    },
    {
      "epoch": 0.4748375116063138,
      "grad_norm": 1.619344657264884,
      "learning_rate": 1e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8516392707824707,
      "num_tokens": 88922549.0,
      "step": 2557
    },
    {
      "epoch": 0.47502321262766944,
      "grad_norm": 1.6019849818319063,
      "learning_rate": 1e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8583074808120728,
      "num_tokens": 88953381.0,
      "step": 2558
    },
    {
      "epoch": 0.47520891364902507,
      "grad_norm": 1.3843824557873972,
      "learning_rate": 1e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8581965565681458,
      "num_tokens": 88992746.0,
      "step": 2559
    },
    {
      "epoch": 0.4753946146703807,
      "grad_norm": 1.475829394452394,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8703745007514954,
      "num_tokens": 89024127.0,
      "step": 2560
    },
    {
      "epoch": 0.4755803156917363,
      "grad_norm": 1.6440119095212942,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8598818182945251,
      "num_tokens": 89055744.0,
      "step": 2561
    },
    {
      "epoch": 0.47576601671309193,
      "grad_norm": 1.372320109290195,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8692358732223511,
      "num_tokens": 89094706.0,
      "step": 2562
    },
    {
      "epoch": 0.47595171773444755,
      "grad_norm": 1.6272436303007516,
      "learning_rate": 1e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8419284820556641,
      "num_tokens": 89129207.0,
      "step": 2563
    },
    {
      "epoch": 0.4761374187558032,
      "grad_norm": 1.4701217013828007,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8654265403747559,
      "num_tokens": 89162869.0,
      "step": 2564
    },
    {
      "epoch": 0.4763231197771588,
      "grad_norm": 1.5386834406821492,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8689425587654114,
      "num_tokens": 89191107.0,
      "step": 2565
    },
    {
      "epoch": 0.47650882079851437,
      "grad_norm": 1.5839692633039348,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8637043237686157,
      "num_tokens": 89226480.0,
      "step": 2566
    },
    {
      "epoch": 0.47669452181987,
      "grad_norm": 1.4079986350015017,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8876742124557495,
      "num_tokens": 89263319.0,
      "step": 2567
    },
    {
      "epoch": 0.4768802228412256,
      "grad_norm": 1.6497985391111416,
      "learning_rate": 1e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8646209239959717,
      "num_tokens": 89291235.0,
      "step": 2568
    },
    {
      "epoch": 0.47706592386258123,
      "grad_norm": 1.6628321172506568,
      "learning_rate": 1e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8573875427246094,
      "num_tokens": 89320305.0,
      "step": 2569
    },
    {
      "epoch": 0.47725162488393685,
      "grad_norm": 1.4033831428560124,
      "learning_rate": 1e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8593083620071411,
      "num_tokens": 89359101.0,
      "step": 2570
    },
    {
      "epoch": 0.4774373259052925,
      "grad_norm": 1.448678261007029,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8624513149261475,
      "num_tokens": 89394902.0,
      "step": 2571
    },
    {
      "epoch": 0.4776230269266481,
      "grad_norm": 1.4747026870750326,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8760461211204529,
      "num_tokens": 89428135.0,
      "step": 2572
    },
    {
      "epoch": 0.4778087279480037,
      "grad_norm": 1.762168435541292,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8735761642456055,
      "num_tokens": 89456788.0,
      "step": 2573
    },
    {
      "epoch": 0.47799442896935934,
      "grad_norm": 1.580009380019218,
      "learning_rate": 1e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8540520668029785,
      "num_tokens": 89490926.0,
      "step": 2574
    },
    {
      "epoch": 0.47818012999071496,
      "grad_norm": 1.3883091941210979,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8725965619087219,
      "num_tokens": 89527442.0,
      "step": 2575
    },
    {
      "epoch": 0.4783658310120706,
      "grad_norm": 1.3647921584443916,
      "learning_rate": 1e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8564075231552124,
      "num_tokens": 89569408.0,
      "step": 2576
    },
    {
      "epoch": 0.4785515320334262,
      "grad_norm": 1.411250090207625,
      "learning_rate": 1e-06,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8591213822364807,
      "num_tokens": 89608369.0,
      "step": 2577
    },
    {
      "epoch": 0.4787372330547818,
      "grad_norm": 1.5078671681758677,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8596417903900146,
      "num_tokens": 89644066.0,
      "step": 2578
    },
    {
      "epoch": 0.4789229340761374,
      "grad_norm": 1.4032732551400031,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8695064187049866,
      "num_tokens": 89680044.0,
      "step": 2579
    },
    {
      "epoch": 0.479108635097493,
      "grad_norm": 1.4607904284297824,
      "learning_rate": 1e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8665454387664795,
      "num_tokens": 89716590.0,
      "step": 2580
    },
    {
      "epoch": 0.47929433611884864,
      "grad_norm": 1.5623522273322714,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8614006042480469,
      "num_tokens": 89745555.0,
      "step": 2581
    },
    {
      "epoch": 0.47948003714020426,
      "grad_norm": 1.3712133464767662,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.865392804145813,
      "num_tokens": 89789117.0,
      "step": 2582
    },
    {
      "epoch": 0.4796657381615599,
      "grad_norm": 1.4817025297094772,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8627424240112305,
      "num_tokens": 89825291.0,
      "step": 2583
    },
    {
      "epoch": 0.4798514391829155,
      "grad_norm": 1.5048130744165797,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.874997615814209,
      "num_tokens": 89856852.0,
      "step": 2584
    },
    {
      "epoch": 0.48003714020427113,
      "grad_norm": 1.3508983902925056,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8611755967140198,
      "num_tokens": 89896699.0,
      "step": 2585
    },
    {
      "epoch": 0.48022284122562675,
      "grad_norm": 1.5507536505183477,
      "learning_rate": 1e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8555951118469238,
      "num_tokens": 89932155.0,
      "step": 2586
    },
    {
      "epoch": 0.4804085422469824,
      "grad_norm": 1.42542226904399,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8650190234184265,
      "num_tokens": 89967343.0,
      "step": 2587
    },
    {
      "epoch": 0.480594243268338,
      "grad_norm": 1.5760414714070152,
      "learning_rate": 1e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8390359878540039,
      "num_tokens": 90001877.0,
      "step": 2588
    },
    {
      "epoch": 0.4807799442896936,
      "grad_norm": 1.4075538291449867,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8713392019271851,
      "num_tokens": 90035988.0,
      "step": 2589
    },
    {
      "epoch": 0.4809656453110492,
      "grad_norm": 1.5127240441511691,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8672038912773132,
      "num_tokens": 90070492.0,
      "step": 2590
    },
    {
      "epoch": 0.4811513463324048,
      "grad_norm": 1.6082022626202361,
      "learning_rate": 1e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8541421890258789,
      "num_tokens": 90105127.0,
      "step": 2591
    },
    {
      "epoch": 0.4813370473537604,
      "grad_norm": 1.5205930762961561,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8608856201171875,
      "num_tokens": 90140236.0,
      "step": 2592
    },
    {
      "epoch": 0.48152274837511605,
      "grad_norm": 1.624877597868651,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8642549514770508,
      "num_tokens": 90167630.0,
      "step": 2593
    },
    {
      "epoch": 0.48170844939647167,
      "grad_norm": 1.4488169868124416,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8808587193489075,
      "num_tokens": 90201287.0,
      "step": 2594
    },
    {
      "epoch": 0.4818941504178273,
      "grad_norm": 1.4469269593679204,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8665990233421326,
      "num_tokens": 90235921.0,
      "step": 2595
    },
    {
      "epoch": 0.4820798514391829,
      "grad_norm": 1.4645414428437842,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.874538242816925,
      "num_tokens": 90270060.0,
      "step": 2596
    },
    {
      "epoch": 0.48226555246053854,
      "grad_norm": 1.523235154973095,
      "learning_rate": 1e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8706342577934265,
      "num_tokens": 90305744.0,
      "step": 2597
    },
    {
      "epoch": 0.48245125348189416,
      "grad_norm": 1.3609101655174785,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8656476736068726,
      "num_tokens": 90343731.0,
      "step": 2598
    },
    {
      "epoch": 0.4826369545032498,
      "grad_norm": 1.634504446009892,
      "learning_rate": 1e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8542942404747009,
      "num_tokens": 90373027.0,
      "step": 2599
    },
    {
      "epoch": 0.4828226555246054,
      "grad_norm": 1.5093393442102156,
      "learning_rate": 1e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8432455062866211,
      "num_tokens": 90412377.0,
      "step": 2600
    },
    {
      "epoch": 0.483008356545961,
      "grad_norm": 1.4784297683333643,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8568601012229919,
      "num_tokens": 90449792.0,
      "step": 2601
    },
    {
      "epoch": 0.4831940575673166,
      "grad_norm": 1.6268306692849215,
      "learning_rate": 1e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8483590483665466,
      "num_tokens": 90480958.0,
      "step": 2602
    },
    {
      "epoch": 0.4833797585886722,
      "grad_norm": 1.677488583751221,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8671483993530273,
      "num_tokens": 90510643.0,
      "step": 2603
    },
    {
      "epoch": 0.48356545961002784,
      "grad_norm": 1.5088217561675918,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8631008863449097,
      "num_tokens": 90546511.0,
      "step": 2604
    },
    {
      "epoch": 0.48375116063138346,
      "grad_norm": 1.5710109053745627,
      "learning_rate": 1e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8603984117507935,
      "num_tokens": 90576714.0,
      "step": 2605
    },
    {
      "epoch": 0.4839368616527391,
      "grad_norm": 1.6094878433100457,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8593742251396179,
      "num_tokens": 90607827.0,
      "step": 2606
    },
    {
      "epoch": 0.4841225626740947,
      "grad_norm": 1.468312755053183,
      "learning_rate": 1e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8583530187606812,
      "num_tokens": 90645439.0,
      "step": 2607
    },
    {
      "epoch": 0.4843082636954503,
      "grad_norm": 1.5331134543011236,
      "learning_rate": 1e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8615167140960693,
      "num_tokens": 90677845.0,
      "step": 2608
    },
    {
      "epoch": 0.48449396471680595,
      "grad_norm": 1.5759491234904808,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8696230053901672,
      "num_tokens": 90711811.0,
      "step": 2609
    },
    {
      "epoch": 0.48467966573816157,
      "grad_norm": 1.6385294304014888,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.852484405040741,
      "num_tokens": 90741880.0,
      "step": 2610
    },
    {
      "epoch": 0.4848653667595172,
      "grad_norm": 1.4326691673090086,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8745974898338318,
      "num_tokens": 90775695.0,
      "step": 2611
    },
    {
      "epoch": 0.4850510677808728,
      "grad_norm": 1.495013705496463,
      "learning_rate": 1e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8565428256988525,
      "num_tokens": 90809626.0,
      "step": 2612
    },
    {
      "epoch": 0.48523676880222844,
      "grad_norm": 1.5117459422672312,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8689405918121338,
      "num_tokens": 90839659.0,
      "step": 2613
    },
    {
      "epoch": 0.48542246982358406,
      "grad_norm": 1.7147911912140297,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8763465285301208,
      "num_tokens": 90865557.0,
      "step": 2614
    },
    {
      "epoch": 0.4856081708449396,
      "grad_norm": 1.646104002989523,
      "learning_rate": 1e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8531743288040161,
      "num_tokens": 90897301.0,
      "step": 2615
    },
    {
      "epoch": 0.48579387186629525,
      "grad_norm": 1.486980227773111,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8737233877182007,
      "num_tokens": 90928863.0,
      "step": 2616
    },
    {
      "epoch": 0.48597957288765087,
      "grad_norm": 1.3758777945501548,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8706231117248535,
      "num_tokens": 90967188.0,
      "step": 2617
    },
    {
      "epoch": 0.4861652739090065,
      "grad_norm": 1.5769496255442335,
      "learning_rate": 1e-06,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8459793329238892,
      "num_tokens": 91005330.0,
      "step": 2618
    },
    {
      "epoch": 0.4863509749303621,
      "grad_norm": 1.6009811344227758,
      "learning_rate": 1e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8577840924263,
      "num_tokens": 91034372.0,
      "step": 2619
    },
    {
      "epoch": 0.48653667595171773,
      "grad_norm": 1.4350202587561705,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8732823729515076,
      "num_tokens": 91070352.0,
      "step": 2620
    },
    {
      "epoch": 0.48672237697307336,
      "grad_norm": 1.45384945263812,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8610942363739014,
      "num_tokens": 91105320.0,
      "step": 2621
    },
    {
      "epoch": 0.486908077994429,
      "grad_norm": 1.378854008077259,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8622243404388428,
      "num_tokens": 91142600.0,
      "step": 2622
    },
    {
      "epoch": 0.4870937790157846,
      "grad_norm": 1.5126279115391474,
      "learning_rate": 1e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.860232949256897,
      "num_tokens": 91179516.0,
      "step": 2623
    },
    {
      "epoch": 0.4872794800371402,
      "grad_norm": 1.4094927577228447,
      "learning_rate": 1e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8522594571113586,
      "num_tokens": 91222048.0,
      "step": 2624
    },
    {
      "epoch": 0.48746518105849584,
      "grad_norm": 1.6392662243273424,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8699880838394165,
      "num_tokens": 91246652.0,
      "step": 2625
    },
    {
      "epoch": 0.48765088207985147,
      "grad_norm": 1.5605812078694308,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.867252767086029,
      "num_tokens": 91277902.0,
      "step": 2626
    },
    {
      "epoch": 0.48783658310120703,
      "grad_norm": 1.4486689969004027,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.863455057144165,
      "num_tokens": 91317535.0,
      "step": 2627
    },
    {
      "epoch": 0.48802228412256266,
      "grad_norm": 1.4527898485942452,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.863251805305481,
      "num_tokens": 91357230.0,
      "step": 2628
    },
    {
      "epoch": 0.4882079851439183,
      "grad_norm": 1.4337022180314405,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8645446300506592,
      "num_tokens": 91392609.0,
      "step": 2629
    },
    {
      "epoch": 0.4883936861652739,
      "grad_norm": 1.6468384146096584,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8569043874740601,
      "num_tokens": 91427561.0,
      "step": 2630
    },
    {
      "epoch": 0.4885793871866295,
      "grad_norm": 1.4976641454727861,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8655683994293213,
      "num_tokens": 91460267.0,
      "step": 2631
    },
    {
      "epoch": 0.48876508820798514,
      "grad_norm": 1.5220998623143407,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8686851859092712,
      "num_tokens": 91492643.0,
      "step": 2632
    },
    {
      "epoch": 0.48895078922934077,
      "grad_norm": 1.5925473135218773,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8704527616500854,
      "num_tokens": 91526428.0,
      "step": 2633
    },
    {
      "epoch": 0.4891364902506964,
      "grad_norm": 1.33537420756659,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8804153203964233,
      "num_tokens": 91563375.0,
      "step": 2634
    },
    {
      "epoch": 0.489322191272052,
      "grad_norm": 1.4554217190115502,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8589562177658081,
      "num_tokens": 91597304.0,
      "step": 2635
    },
    {
      "epoch": 0.48950789229340763,
      "grad_norm": 1.517908832244657,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8633760213851929,
      "num_tokens": 91634009.0,
      "step": 2636
    },
    {
      "epoch": 0.48969359331476325,
      "grad_norm": 1.54504913477987,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8612509965896606,
      "num_tokens": 91667923.0,
      "step": 2637
    },
    {
      "epoch": 0.4898792943361189,
      "grad_norm": 1.3490111586054672,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8693673014640808,
      "num_tokens": 91708633.0,
      "step": 2638
    },
    {
      "epoch": 0.49006499535747444,
      "grad_norm": 1.3792472940813898,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8680287599563599,
      "num_tokens": 91747859.0,
      "step": 2639
    },
    {
      "epoch": 0.49025069637883006,
      "grad_norm": 1.5464534679298119,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8620193004608154,
      "num_tokens": 91783497.0,
      "step": 2640
    },
    {
      "epoch": 0.4904363974001857,
      "grad_norm": 1.5808575737869501,
      "learning_rate": 1e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8558934330940247,
      "num_tokens": 91817318.0,
      "step": 2641
    },
    {
      "epoch": 0.4906220984215413,
      "grad_norm": 1.6283965529943563,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8629213571548462,
      "num_tokens": 91851160.0,
      "step": 2642
    },
    {
      "epoch": 0.49080779944289693,
      "grad_norm": 1.4660427525341426,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8597621917724609,
      "num_tokens": 91886179.0,
      "step": 2643
    },
    {
      "epoch": 0.49099350046425255,
      "grad_norm": 1.5116270512022445,
      "learning_rate": 1e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8619177937507629,
      "num_tokens": 91922947.0,
      "step": 2644
    },
    {
      "epoch": 0.4911792014856082,
      "grad_norm": 1.4403600740674452,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8666180968284607,
      "num_tokens": 91961678.0,
      "step": 2645
    },
    {
      "epoch": 0.4913649025069638,
      "grad_norm": 1.5521632975465862,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8756136894226074,
      "num_tokens": 91992859.0,
      "step": 2646
    },
    {
      "epoch": 0.4915506035283194,
      "grad_norm": 1.4447506614141874,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8689825534820557,
      "num_tokens": 92026851.0,
      "step": 2647
    },
    {
      "epoch": 0.49173630454967504,
      "grad_norm": 1.5633635917155282,
      "learning_rate": 1e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8645678758621216,
      "num_tokens": 92058103.0,
      "step": 2648
    },
    {
      "epoch": 0.49192200557103066,
      "grad_norm": 1.4893099286895444,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8691108822822571,
      "num_tokens": 92096735.0,
      "step": 2649
    },
    {
      "epoch": 0.4921077065923863,
      "grad_norm": 1.3466720457475003,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8763227462768555,
      "num_tokens": 92133338.0,
      "step": 2650
    },
    {
      "epoch": 0.49229340761374185,
      "grad_norm": 1.4367600353966057,
      "learning_rate": 1e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8597482442855835,
      "num_tokens": 92171757.0,
      "step": 2651
    },
    {
      "epoch": 0.4924791086350975,
      "grad_norm": 1.5173614917583198,
      "learning_rate": 1e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8560079336166382,
      "num_tokens": 92205639.0,
      "step": 2652
    },
    {
      "epoch": 0.4926648096564531,
      "grad_norm": 1.47697768045868,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8631178140640259,
      "num_tokens": 92242449.0,
      "step": 2653
    },
    {
      "epoch": 0.4928505106778087,
      "grad_norm": 1.461901518102434,
      "learning_rate": 1e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8546559810638428,
      "num_tokens": 92277327.0,
      "step": 2654
    },
    {
      "epoch": 0.49303621169916434,
      "grad_norm": 1.3599461688937438,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8635879158973694,
      "num_tokens": 92315934.0,
      "step": 2655
    },
    {
      "epoch": 0.49322191272051996,
      "grad_norm": 1.430959043716517,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8689563274383545,
      "num_tokens": 92351937.0,
      "step": 2656
    },
    {
      "epoch": 0.4934076137418756,
      "grad_norm": 1.6861715254700618,
      "learning_rate": 1e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.862156867980957,
      "num_tokens": 92378682.0,
      "step": 2657
    },
    {
      "epoch": 0.4935933147632312,
      "grad_norm": 1.4028789017006738,
      "learning_rate": 1e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8549296855926514,
      "num_tokens": 92419940.0,
      "step": 2658
    },
    {
      "epoch": 0.4937790157845868,
      "grad_norm": 1.4772460381849768,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8601391315460205,
      "num_tokens": 92458129.0,
      "step": 2659
    },
    {
      "epoch": 0.49396471680594245,
      "grad_norm": 1.347023496212969,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8801515102386475,
      "num_tokens": 92495709.0,
      "step": 2660
    },
    {
      "epoch": 0.49415041782729807,
      "grad_norm": 1.4137654832190485,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8622760772705078,
      "num_tokens": 92532359.0,
      "step": 2661
    },
    {
      "epoch": 0.4943361188486537,
      "grad_norm": 1.6160457296900723,
      "learning_rate": 1e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8485724925994873,
      "num_tokens": 92563058.0,
      "step": 2662
    },
    {
      "epoch": 0.49452181987000926,
      "grad_norm": 1.5628467208760652,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8615024089813232,
      "num_tokens": 92595546.0,
      "step": 2663
    },
    {
      "epoch": 0.4947075208913649,
      "grad_norm": 1.4803830714100825,
      "learning_rate": 1e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8643202781677246,
      "num_tokens": 92631350.0,
      "step": 2664
    },
    {
      "epoch": 0.4948932219127205,
      "grad_norm": 1.394733089113639,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8741898536682129,
      "num_tokens": 92672514.0,
      "step": 2665
    },
    {
      "epoch": 0.4950789229340761,
      "grad_norm": 1.6238008004993447,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8617097735404968,
      "num_tokens": 92700320.0,
      "step": 2666
    },
    {
      "epoch": 0.49526462395543175,
      "grad_norm": 1.4826115601169103,
      "learning_rate": 1e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8519027233123779,
      "num_tokens": 92740330.0,
      "step": 2667
    },
    {
      "epoch": 0.49545032497678737,
      "grad_norm": 1.3767188535862889,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8620625734329224,
      "num_tokens": 92779666.0,
      "step": 2668
    },
    {
      "epoch": 0.495636025998143,
      "grad_norm": 1.4800451834322907,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8748610019683838,
      "num_tokens": 92812268.0,
      "step": 2669
    },
    {
      "epoch": 0.4958217270194986,
      "grad_norm": 1.4869751894456047,
      "learning_rate": 1e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8605620861053467,
      "num_tokens": 92853852.0,
      "step": 2670
    },
    {
      "epoch": 0.49600742804085424,
      "grad_norm": 1.484355928465609,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8584905862808228,
      "num_tokens": 92888166.0,
      "step": 2671
    },
    {
      "epoch": 0.49619312906220986,
      "grad_norm": 1.5233000026051737,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8805509805679321,
      "num_tokens": 92922823.0,
      "step": 2672
    },
    {
      "epoch": 0.4963788300835655,
      "grad_norm": 1.5253624791811087,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8738940954208374,
      "num_tokens": 92955117.0,
      "step": 2673
    },
    {
      "epoch": 0.4965645311049211,
      "grad_norm": 1.509774513096776,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8629574775695801,
      "num_tokens": 92989485.0,
      "step": 2674
    },
    {
      "epoch": 0.49675023212627667,
      "grad_norm": 1.6494615194680258,
      "learning_rate": 1e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8530597686767578,
      "num_tokens": 93019122.0,
      "step": 2675
    },
    {
      "epoch": 0.4969359331476323,
      "grad_norm": 1.5932775398090087,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.866684079170227,
      "num_tokens": 93047045.0,
      "step": 2676
    },
    {
      "epoch": 0.4971216341689879,
      "grad_norm": 1.6704770873113841,
      "learning_rate": 1e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8577831387519836,
      "num_tokens": 93078104.0,
      "step": 2677
    },
    {
      "epoch": 0.49730733519034354,
      "grad_norm": 1.4378794293429358,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8636274337768555,
      "num_tokens": 93117428.0,
      "step": 2678
    },
    {
      "epoch": 0.49749303621169916,
      "grad_norm": 1.5975288756223414,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.866500973701477,
      "num_tokens": 93151671.0,
      "step": 2679
    },
    {
      "epoch": 0.4976787372330548,
      "grad_norm": 1.8888740014613636,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8622790575027466,
      "num_tokens": 93180009.0,
      "step": 2680
    },
    {
      "epoch": 0.4978644382544104,
      "grad_norm": 1.4676755804407184,
      "learning_rate": 1e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8519892692565918,
      "num_tokens": 93216744.0,
      "step": 2681
    },
    {
      "epoch": 0.498050139275766,
      "grad_norm": 1.771752941489223,
      "learning_rate": 1e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8599228262901306,
      "num_tokens": 93244444.0,
      "step": 2682
    },
    {
      "epoch": 0.49823584029712165,
      "grad_norm": 1.5431723314734795,
      "learning_rate": 1e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8593106865882874,
      "num_tokens": 93277064.0,
      "step": 2683
    },
    {
      "epoch": 0.49842154131847727,
      "grad_norm": 1.6511129512830085,
      "learning_rate": 1e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8555334806442261,
      "num_tokens": 93304584.0,
      "step": 2684
    },
    {
      "epoch": 0.4986072423398329,
      "grad_norm": 1.495185887299927,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8763507604598999,
      "num_tokens": 93340021.0,
      "step": 2685
    },
    {
      "epoch": 0.4987929433611885,
      "grad_norm": 1.3920237307293362,
      "learning_rate": 1e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8548563122749329,
      "num_tokens": 93381618.0,
      "step": 2686
    },
    {
      "epoch": 0.4989786443825441,
      "grad_norm": 1.3991748896201244,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8708717226982117,
      "num_tokens": 93421362.0,
      "step": 2687
    },
    {
      "epoch": 0.4991643454038997,
      "grad_norm": 1.3886602733837277,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8678598403930664,
      "num_tokens": 93457099.0,
      "step": 2688
    },
    {
      "epoch": 0.4993500464252553,
      "grad_norm": 1.4954036572120786,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8838626742362976,
      "num_tokens": 93489014.0,
      "step": 2689
    },
    {
      "epoch": 0.49953574744661094,
      "grad_norm": 1.6464111041374596,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8561607003211975,
      "num_tokens": 93517531.0,
      "step": 2690
    },
    {
      "epoch": 0.49972144846796657,
      "grad_norm": 1.4296937967767953,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.867764949798584,
      "num_tokens": 93561097.0,
      "step": 2691
    },
    {
      "epoch": 0.4999071494893222,
      "grad_norm": 1.362377876297636,
      "learning_rate": 1e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8375580906867981,
      "num_tokens": 93608898.0,
      "step": 2692
    },
    {
      "epoch": 0.5000928505106778,
      "grad_norm": 1.3941199676881602,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8833483457565308,
      "num_tokens": 93640891.0,
      "step": 2693
    },
    {
      "epoch": 0.5002785515320334,
      "grad_norm": 1.4280637689709486,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8741714358329773,
      "num_tokens": 93678227.0,
      "step": 2694
    },
    {
      "epoch": 0.500464252553389,
      "grad_norm": 1.5087833474820638,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.867081344127655,
      "num_tokens": 93713177.0,
      "step": 2695
    },
    {
      "epoch": 0.5006499535747446,
      "grad_norm": 1.4867399120608416,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8695032596588135,
      "num_tokens": 93750080.0,
      "step": 2696
    },
    {
      "epoch": 0.5008356545961002,
      "grad_norm": 1.4437594127702376,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8651351928710938,
      "num_tokens": 93787564.0,
      "step": 2697
    },
    {
      "epoch": 0.5010213556174559,
      "grad_norm": 1.3892874031415814,
      "learning_rate": 1e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8695120811462402,
      "num_tokens": 93824230.0,
      "step": 2698
    },
    {
      "epoch": 0.5012070566388115,
      "grad_norm": 1.3989395568381997,
      "learning_rate": 1e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8562138080596924,
      "num_tokens": 93860721.0,
      "step": 2699
    },
    {
      "epoch": 0.5013927576601671,
      "grad_norm": 1.4899750665137177,
      "learning_rate": 1e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8631459474563599,
      "num_tokens": 93895266.0,
      "step": 2700
    },
    {
      "epoch": 0.5015784586815227,
      "grad_norm": 1.5494760252055597,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8645416498184204,
      "num_tokens": 93933922.0,
      "step": 2701
    },
    {
      "epoch": 0.5017641597028784,
      "grad_norm": 1.3505208433632194,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8718623518943787,
      "num_tokens": 93975029.0,
      "step": 2702
    },
    {
      "epoch": 0.501949860724234,
      "grad_norm": 1.6038554050276774,
      "learning_rate": 1e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.843631386756897,
      "num_tokens": 94012333.0,
      "step": 2703
    },
    {
      "epoch": 0.5021355617455896,
      "grad_norm": 1.483004732590096,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.866756796836853,
      "num_tokens": 94048081.0,
      "step": 2704
    },
    {
      "epoch": 0.5023212627669452,
      "grad_norm": 1.4082895410781258,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8769446015357971,
      "num_tokens": 94084255.0,
      "step": 2705
    },
    {
      "epoch": 0.5025069637883008,
      "grad_norm": 1.423795715854413,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8663549423217773,
      "num_tokens": 94117201.0,
      "step": 2706
    },
    {
      "epoch": 0.5026926648096565,
      "grad_norm": 1.4725849635208186,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8791630268096924,
      "num_tokens": 94151568.0,
      "step": 2707
    },
    {
      "epoch": 0.5028783658310121,
      "grad_norm": 1.5169017344152427,
      "learning_rate": 1e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8529129028320312,
      "num_tokens": 94186461.0,
      "step": 2708
    },
    {
      "epoch": 0.5030640668523677,
      "grad_norm": 1.4861658145400878,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8744450807571411,
      "num_tokens": 94222192.0,
      "step": 2709
    },
    {
      "epoch": 0.5032497678737233,
      "grad_norm": 1.512861559949825,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8574451208114624,
      "num_tokens": 94259058.0,
      "step": 2710
    },
    {
      "epoch": 0.503435468895079,
      "grad_norm": 1.5615205529352316,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8737943172454834,
      "num_tokens": 94288091.0,
      "step": 2711
    },
    {
      "epoch": 0.5036211699164346,
      "grad_norm": 1.4935276494774026,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.872843861579895,
      "num_tokens": 94319484.0,
      "step": 2712
    },
    {
      "epoch": 0.5038068709377902,
      "grad_norm": 1.571632236340144,
      "learning_rate": 1e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8374654054641724,
      "num_tokens": 94354350.0,
      "step": 2713
    },
    {
      "epoch": 0.5039925719591458,
      "grad_norm": 1.4239982462043876,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8597431182861328,
      "num_tokens": 94394563.0,
      "step": 2714
    },
    {
      "epoch": 0.5041782729805014,
      "grad_norm": 1.5080896670242077,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.863508939743042,
      "num_tokens": 94427828.0,
      "step": 2715
    },
    {
      "epoch": 0.5043639740018571,
      "grad_norm": 1.4635540035056718,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8672955632209778,
      "num_tokens": 94464219.0,
      "step": 2716
    },
    {
      "epoch": 0.5045496750232126,
      "grad_norm": 1.4881953211186414,
      "learning_rate": 1e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8504488468170166,
      "num_tokens": 94500293.0,
      "step": 2717
    },
    {
      "epoch": 0.5047353760445682,
      "grad_norm": 1.5255617172947502,
      "learning_rate": 1e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8546329140663147,
      "num_tokens": 94535348.0,
      "step": 2718
    },
    {
      "epoch": 0.5049210770659238,
      "grad_norm": 1.467455139265692,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8720859885215759,
      "num_tokens": 94567559.0,
      "step": 2719
    },
    {
      "epoch": 0.5051067780872794,
      "grad_norm": 1.7274071044236294,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8702701330184937,
      "num_tokens": 94593758.0,
      "step": 2720
    },
    {
      "epoch": 0.5052924791086351,
      "grad_norm": 1.4883164095044419,
      "learning_rate": 1e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8574252724647522,
      "num_tokens": 94629026.0,
      "step": 2721
    },
    {
      "epoch": 0.5054781801299907,
      "grad_norm": 1.4345801874617108,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8634036779403687,
      "num_tokens": 94664225.0,
      "step": 2722
    },
    {
      "epoch": 0.5056638811513463,
      "grad_norm": 1.3254154118813946,
      "learning_rate": 1e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.854405403137207,
      "num_tokens": 94704979.0,
      "step": 2723
    },
    {
      "epoch": 0.5058495821727019,
      "grad_norm": 1.4817077073779281,
      "learning_rate": 1e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.848025918006897,
      "num_tokens": 94742176.0,
      "step": 2724
    },
    {
      "epoch": 0.5060352831940576,
      "grad_norm": 1.396700822172162,
      "learning_rate": 1e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8504505157470703,
      "num_tokens": 94782195.0,
      "step": 2725
    },
    {
      "epoch": 0.5062209842154132,
      "grad_norm": 1.592010091323059,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8579667210578918,
      "num_tokens": 94813161.0,
      "step": 2726
    },
    {
      "epoch": 0.5064066852367688,
      "grad_norm": 1.7184157275293548,
      "learning_rate": 1e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8355862498283386,
      "num_tokens": 94844961.0,
      "step": 2727
    },
    {
      "epoch": 0.5065923862581244,
      "grad_norm": 1.6297813835262813,
      "learning_rate": 1e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8497568964958191,
      "num_tokens": 94875540.0,
      "step": 2728
    },
    {
      "epoch": 0.50677808727948,
      "grad_norm": 1.3998547321773733,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8706650733947754,
      "num_tokens": 94915469.0,
      "step": 2729
    },
    {
      "epoch": 0.5069637883008357,
      "grad_norm": 1.4025812659375452,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8763545155525208,
      "num_tokens": 94950364.0,
      "step": 2730
    },
    {
      "epoch": 0.5071494893221913,
      "grad_norm": 1.6135703726572408,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8597022891044617,
      "num_tokens": 94980476.0,
      "step": 2731
    },
    {
      "epoch": 0.5073351903435469,
      "grad_norm": 1.3957870669909573,
      "learning_rate": 1e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8604860305786133,
      "num_tokens": 95018046.0,
      "step": 2732
    },
    {
      "epoch": 0.5075208913649025,
      "grad_norm": 1.6201495384598465,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8647134304046631,
      "num_tokens": 95051804.0,
      "step": 2733
    },
    {
      "epoch": 0.5077065923862581,
      "grad_norm": 1.474613081559692,
      "learning_rate": 1e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8516402840614319,
      "num_tokens": 95092296.0,
      "step": 2734
    },
    {
      "epoch": 0.5078922934076138,
      "grad_norm": 1.504831933144384,
      "learning_rate": 1e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8688510656356812,
      "num_tokens": 95127627.0,
      "step": 2735
    },
    {
      "epoch": 0.5080779944289694,
      "grad_norm": 1.5006287137507184,
      "learning_rate": 1e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8607581853866577,
      "num_tokens": 95161004.0,
      "step": 2736
    },
    {
      "epoch": 0.508263695450325,
      "grad_norm": 1.386202756369923,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8710346817970276,
      "num_tokens": 95199972.0,
      "step": 2737
    },
    {
      "epoch": 0.5084493964716806,
      "grad_norm": 1.4358455659308185,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8716256618499756,
      "num_tokens": 95236836.0,
      "step": 2738
    },
    {
      "epoch": 0.5086350974930363,
      "grad_norm": 1.5374860352668924,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8666402101516724,
      "num_tokens": 95269732.0,
      "step": 2739
    },
    {
      "epoch": 0.5088207985143919,
      "grad_norm": 1.400229686700418,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8667045831680298,
      "num_tokens": 95306374.0,
      "step": 2740
    },
    {
      "epoch": 0.5090064995357474,
      "grad_norm": 1.4232279786400606,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8653296232223511,
      "num_tokens": 95343363.0,
      "step": 2741
    },
    {
      "epoch": 0.509192200557103,
      "grad_norm": 1.4191463264615185,
      "learning_rate": 1e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8629236221313477,
      "num_tokens": 95386158.0,
      "step": 2742
    },
    {
      "epoch": 0.5093779015784586,
      "grad_norm": 1.5924250568063607,
      "learning_rate": 1e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8535540699958801,
      "num_tokens": 95425004.0,
      "step": 2743
    },
    {
      "epoch": 0.5095636025998143,
      "grad_norm": 1.421481416907017,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8642024397850037,
      "num_tokens": 95466390.0,
      "step": 2744
    },
    {
      "epoch": 0.5097493036211699,
      "grad_norm": 1.4881883661358548,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8601484298706055,
      "num_tokens": 95504723.0,
      "step": 2745
    },
    {
      "epoch": 0.5099350046425255,
      "grad_norm": 1.462739797649939,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8650198578834534,
      "num_tokens": 95540149.0,
      "step": 2746
    },
    {
      "epoch": 0.5101207056638811,
      "grad_norm": 1.7036090123462024,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8566799759864807,
      "num_tokens": 95569843.0,
      "step": 2747
    },
    {
      "epoch": 0.5103064066852367,
      "grad_norm": 1.4718692021271298,
      "learning_rate": 1e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8589428663253784,
      "num_tokens": 95606223.0,
      "step": 2748
    },
    {
      "epoch": 0.5104921077065924,
      "grad_norm": 1.4343793559795885,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8777670860290527,
      "num_tokens": 95642015.0,
      "step": 2749
    },
    {
      "epoch": 0.510677808727948,
      "grad_norm": 1.5182935729098195,
      "learning_rate": 1e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8696175813674927,
      "num_tokens": 95673962.0,
      "step": 2750
    },
    {
      "epoch": 0.5108635097493036,
      "grad_norm": 1.3815640579791841,
      "learning_rate": 1e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8647259473800659,
      "num_tokens": 95715040.0,
      "step": 2751
    },
    {
      "epoch": 0.5110492107706592,
      "grad_norm": 1.6033132832944517,
      "learning_rate": 1e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8462210893630981,
      "num_tokens": 95746657.0,
      "step": 2752
    },
    {
      "epoch": 0.5112349117920149,
      "grad_norm": 1.50008184596168,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8690181970596313,
      "num_tokens": 95778738.0,
      "step": 2753
    },
    {
      "epoch": 0.5114206128133705,
      "grad_norm": 1.2884716328866148,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8876348733901978,
      "num_tokens": 95817377.0,
      "step": 2754
    },
    {
      "epoch": 0.5116063138347261,
      "grad_norm": 1.622100536316047,
      "learning_rate": 1e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.845292329788208,
      "num_tokens": 95853988.0,
      "step": 2755
    },
    {
      "epoch": 0.5117920148560817,
      "grad_norm": 1.5155012119146936,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8636219501495361,
      "num_tokens": 95888817.0,
      "step": 2756
    },
    {
      "epoch": 0.5119777158774373,
      "grad_norm": 1.4926778840413903,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8856939077377319,
      "num_tokens": 95919689.0,
      "step": 2757
    },
    {
      "epoch": 0.512163416898793,
      "grad_norm": 1.5061671273047257,
      "learning_rate": 1e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8665695190429688,
      "num_tokens": 95953754.0,
      "step": 2758
    },
    {
      "epoch": 0.5123491179201486,
      "grad_norm": 1.6883648377461375,
      "learning_rate": 1e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8495973348617554,
      "num_tokens": 95985626.0,
      "step": 2759
    },
    {
      "epoch": 0.5125348189415042,
      "grad_norm": 1.3358071101794915,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8734080791473389,
      "num_tokens": 96024753.0,
      "step": 2760
    },
    {
      "epoch": 0.5127205199628598,
      "grad_norm": 1.3468468716376272,
      "learning_rate": 1e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8563942909240723,
      "num_tokens": 96067241.0,
      "step": 2761
    },
    {
      "epoch": 0.5129062209842155,
      "grad_norm": 1.4609333359483359,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8710190057754517,
      "num_tokens": 96100562.0,
      "step": 2762
    },
    {
      "epoch": 0.5130919220055711,
      "grad_norm": 1.5379009492279132,
      "learning_rate": 1e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8629493713378906,
      "num_tokens": 96133663.0,
      "step": 2763
    },
    {
      "epoch": 0.5132776230269267,
      "grad_norm": 1.5708500702351684,
      "learning_rate": 1e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8513648509979248,
      "num_tokens": 96169266.0,
      "step": 2764
    },
    {
      "epoch": 0.5134633240482822,
      "grad_norm": 1.484267596984409,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8825684189796448,
      "num_tokens": 96200316.0,
      "step": 2765
    },
    {
      "epoch": 0.5136490250696378,
      "grad_norm": 1.468174111939501,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8673542141914368,
      "num_tokens": 96235676.0,
      "step": 2766
    },
    {
      "epoch": 0.5138347260909935,
      "grad_norm": 1.5075744841236447,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.864750862121582,
      "num_tokens": 96268885.0,
      "step": 2767
    },
    {
      "epoch": 0.5140204271123491,
      "grad_norm": 1.4599873561868468,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8714138269424438,
      "num_tokens": 96308455.0,
      "step": 2768
    },
    {
      "epoch": 0.5142061281337047,
      "grad_norm": 1.4628421735859203,
      "learning_rate": 1e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8644660711288452,
      "num_tokens": 96347286.0,
      "step": 2769
    },
    {
      "epoch": 0.5143918291550603,
      "grad_norm": 1.4622097490457027,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8707377314567566,
      "num_tokens": 96387949.0,
      "step": 2770
    },
    {
      "epoch": 0.5145775301764159,
      "grad_norm": 1.4973835524918202,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8746535778045654,
      "num_tokens": 96419000.0,
      "step": 2771
    },
    {
      "epoch": 0.5147632311977716,
      "grad_norm": 1.542148204922505,
      "learning_rate": 1e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8465165495872498,
      "num_tokens": 96453355.0,
      "step": 2772
    },
    {
      "epoch": 0.5149489322191272,
      "grad_norm": 1.484103094975706,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.876849353313446,
      "num_tokens": 96488467.0,
      "step": 2773
    },
    {
      "epoch": 0.5151346332404828,
      "grad_norm": 1.4461688443031109,
      "learning_rate": 1e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8611165881156921,
      "num_tokens": 96523264.0,
      "step": 2774
    },
    {
      "epoch": 0.5153203342618384,
      "grad_norm": 1.5120383904656145,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8612399697303772,
      "num_tokens": 96560166.0,
      "step": 2775
    },
    {
      "epoch": 0.515506035283194,
      "grad_norm": 1.6561424412080172,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8550799489021301,
      "num_tokens": 96587708.0,
      "step": 2776
    },
    {
      "epoch": 0.5156917363045497,
      "grad_norm": 1.407182677672185,
      "learning_rate": 1e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8568763732910156,
      "num_tokens": 96627796.0,
      "step": 2777
    },
    {
      "epoch": 0.5158774373259053,
      "grad_norm": 1.470025491903255,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8669443130493164,
      "num_tokens": 96663506.0,
      "step": 2778
    },
    {
      "epoch": 0.5160631383472609,
      "grad_norm": 1.6294279118296158,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8611515760421753,
      "num_tokens": 96701745.0,
      "step": 2779
    },
    {
      "epoch": 0.5162488393686165,
      "grad_norm": 1.590521327788118,
      "learning_rate": 1e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8729023933410645,
      "num_tokens": 96735954.0,
      "step": 2780
    },
    {
      "epoch": 0.5164345403899722,
      "grad_norm": 1.4033408344540546,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8671910762786865,
      "num_tokens": 96771474.0,
      "step": 2781
    },
    {
      "epoch": 0.5166202414113278,
      "grad_norm": 1.43430217790784,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.863899290561676,
      "num_tokens": 96805353.0,
      "step": 2782
    },
    {
      "epoch": 0.5168059424326834,
      "grad_norm": 1.4945182141500837,
      "learning_rate": 1e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8511945009231567,
      "num_tokens": 96841616.0,
      "step": 2783
    },
    {
      "epoch": 0.516991643454039,
      "grad_norm": 1.381955047200711,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8685745000839233,
      "num_tokens": 96881457.0,
      "step": 2784
    },
    {
      "epoch": 0.5171773444753947,
      "grad_norm": 1.6032267349841811,
      "learning_rate": 1e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8621417284011841,
      "num_tokens": 96908276.0,
      "step": 2785
    },
    {
      "epoch": 0.5173630454967503,
      "grad_norm": 1.6084753867163422,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8727463483810425,
      "num_tokens": 96938779.0,
      "step": 2786
    },
    {
      "epoch": 0.5175487465181059,
      "grad_norm": 1.4590501799594109,
      "learning_rate": 1e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8621960878372192,
      "num_tokens": 96977150.0,
      "step": 2787
    },
    {
      "epoch": 0.5177344475394615,
      "grad_norm": 1.4362776750106132,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8688535094261169,
      "num_tokens": 97013747.0,
      "step": 2788
    },
    {
      "epoch": 0.5179201485608171,
      "grad_norm": 1.5013683304672587,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8698375821113586,
      "num_tokens": 97048115.0,
      "step": 2789
    },
    {
      "epoch": 0.5181058495821727,
      "grad_norm": 1.4488255367028426,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8715077638626099,
      "num_tokens": 97084218.0,
      "step": 2790
    },
    {
      "epoch": 0.5182915506035283,
      "grad_norm": 1.592795129235621,
      "learning_rate": 1e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8649078011512756,
      "num_tokens": 97113901.0,
      "step": 2791
    },
    {
      "epoch": 0.5184772516248839,
      "grad_norm": 1.6605978854995922,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8610251545906067,
      "num_tokens": 97143025.0,
      "step": 2792
    },
    {
      "epoch": 0.5186629526462395,
      "grad_norm": 1.7920538559009114,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8561311364173889,
      "num_tokens": 97171014.0,
      "step": 2793
    },
    {
      "epoch": 0.5188486536675951,
      "grad_norm": 1.571768259076369,
      "learning_rate": 1e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8634731769561768,
      "num_tokens": 97206785.0,
      "step": 2794
    },
    {
      "epoch": 0.5190343546889508,
      "grad_norm": 1.5744353475068391,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8690482378005981,
      "num_tokens": 97243326.0,
      "step": 2795
    },
    {
      "epoch": 0.5192200557103064,
      "grad_norm": 1.5261181180986263,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8656908869743347,
      "num_tokens": 97275983.0,
      "step": 2796
    },
    {
      "epoch": 0.519405756731662,
      "grad_norm": 1.5618657052143274,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8745801448822021,
      "num_tokens": 97306264.0,
      "step": 2797
    },
    {
      "epoch": 0.5195914577530176,
      "grad_norm": 1.6551289528881423,
      "learning_rate": 1e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8574623465538025,
      "num_tokens": 97342236.0,
      "step": 2798
    },
    {
      "epoch": 0.5197771587743732,
      "grad_norm": 1.651032590636672,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8812564611434937,
      "num_tokens": 97374358.0,
      "step": 2799
    },
    {
      "epoch": 0.5199628597957289,
      "grad_norm": 1.5857925988144475,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.877658486366272,
      "num_tokens": 97403087.0,
      "step": 2800
    },
    {
      "epoch": 0.5201485608170845,
      "grad_norm": 1.4371582538131122,
      "learning_rate": 1e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.859759509563446,
      "num_tokens": 97442071.0,
      "step": 2801
    },
    {
      "epoch": 0.5203342618384401,
      "grad_norm": 1.6160631221424595,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.871948778629303,
      "num_tokens": 97473913.0,
      "step": 2802
    },
    {
      "epoch": 0.5205199628597957,
      "grad_norm": 1.7253513324676886,
      "learning_rate": 1e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8580005168914795,
      "num_tokens": 97500529.0,
      "step": 2803
    },
    {
      "epoch": 0.5207056638811514,
      "grad_norm": 1.4884007708876659,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8672870397567749,
      "num_tokens": 97537210.0,
      "step": 2804
    },
    {
      "epoch": 0.520891364902507,
      "grad_norm": 1.4887498987890622,
      "learning_rate": 1e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8559554219245911,
      "num_tokens": 97572340.0,
      "step": 2805
    },
    {
      "epoch": 0.5210770659238626,
      "grad_norm": 1.5097921715164135,
      "learning_rate": 1e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8583738803863525,
      "num_tokens": 97608582.0,
      "step": 2806
    },
    {
      "epoch": 0.5212627669452182,
      "grad_norm": 1.5316220754709882,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8668302297592163,
      "num_tokens": 97645911.0,
      "step": 2807
    },
    {
      "epoch": 0.5214484679665738,
      "grad_norm": 1.5674193155037852,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8647474646568298,
      "num_tokens": 97682212.0,
      "step": 2808
    },
    {
      "epoch": 0.5216341689879295,
      "grad_norm": 1.6346819753046162,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8648872971534729,
      "num_tokens": 97714627.0,
      "step": 2809
    },
    {
      "epoch": 0.5218198700092851,
      "grad_norm": 1.5013462617790205,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.85955810546875,
      "num_tokens": 97750452.0,
      "step": 2810
    },
    {
      "epoch": 0.5220055710306407,
      "grad_norm": 1.5644616558701203,
      "learning_rate": 1e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8564329147338867,
      "num_tokens": 97783191.0,
      "step": 2811
    },
    {
      "epoch": 0.5221912720519963,
      "grad_norm": 1.6017121016796139,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.879010796546936,
      "num_tokens": 97810719.0,
      "step": 2812
    },
    {
      "epoch": 0.522376973073352,
      "grad_norm": 1.4909749774013792,
      "learning_rate": 1e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8540467023849487,
      "num_tokens": 97846223.0,
      "step": 2813
    },
    {
      "epoch": 0.5225626740947075,
      "grad_norm": 1.5270737983960616,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8666413426399231,
      "num_tokens": 97879198.0,
      "step": 2814
    },
    {
      "epoch": 0.5227483751160631,
      "grad_norm": 1.683821572556222,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.869661808013916,
      "num_tokens": 97907458.0,
      "step": 2815
    },
    {
      "epoch": 0.5229340761374187,
      "grad_norm": 1.548365994197455,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8609036803245544,
      "num_tokens": 97942491.0,
      "step": 2816
    },
    {
      "epoch": 0.5231197771587743,
      "grad_norm": 1.4503348685726354,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.865370512008667,
      "num_tokens": 97977747.0,
      "step": 2817
    },
    {
      "epoch": 0.52330547818013,
      "grad_norm": 1.4634068610467292,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.859792947769165,
      "num_tokens": 98012899.0,
      "step": 2818
    },
    {
      "epoch": 0.5234911792014856,
      "grad_norm": 1.574449429643063,
      "learning_rate": 1e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8535483479499817,
      "num_tokens": 98047325.0,
      "step": 2819
    },
    {
      "epoch": 0.5236768802228412,
      "grad_norm": 1.5857834572250848,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8709007501602173,
      "num_tokens": 98073296.0,
      "step": 2820
    },
    {
      "epoch": 0.5238625812441968,
      "grad_norm": 1.4657925268823768,
      "learning_rate": 1e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8529247045516968,
      "num_tokens": 98114518.0,
      "step": 2821
    },
    {
      "epoch": 0.5240482822655524,
      "grad_norm": 1.4773394195538816,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8600748777389526,
      "num_tokens": 98152675.0,
      "step": 2822
    },
    {
      "epoch": 0.5242339832869081,
      "grad_norm": 1.382397408971622,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8705881237983704,
      "num_tokens": 98193899.0,
      "step": 2823
    },
    {
      "epoch": 0.5244196843082637,
      "grad_norm": 1.4938265122911405,
      "learning_rate": 1e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8584161400794983,
      "num_tokens": 98229116.0,
      "step": 2824
    },
    {
      "epoch": 0.5246053853296193,
      "grad_norm": 1.3511678428579572,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8657841682434082,
      "num_tokens": 98268278.0,
      "step": 2825
    },
    {
      "epoch": 0.5247910863509749,
      "grad_norm": 1.4814748010026793,
      "learning_rate": 1e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.859167218208313,
      "num_tokens": 98307244.0,
      "step": 2826
    },
    {
      "epoch": 0.5249767873723306,
      "grad_norm": 1.5311120550938486,
      "learning_rate": 1e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8622195720672607,
      "num_tokens": 98345920.0,
      "step": 2827
    },
    {
      "epoch": 0.5251624883936862,
      "grad_norm": 1.2988106931692327,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8826352953910828,
      "num_tokens": 98387343.0,
      "step": 2828
    },
    {
      "epoch": 0.5253481894150418,
      "grad_norm": 1.4480783739846783,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8661586046218872,
      "num_tokens": 98427918.0,
      "step": 2829
    },
    {
      "epoch": 0.5255338904363974,
      "grad_norm": 1.5929786411356492,
      "learning_rate": 1e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8564870357513428,
      "num_tokens": 98461566.0,
      "step": 2830
    },
    {
      "epoch": 0.525719591457753,
      "grad_norm": 1.4997153168875617,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.880007266998291,
      "num_tokens": 98494213.0,
      "step": 2831
    },
    {
      "epoch": 0.5259052924791087,
      "grad_norm": 1.4556450539252688,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.877192497253418,
      "num_tokens": 98531169.0,
      "step": 2832
    },
    {
      "epoch": 0.5260909935004643,
      "grad_norm": 1.6193085482794438,
      "learning_rate": 1e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8553762435913086,
      "num_tokens": 98564574.0,
      "step": 2833
    },
    {
      "epoch": 0.5262766945218199,
      "grad_norm": 1.567039982861589,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8665999174118042,
      "num_tokens": 98595050.0,
      "step": 2834
    },
    {
      "epoch": 0.5264623955431755,
      "grad_norm": 1.6291141456711313,
      "learning_rate": 1e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8538899421691895,
      "num_tokens": 98626138.0,
      "step": 2835
    },
    {
      "epoch": 0.5266480965645312,
      "grad_norm": 1.4142847320086331,
      "learning_rate": 1e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.860814094543457,
      "num_tokens": 98665050.0,
      "step": 2836
    },
    {
      "epoch": 0.5268337975858868,
      "grad_norm": 1.3542489103230355,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.872441291809082,
      "num_tokens": 98706257.0,
      "step": 2837
    },
    {
      "epoch": 0.5270194986072423,
      "grad_norm": 1.3608631674721565,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8744219541549683,
      "num_tokens": 98746298.0,
      "step": 2838
    },
    {
      "epoch": 0.5272051996285979,
      "grad_norm": 1.4320131026638578,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8691041469573975,
      "num_tokens": 98784220.0,
      "step": 2839
    },
    {
      "epoch": 0.5273909006499535,
      "grad_norm": 1.530615103533247,
      "learning_rate": 1e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8507607579231262,
      "num_tokens": 98819583.0,
      "step": 2840
    },
    {
      "epoch": 0.5275766016713092,
      "grad_norm": 1.5161873989547647,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8631776571273804,
      "num_tokens": 98852940.0,
      "step": 2841
    },
    {
      "epoch": 0.5277623026926648,
      "grad_norm": 1.5433061015246603,
      "learning_rate": 1e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8547441959381104,
      "num_tokens": 98887900.0,
      "step": 2842
    },
    {
      "epoch": 0.5279480037140204,
      "grad_norm": 1.528103230812244,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8739967346191406,
      "num_tokens": 98923053.0,
      "step": 2843
    },
    {
      "epoch": 0.528133704735376,
      "grad_norm": 1.4673580052906827,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.872785210609436,
      "num_tokens": 98957769.0,
      "step": 2844
    },
    {
      "epoch": 0.5283194057567316,
      "grad_norm": 1.5669579427473965,
      "learning_rate": 1e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8591603636741638,
      "num_tokens": 98992361.0,
      "step": 2845
    },
    {
      "epoch": 0.5285051067780873,
      "grad_norm": 1.4240882988315673,
      "learning_rate": 1e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8617700338363647,
      "num_tokens": 99030584.0,
      "step": 2846
    },
    {
      "epoch": 0.5286908077994429,
      "grad_norm": 1.5691530482933083,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8620111346244812,
      "num_tokens": 99064023.0,
      "step": 2847
    },
    {
      "epoch": 0.5288765088207985,
      "grad_norm": 1.4482022474316274,
      "learning_rate": 1e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8668277263641357,
      "num_tokens": 99100242.0,
      "step": 2848
    },
    {
      "epoch": 0.5290622098421541,
      "grad_norm": 1.621629474820263,
      "learning_rate": 1e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8554064035415649,
      "num_tokens": 99134185.0,
      "step": 2849
    },
    {
      "epoch": 0.5292479108635098,
      "grad_norm": 1.5575107414370688,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8573987483978271,
      "num_tokens": 99168788.0,
      "step": 2850
    },
    {
      "epoch": 0.5294336118848654,
      "grad_norm": 1.6990117820748687,
      "learning_rate": 1e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8547195792198181,
      "num_tokens": 99201578.0,
      "step": 2851
    },
    {
      "epoch": 0.529619312906221,
      "grad_norm": 1.5094586966099974,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8678290843963623,
      "num_tokens": 99237348.0,
      "step": 2852
    },
    {
      "epoch": 0.5298050139275766,
      "grad_norm": 1.4733931916715892,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8728562593460083,
      "num_tokens": 99272692.0,
      "step": 2853
    },
    {
      "epoch": 0.5299907149489322,
      "grad_norm": 1.394806844941924,
      "learning_rate": 1e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8640584945678711,
      "num_tokens": 99314449.0,
      "step": 2854
    },
    {
      "epoch": 0.5301764159702879,
      "grad_norm": 1.5136646961124787,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8581292629241943,
      "num_tokens": 99353938.0,
      "step": 2855
    },
    {
      "epoch": 0.5303621169916435,
      "grad_norm": 1.4625952958679944,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8667693138122559,
      "num_tokens": 99389095.0,
      "step": 2856
    },
    {
      "epoch": 0.5305478180129991,
      "grad_norm": 1.6605220310147766,
      "learning_rate": 1e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8421355485916138,
      "num_tokens": 99418186.0,
      "step": 2857
    },
    {
      "epoch": 0.5307335190343547,
      "grad_norm": 1.5701229567499224,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8676894903182983,
      "num_tokens": 99452137.0,
      "step": 2858
    },
    {
      "epoch": 0.5309192200557104,
      "grad_norm": 1.4443180034077965,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8675656318664551,
      "num_tokens": 99492717.0,
      "step": 2859
    },
    {
      "epoch": 0.531104921077066,
      "grad_norm": 1.623280060614991,
      "learning_rate": 1e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8513433337211609,
      "num_tokens": 99521806.0,
      "step": 2860
    },
    {
      "epoch": 0.5312906220984216,
      "grad_norm": 1.420329340739875,
      "learning_rate": 1e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8682754635810852,
      "num_tokens": 99558085.0,
      "step": 2861
    },
    {
      "epoch": 0.5314763231197771,
      "grad_norm": 1.7172358568726969,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.862648606300354,
      "num_tokens": 99587780.0,
      "step": 2862
    },
    {
      "epoch": 0.5316620241411327,
      "grad_norm": 1.3731771722060353,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8656612634658813,
      "num_tokens": 99626789.0,
      "step": 2863
    },
    {
      "epoch": 0.5318477251624883,
      "grad_norm": 1.5601718058942518,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8678364753723145,
      "num_tokens": 99660783.0,
      "step": 2864
    },
    {
      "epoch": 0.532033426183844,
      "grad_norm": 1.4889038270772585,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.864603579044342,
      "num_tokens": 99698735.0,
      "step": 2865
    },
    {
      "epoch": 0.5322191272051996,
      "grad_norm": 1.634136941752622,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8768436908721924,
      "num_tokens": 99725124.0,
      "step": 2866
    },
    {
      "epoch": 0.5324048282265552,
      "grad_norm": 1.5276040745455723,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.856352686882019,
      "num_tokens": 99756815.0,
      "step": 2867
    },
    {
      "epoch": 0.5325905292479108,
      "grad_norm": 1.5926134624075936,
      "learning_rate": 1e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8617438673973083,
      "num_tokens": 99788728.0,
      "step": 2868
    },
    {
      "epoch": 0.5327762302692665,
      "grad_norm": 1.6373955453141196,
      "learning_rate": 1e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8489245772361755,
      "num_tokens": 99821028.0,
      "step": 2869
    },
    {
      "epoch": 0.5329619312906221,
      "grad_norm": 1.5420689364402,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8638663291931152,
      "num_tokens": 99857585.0,
      "step": 2870
    },
    {
      "epoch": 0.5331476323119777,
      "grad_norm": 1.5569788190194116,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.871908962726593,
      "num_tokens": 99885977.0,
      "step": 2871
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.5342188427637748,
      "learning_rate": 1e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8645557165145874,
      "num_tokens": 99921316.0,
      "step": 2872
    },
    {
      "epoch": 0.533519034354689,
      "grad_norm": 1.5443743556533918,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.862459659576416,
      "num_tokens": 99950925.0,
      "step": 2873
    },
    {
      "epoch": 0.5337047353760446,
      "grad_norm": 1.4576148528033706,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8656730651855469,
      "num_tokens": 99987140.0,
      "step": 2874
    },
    {
      "epoch": 0.5338904363974002,
      "grad_norm": 1.6387255887849954,
      "learning_rate": 1e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8512827754020691,
      "num_tokens": 100020133.0,
      "step": 2875
    },
    {
      "epoch": 0.5340761374187558,
      "grad_norm": 1.513500188969296,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8703122138977051,
      "num_tokens": 100054650.0,
      "step": 2876
    },
    {
      "epoch": 0.5342618384401114,
      "grad_norm": 1.596036929810501,
      "learning_rate": 1e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8431729674339294,
      "num_tokens": 100090141.0,
      "step": 2877
    },
    {
      "epoch": 0.5344475394614671,
      "grad_norm": 1.6200913540792425,
      "learning_rate": 1e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8591353297233582,
      "num_tokens": 100120164.0,
      "step": 2878
    },
    {
      "epoch": 0.5346332404828227,
      "grad_norm": 1.401027885487848,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.871410608291626,
      "num_tokens": 100157855.0,
      "step": 2879
    },
    {
      "epoch": 0.5348189415041783,
      "grad_norm": 1.3359881495428076,
      "learning_rate": 1e-06,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8790014982223511,
      "num_tokens": 100198020.0,
      "step": 2880
    },
    {
      "epoch": 0.5350046425255339,
      "grad_norm": 1.4243722744405007,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8691232204437256,
      "num_tokens": 100235683.0,
      "step": 2881
    },
    {
      "epoch": 0.5351903435468895,
      "grad_norm": 1.5152425860936665,
      "learning_rate": 1e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.852527379989624,
      "num_tokens": 100272393.0,
      "step": 2882
    },
    {
      "epoch": 0.5353760445682452,
      "grad_norm": 1.653726318978251,
      "learning_rate": 1e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8340209126472473,
      "num_tokens": 100304915.0,
      "step": 2883
    },
    {
      "epoch": 0.5355617455896008,
      "grad_norm": 1.4693452214244795,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8683210015296936,
      "num_tokens": 100339735.0,
      "step": 2884
    },
    {
      "epoch": 0.5357474466109564,
      "grad_norm": 1.494000020590255,
      "learning_rate": 1e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8528174757957458,
      "num_tokens": 100378188.0,
      "step": 2885
    },
    {
      "epoch": 0.5359331476323119,
      "grad_norm": 1.4284638137752017,
      "learning_rate": 1e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8549193143844604,
      "num_tokens": 100417115.0,
      "step": 2886
    },
    {
      "epoch": 0.5361188486536675,
      "grad_norm": 1.430923658089092,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8695957064628601,
      "num_tokens": 100452955.0,
      "step": 2887
    },
    {
      "epoch": 0.5363045496750232,
      "grad_norm": 1.7504491203557606,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8548953533172607,
      "num_tokens": 100483192.0,
      "step": 2888
    },
    {
      "epoch": 0.5364902506963788,
      "grad_norm": 1.5724327381529934,
      "learning_rate": 1e-06,
      "loss": 0.4292,
      "mean_token_accuracy": 0.857640266418457,
      "num_tokens": 100518788.0,
      "step": 2889
    },
    {
      "epoch": 0.5366759517177344,
      "grad_norm": 1.6560579232052703,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8727930784225464,
      "num_tokens": 100544017.0,
      "step": 2890
    },
    {
      "epoch": 0.53686165273909,
      "grad_norm": 1.4600478867468878,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8747321367263794,
      "num_tokens": 100576681.0,
      "step": 2891
    },
    {
      "epoch": 0.5370473537604457,
      "grad_norm": 1.3546765144191926,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8741657733917236,
      "num_tokens": 100617697.0,
      "step": 2892
    },
    {
      "epoch": 0.5372330547818013,
      "grad_norm": 1.6082814617262835,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8736393451690674,
      "num_tokens": 100648954.0,
      "step": 2893
    },
    {
      "epoch": 0.5374187558031569,
      "grad_norm": 1.5696800356308687,
      "learning_rate": 1e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8565725088119507,
      "num_tokens": 100682546.0,
      "step": 2894
    },
    {
      "epoch": 0.5376044568245125,
      "grad_norm": 1.5732298874221462,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8620322346687317,
      "num_tokens": 100717110.0,
      "step": 2895
    },
    {
      "epoch": 0.5377901578458681,
      "grad_norm": 1.6083899298421516,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8755669593811035,
      "num_tokens": 100749100.0,
      "step": 2896
    },
    {
      "epoch": 0.5379758588672238,
      "grad_norm": 1.4882499948705648,
      "learning_rate": 1e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8577102422714233,
      "num_tokens": 100783094.0,
      "step": 2897
    },
    {
      "epoch": 0.5381615598885794,
      "grad_norm": 1.4545318773949008,
      "learning_rate": 1e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8731969594955444,
      "num_tokens": 100820150.0,
      "step": 2898
    },
    {
      "epoch": 0.538347260909935,
      "grad_norm": 1.4196890084317826,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8730120658874512,
      "num_tokens": 100857011.0,
      "step": 2899
    },
    {
      "epoch": 0.5385329619312906,
      "grad_norm": 1.5567121193766076,
      "learning_rate": 1e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8516745567321777,
      "num_tokens": 100891488.0,
      "step": 2900
    },
    {
      "epoch": 0.5387186629526463,
      "grad_norm": 1.6818406109499233,
      "learning_rate": 1e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8502145409584045,
      "num_tokens": 100924997.0,
      "step": 2901
    },
    {
      "epoch": 0.5389043639740019,
      "grad_norm": 1.4730809999343204,
      "learning_rate": 1e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.859074592590332,
      "num_tokens": 100961585.0,
      "step": 2902
    },
    {
      "epoch": 0.5390900649953575,
      "grad_norm": 1.7363755010924833,
      "learning_rate": 1e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8297278881072998,
      "num_tokens": 100994229.0,
      "step": 2903
    },
    {
      "epoch": 0.5392757660167131,
      "grad_norm": 1.6024519099922139,
      "learning_rate": 1e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8648697137832642,
      "num_tokens": 101024064.0,
      "step": 2904
    },
    {
      "epoch": 0.5394614670380687,
      "grad_norm": 1.5200783733266139,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8671908974647522,
      "num_tokens": 101057981.0,
      "step": 2905
    },
    {
      "epoch": 0.5396471680594244,
      "grad_norm": 1.5025104543118997,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8741065263748169,
      "num_tokens": 101091520.0,
      "step": 2906
    },
    {
      "epoch": 0.53983286908078,
      "grad_norm": 1.3839925146015455,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8659889698028564,
      "num_tokens": 101128579.0,
      "step": 2907
    },
    {
      "epoch": 0.5400185701021356,
      "grad_norm": 1.4270150975852849,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8673994541168213,
      "num_tokens": 101164499.0,
      "step": 2908
    },
    {
      "epoch": 0.5402042711234912,
      "grad_norm": 1.5357746953486786,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8763875961303711,
      "num_tokens": 101195246.0,
      "step": 2909
    },
    {
      "epoch": 0.5403899721448467,
      "grad_norm": 1.35389872912575,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8756681680679321,
      "num_tokens": 101236493.0,
      "step": 2910
    },
    {
      "epoch": 0.5405756731662024,
      "grad_norm": 1.5164362177559492,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8698724508285522,
      "num_tokens": 101268747.0,
      "step": 2911
    },
    {
      "epoch": 0.540761374187558,
      "grad_norm": 1.5146827638324292,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8648736476898193,
      "num_tokens": 101306793.0,
      "step": 2912
    },
    {
      "epoch": 0.5409470752089136,
      "grad_norm": 1.712878430682924,
      "learning_rate": 1e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8479503989219666,
      "num_tokens": 101338519.0,
      "step": 2913
    },
    {
      "epoch": 0.5411327762302692,
      "grad_norm": 1.4517108006688255,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8710267543792725,
      "num_tokens": 101375681.0,
      "step": 2914
    },
    {
      "epoch": 0.5413184772516249,
      "grad_norm": 1.542511998856839,
      "learning_rate": 1e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8562476634979248,
      "num_tokens": 101411133.0,
      "step": 2915
    },
    {
      "epoch": 0.5415041782729805,
      "grad_norm": 1.4976307029928955,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.868091344833374,
      "num_tokens": 101442220.0,
      "step": 2916
    },
    {
      "epoch": 0.5416898792943361,
      "grad_norm": 1.460590926593365,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.879054069519043,
      "num_tokens": 101476955.0,
      "step": 2917
    },
    {
      "epoch": 0.5418755803156917,
      "grad_norm": 1.5076892981964392,
      "learning_rate": 1e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.852984607219696,
      "num_tokens": 101512655.0,
      "step": 2918
    },
    {
      "epoch": 0.5420612813370473,
      "grad_norm": 1.3917223951335609,
      "learning_rate": 1e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.856604814529419,
      "num_tokens": 101554794.0,
      "step": 2919
    },
    {
      "epoch": 0.542246982358403,
      "grad_norm": 1.4500136932550225,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8700080513954163,
      "num_tokens": 101591161.0,
      "step": 2920
    },
    {
      "epoch": 0.5424326833797586,
      "grad_norm": 1.6086796067979472,
      "learning_rate": 1e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8499972820281982,
      "num_tokens": 101623317.0,
      "step": 2921
    },
    {
      "epoch": 0.5426183844011142,
      "grad_norm": 1.3847709669296921,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8660696744918823,
      "num_tokens": 101659982.0,
      "step": 2922
    },
    {
      "epoch": 0.5428040854224698,
      "grad_norm": 1.480697525582776,
      "learning_rate": 1e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8552966117858887,
      "num_tokens": 101692915.0,
      "step": 2923
    },
    {
      "epoch": 0.5429897864438255,
      "grad_norm": 1.51154846685713,
      "learning_rate": 1e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8564287424087524,
      "num_tokens": 101725460.0,
      "step": 2924
    },
    {
      "epoch": 0.5431754874651811,
      "grad_norm": 1.4689344595425584,
      "learning_rate": 1e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8563834428787231,
      "num_tokens": 101760459.0,
      "step": 2925
    },
    {
      "epoch": 0.5433611884865367,
      "grad_norm": 1.4771114010831208,
      "learning_rate": 1e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8539972305297852,
      "num_tokens": 101798750.0,
      "step": 2926
    },
    {
      "epoch": 0.5435468895078923,
      "grad_norm": 1.4299880434336225,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8682900071144104,
      "num_tokens": 101834660.0,
      "step": 2927
    },
    {
      "epoch": 0.5437325905292479,
      "grad_norm": 1.4251846846103975,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.873516321182251,
      "num_tokens": 101869154.0,
      "step": 2928
    },
    {
      "epoch": 0.5439182915506036,
      "grad_norm": 1.6348550430344204,
      "learning_rate": 1e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8554781079292297,
      "num_tokens": 101901349.0,
      "step": 2929
    },
    {
      "epoch": 0.5441039925719592,
      "grad_norm": 1.647594544869226,
      "learning_rate": 1e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8524038791656494,
      "num_tokens": 101930187.0,
      "step": 2930
    },
    {
      "epoch": 0.5442896935933148,
      "grad_norm": 1.3397506614735777,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8751116394996643,
      "num_tokens": 101965500.0,
      "step": 2931
    },
    {
      "epoch": 0.5444753946146704,
      "grad_norm": 1.4527129596251633,
      "learning_rate": 1e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8563196659088135,
      "num_tokens": 102001597.0,
      "step": 2932
    },
    {
      "epoch": 0.544661095636026,
      "grad_norm": 1.4814591896495848,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8687630891799927,
      "num_tokens": 102036548.0,
      "step": 2933
    },
    {
      "epoch": 0.5448467966573816,
      "grad_norm": 1.407545607627448,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.864851713180542,
      "num_tokens": 102074246.0,
      "step": 2934
    },
    {
      "epoch": 0.5450324976787372,
      "grad_norm": 1.4875191744648462,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8614895939826965,
      "num_tokens": 102108525.0,
      "step": 2935
    },
    {
      "epoch": 0.5452181987000928,
      "grad_norm": 1.4744522737258803,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8659015893936157,
      "num_tokens": 102148468.0,
      "step": 2936
    },
    {
      "epoch": 0.5454038997214484,
      "grad_norm": 1.4847831391063722,
      "learning_rate": 1e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8507586717605591,
      "num_tokens": 102187144.0,
      "step": 2937
    },
    {
      "epoch": 0.545589600742804,
      "grad_norm": 1.5236577749246643,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8680597543716431,
      "num_tokens": 102217916.0,
      "step": 2938
    },
    {
      "epoch": 0.5457753017641597,
      "grad_norm": 1.45979730162288,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8721247911453247,
      "num_tokens": 102252482.0,
      "step": 2939
    },
    {
      "epoch": 0.5459610027855153,
      "grad_norm": 1.5237030060578431,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8606069087982178,
      "num_tokens": 102287921.0,
      "step": 2940
    },
    {
      "epoch": 0.5461467038068709,
      "grad_norm": 1.3914405523844278,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8653429746627808,
      "num_tokens": 102328120.0,
      "step": 2941
    },
    {
      "epoch": 0.5463324048282265,
      "grad_norm": 1.4394769279347461,
      "learning_rate": 1e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8554578423500061,
      "num_tokens": 102365846.0,
      "step": 2942
    },
    {
      "epoch": 0.5465181058495822,
      "grad_norm": 1.4959897906665465,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8725376129150391,
      "num_tokens": 102397989.0,
      "step": 2943
    },
    {
      "epoch": 0.5467038068709378,
      "grad_norm": 1.4839523456189765,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8602567911148071,
      "num_tokens": 102433346.0,
      "step": 2944
    },
    {
      "epoch": 0.5468895078922934,
      "grad_norm": 1.5158773080414576,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8759093880653381,
      "num_tokens": 102463711.0,
      "step": 2945
    },
    {
      "epoch": 0.547075208913649,
      "grad_norm": 1.415911752252337,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8705945611000061,
      "num_tokens": 102506294.0,
      "step": 2946
    },
    {
      "epoch": 0.5472609099350046,
      "grad_norm": 1.4606106646709578,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8670142889022827,
      "num_tokens": 102547268.0,
      "step": 2947
    },
    {
      "epoch": 0.5474466109563603,
      "grad_norm": 1.5286305644337401,
      "learning_rate": 1e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.856465756893158,
      "num_tokens": 102581220.0,
      "step": 2948
    },
    {
      "epoch": 0.5476323119777159,
      "grad_norm": 1.4429854294998052,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.859279453754425,
      "num_tokens": 102618540.0,
      "step": 2949
    },
    {
      "epoch": 0.5478180129990715,
      "grad_norm": 1.4449352874548782,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8707987070083618,
      "num_tokens": 102654683.0,
      "step": 2950
    },
    {
      "epoch": 0.5480037140204271,
      "grad_norm": 1.5028546563730942,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8698039054870605,
      "num_tokens": 102685197.0,
      "step": 2951
    },
    {
      "epoch": 0.5481894150417828,
      "grad_norm": 1.4642922318184888,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.869281530380249,
      "num_tokens": 102719899.0,
      "step": 2952
    },
    {
      "epoch": 0.5483751160631384,
      "grad_norm": 1.4893466429243012,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8712173104286194,
      "num_tokens": 102758756.0,
      "step": 2953
    },
    {
      "epoch": 0.548560817084494,
      "grad_norm": 1.5496248299509325,
      "learning_rate": 1e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.8459239602088928,
      "num_tokens": 102795019.0,
      "step": 2954
    },
    {
      "epoch": 0.5487465181058496,
      "grad_norm": 1.6133595033795065,
      "learning_rate": 1e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8441295623779297,
      "num_tokens": 102827671.0,
      "step": 2955
    },
    {
      "epoch": 0.5489322191272052,
      "grad_norm": 1.4985964915289758,
      "learning_rate": 1e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8571110963821411,
      "num_tokens": 102863094.0,
      "step": 2956
    },
    {
      "epoch": 0.5491179201485609,
      "grad_norm": 1.4956443308286338,
      "learning_rate": 1e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8655492067337036,
      "num_tokens": 102895619.0,
      "step": 2957
    },
    {
      "epoch": 0.5493036211699165,
      "grad_norm": 1.4339746372337492,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8738856315612793,
      "num_tokens": 102934176.0,
      "step": 2958
    },
    {
      "epoch": 0.549489322191272,
      "grad_norm": 1.5940626040251502,
      "learning_rate": 1e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8426952362060547,
      "num_tokens": 102970174.0,
      "step": 2959
    },
    {
      "epoch": 0.5496750232126276,
      "grad_norm": 1.4526764006714643,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8630515933036804,
      "num_tokens": 103004650.0,
      "step": 2960
    },
    {
      "epoch": 0.5498607242339832,
      "grad_norm": 1.3124081861331527,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8635169267654419,
      "num_tokens": 103048567.0,
      "step": 2961
    },
    {
      "epoch": 0.5500464252553389,
      "grad_norm": 1.5133424866773442,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8716148734092712,
      "num_tokens": 103082866.0,
      "step": 2962
    },
    {
      "epoch": 0.5502321262766945,
      "grad_norm": 1.6526282313136407,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.860297679901123,
      "num_tokens": 103112876.0,
      "step": 2963
    },
    {
      "epoch": 0.5504178272980501,
      "grad_norm": 1.4384757229879275,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8746281862258911,
      "num_tokens": 103151536.0,
      "step": 2964
    },
    {
      "epoch": 0.5506035283194057,
      "grad_norm": 1.4868996580593155,
      "learning_rate": 1e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.863506555557251,
      "num_tokens": 103187635.0,
      "step": 2965
    },
    {
      "epoch": 0.5507892293407614,
      "grad_norm": 1.410217116010587,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8784680962562561,
      "num_tokens": 103222050.0,
      "step": 2966
    },
    {
      "epoch": 0.550974930362117,
      "grad_norm": 1.5498414014919468,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8649082183837891,
      "num_tokens": 103255172.0,
      "step": 2967
    },
    {
      "epoch": 0.5511606313834726,
      "grad_norm": 1.636908113572874,
      "learning_rate": 1e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8462748527526855,
      "num_tokens": 103293503.0,
      "step": 2968
    },
    {
      "epoch": 0.5513463324048282,
      "grad_norm": 1.4533956081932222,
      "learning_rate": 1e-06,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8659592270851135,
      "num_tokens": 103330586.0,
      "step": 2969
    },
    {
      "epoch": 0.5515320334261838,
      "grad_norm": 1.6035659417195043,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.867678165435791,
      "num_tokens": 103359727.0,
      "step": 2970
    },
    {
      "epoch": 0.5517177344475395,
      "grad_norm": 1.4644143285427464,
      "learning_rate": 1e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8490311503410339,
      "num_tokens": 103397095.0,
      "step": 2971
    },
    {
      "epoch": 0.5519034354688951,
      "grad_norm": 1.4993874353200003,
      "learning_rate": 1e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8572006225585938,
      "num_tokens": 103431156.0,
      "step": 2972
    },
    {
      "epoch": 0.5520891364902507,
      "grad_norm": 1.5633387964398566,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.863628089427948,
      "num_tokens": 103466861.0,
      "step": 2973
    },
    {
      "epoch": 0.5522748375116063,
      "grad_norm": 1.5356075438089718,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8674821853637695,
      "num_tokens": 103502269.0,
      "step": 2974
    },
    {
      "epoch": 0.552460538532962,
      "grad_norm": 1.5895032058205973,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8588032722473145,
      "num_tokens": 103533526.0,
      "step": 2975
    },
    {
      "epoch": 0.5526462395543176,
      "grad_norm": 1.577824801612155,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8616085648536682,
      "num_tokens": 103564641.0,
      "step": 2976
    },
    {
      "epoch": 0.5528319405756732,
      "grad_norm": 1.625624088886179,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8642791509628296,
      "num_tokens": 103597215.0,
      "step": 2977
    },
    {
      "epoch": 0.5530176415970288,
      "grad_norm": 1.4426752909710365,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8667638301849365,
      "num_tokens": 103633051.0,
      "step": 2978
    },
    {
      "epoch": 0.5532033426183844,
      "grad_norm": 1.4311421935322957,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8704190850257874,
      "num_tokens": 103670688.0,
      "step": 2979
    },
    {
      "epoch": 0.5533890436397401,
      "grad_norm": 1.5083648782792034,
      "learning_rate": 1e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8667758107185364,
      "num_tokens": 103704204.0,
      "step": 2980
    },
    {
      "epoch": 0.5535747446610957,
      "grad_norm": 1.500960772784753,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.860176682472229,
      "num_tokens": 103741628.0,
      "step": 2981
    },
    {
      "epoch": 0.5537604456824513,
      "grad_norm": 1.4907933684158787,
      "learning_rate": 1e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8684654235839844,
      "num_tokens": 103775886.0,
      "step": 2982
    },
    {
      "epoch": 0.5539461467038068,
      "grad_norm": 1.4196778033556394,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8723414540290833,
      "num_tokens": 103813289.0,
      "step": 2983
    },
    {
      "epoch": 0.5541318477251624,
      "grad_norm": 1.4033037622534752,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8688774108886719,
      "num_tokens": 103848905.0,
      "step": 2984
    },
    {
      "epoch": 0.5543175487465181,
      "grad_norm": 1.573370651302189,
      "learning_rate": 1e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8543153405189514,
      "num_tokens": 103883561.0,
      "step": 2985
    },
    {
      "epoch": 0.5545032497678737,
      "grad_norm": 1.619817671355523,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.864338755607605,
      "num_tokens": 103913622.0,
      "step": 2986
    },
    {
      "epoch": 0.5546889507892293,
      "grad_norm": 1.6996527406036863,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8683823347091675,
      "num_tokens": 103940428.0,
      "step": 2987
    },
    {
      "epoch": 0.5548746518105849,
      "grad_norm": 1.4511534408244964,
      "learning_rate": 1e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8504369258880615,
      "num_tokens": 103978295.0,
      "step": 2988
    },
    {
      "epoch": 0.5550603528319406,
      "grad_norm": 1.4384620865537314,
      "learning_rate": 1e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8610455989837646,
      "num_tokens": 104012569.0,
      "step": 2989
    },
    {
      "epoch": 0.5552460538532962,
      "grad_norm": 1.4239836459652282,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8623053431510925,
      "num_tokens": 104049192.0,
      "step": 2990
    },
    {
      "epoch": 0.5554317548746518,
      "grad_norm": 1.5332395203274112,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8664220571517944,
      "num_tokens": 104081436.0,
      "step": 2991
    },
    {
      "epoch": 0.5556174558960074,
      "grad_norm": 1.4052185451278336,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8740084171295166,
      "num_tokens": 104116520.0,
      "step": 2992
    },
    {
      "epoch": 0.555803156917363,
      "grad_norm": 1.4070953622833362,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8718984127044678,
      "num_tokens": 104154541.0,
      "step": 2993
    },
    {
      "epoch": 0.5559888579387187,
      "grad_norm": 1.3193520900160618,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8809183835983276,
      "num_tokens": 104193171.0,
      "step": 2994
    },
    {
      "epoch": 0.5561745589600743,
      "grad_norm": 1.4733079972184953,
      "learning_rate": 1e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.852556586265564,
      "num_tokens": 104227446.0,
      "step": 2995
    },
    {
      "epoch": 0.5563602599814299,
      "grad_norm": 1.5396575195299256,
      "learning_rate": 1e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8563507795333862,
      "num_tokens": 104262089.0,
      "step": 2996
    },
    {
      "epoch": 0.5565459610027855,
      "grad_norm": 1.5281062080440597,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.865524411201477,
      "num_tokens": 104293153.0,
      "step": 2997
    },
    {
      "epoch": 0.5567316620241411,
      "grad_norm": 1.4714361786135757,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8670596480369568,
      "num_tokens": 104326650.0,
      "step": 2998
    },
    {
      "epoch": 0.5569173630454968,
      "grad_norm": 1.2913424411223973,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8727285265922546,
      "num_tokens": 104370429.0,
      "step": 2999
    },
    {
      "epoch": 0.5571030640668524,
      "grad_norm": 1.4692435419890078,
      "learning_rate": 1e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8577518463134766,
      "num_tokens": 104404658.0,
      "step": 3000
    },
    {
      "epoch": 0.557288765088208,
      "grad_norm": 1.3689794749286393,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8688611388206482,
      "num_tokens": 104442457.0,
      "step": 3001
    },
    {
      "epoch": 0.5574744661095636,
      "grad_norm": 1.4171729520610552,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8651511669158936,
      "num_tokens": 104478329.0,
      "step": 3002
    },
    {
      "epoch": 0.5576601671309193,
      "grad_norm": 1.6111328187868474,
      "learning_rate": 1e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8509615659713745,
      "num_tokens": 104510198.0,
      "step": 3003
    },
    {
      "epoch": 0.5578458681522749,
      "grad_norm": 1.3658320341536978,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8782284259796143,
      "num_tokens": 104547775.0,
      "step": 3004
    },
    {
      "epoch": 0.5580315691736305,
      "grad_norm": 1.4067164084476889,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8748517036437988,
      "num_tokens": 104582307.0,
      "step": 3005
    },
    {
      "epoch": 0.5582172701949861,
      "grad_norm": 1.4423386586465157,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.873904824256897,
      "num_tokens": 104615175.0,
      "step": 3006
    },
    {
      "epoch": 0.5584029712163416,
      "grad_norm": 1.43966412093054,
      "learning_rate": 1e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8640309572219849,
      "num_tokens": 104653400.0,
      "step": 3007
    },
    {
      "epoch": 0.5585886722376973,
      "grad_norm": 1.4603199436959573,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.864680290222168,
      "num_tokens": 104688314.0,
      "step": 3008
    },
    {
      "epoch": 0.5587743732590529,
      "grad_norm": 1.396691207780059,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8724046945571899,
      "num_tokens": 104728993.0,
      "step": 3009
    },
    {
      "epoch": 0.5589600742804085,
      "grad_norm": 1.5278629863331845,
      "learning_rate": 1e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8584188222885132,
      "num_tokens": 104764066.0,
      "step": 3010
    },
    {
      "epoch": 0.5591457753017641,
      "grad_norm": 1.6496914441640622,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8685165643692017,
      "num_tokens": 104795948.0,
      "step": 3011
    },
    {
      "epoch": 0.5593314763231197,
      "grad_norm": 1.7310515972116876,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8614895343780518,
      "num_tokens": 104821847.0,
      "step": 3012
    },
    {
      "epoch": 0.5595171773444754,
      "grad_norm": 1.6415150215565308,
      "learning_rate": 1e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8581854104995728,
      "num_tokens": 104852503.0,
      "step": 3013
    },
    {
      "epoch": 0.559702878365831,
      "grad_norm": 1.458103544185668,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8667424917221069,
      "num_tokens": 104891750.0,
      "step": 3014
    },
    {
      "epoch": 0.5598885793871866,
      "grad_norm": 1.3937285218149238,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8717979788780212,
      "num_tokens": 104927867.0,
      "step": 3015
    },
    {
      "epoch": 0.5600742804085422,
      "grad_norm": 1.4202730697338166,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8785147666931152,
      "num_tokens": 104963193.0,
      "step": 3016
    },
    {
      "epoch": 0.5602599814298979,
      "grad_norm": 1.4343566269003316,
      "learning_rate": 1e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8580384254455566,
      "num_tokens": 105001422.0,
      "step": 3017
    },
    {
      "epoch": 0.5604456824512535,
      "grad_norm": 1.373744370197073,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8672364950180054,
      "num_tokens": 105042440.0,
      "step": 3018
    },
    {
      "epoch": 0.5606313834726091,
      "grad_norm": 1.5754369373888337,
      "learning_rate": 1e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8567699193954468,
      "num_tokens": 105073766.0,
      "step": 3019
    },
    {
      "epoch": 0.5608170844939647,
      "grad_norm": 1.6135308973143945,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8708217144012451,
      "num_tokens": 105105296.0,
      "step": 3020
    },
    {
      "epoch": 0.5610027855153203,
      "grad_norm": 1.4935734028176557,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8719401359558105,
      "num_tokens": 105136333.0,
      "step": 3021
    },
    {
      "epoch": 0.561188486536676,
      "grad_norm": 1.5749444743639027,
      "learning_rate": 1e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8432673215866089,
      "num_tokens": 105168328.0,
      "step": 3022
    },
    {
      "epoch": 0.5613741875580316,
      "grad_norm": 1.5464831222709448,
      "learning_rate": 1e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8600081205368042,
      "num_tokens": 105200791.0,
      "step": 3023
    },
    {
      "epoch": 0.5615598885793872,
      "grad_norm": 1.5416375488759457,
      "learning_rate": 1e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8661925792694092,
      "num_tokens": 105230130.0,
      "step": 3024
    },
    {
      "epoch": 0.5617455896007428,
      "grad_norm": 1.440680417761198,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.866668701171875,
      "num_tokens": 105268668.0,
      "step": 3025
    },
    {
      "epoch": 0.5619312906220985,
      "grad_norm": 1.4814775161606326,
      "learning_rate": 1e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8718854188919067,
      "num_tokens": 105305383.0,
      "step": 3026
    },
    {
      "epoch": 0.5621169916434541,
      "grad_norm": 1.3490642404086912,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8743233680725098,
      "num_tokens": 105344048.0,
      "step": 3027
    },
    {
      "epoch": 0.5623026926648097,
      "grad_norm": 1.7338588951565643,
      "learning_rate": 1e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8667853474617004,
      "num_tokens": 105369752.0,
      "step": 3028
    },
    {
      "epoch": 0.5624883936861653,
      "grad_norm": 1.4559732147620552,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8644461035728455,
      "num_tokens": 105407654.0,
      "step": 3029
    },
    {
      "epoch": 0.5626740947075209,
      "grad_norm": 1.541621819150831,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8581950664520264,
      "num_tokens": 105440965.0,
      "step": 3030
    },
    {
      "epoch": 0.5628597957288765,
      "grad_norm": 1.5393436471461752,
      "learning_rate": 1e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8600356578826904,
      "num_tokens": 105477998.0,
      "step": 3031
    },
    {
      "epoch": 0.5630454967502321,
      "grad_norm": 1.3993720390553732,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8669774532318115,
      "num_tokens": 105515492.0,
      "step": 3032
    },
    {
      "epoch": 0.5632311977715877,
      "grad_norm": 1.528616246388864,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8842512369155884,
      "num_tokens": 105543531.0,
      "step": 3033
    },
    {
      "epoch": 0.5634168987929433,
      "grad_norm": 1.5740819516340299,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8709189891815186,
      "num_tokens": 105575255.0,
      "step": 3034
    },
    {
      "epoch": 0.5636025998142989,
      "grad_norm": 1.4946564277441758,
      "learning_rate": 1e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8564995527267456,
      "num_tokens": 105611094.0,
      "step": 3035
    },
    {
      "epoch": 0.5637883008356546,
      "grad_norm": 1.4976188064510738,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8684770464897156,
      "num_tokens": 105642163.0,
      "step": 3036
    },
    {
      "epoch": 0.5639740018570102,
      "grad_norm": 1.6360175611593202,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8573126196861267,
      "num_tokens": 105670225.0,
      "step": 3037
    },
    {
      "epoch": 0.5641597028783658,
      "grad_norm": 1.3769163515011522,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8742347955703735,
      "num_tokens": 105709886.0,
      "step": 3038
    },
    {
      "epoch": 0.5643454038997214,
      "grad_norm": 1.4359593549805598,
      "learning_rate": 1e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8613690137863159,
      "num_tokens": 105747290.0,
      "step": 3039
    },
    {
      "epoch": 0.564531104921077,
      "grad_norm": 1.4769297189457278,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8676970601081848,
      "num_tokens": 105782740.0,
      "step": 3040
    },
    {
      "epoch": 0.5647168059424327,
      "grad_norm": 1.3548749907473876,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.881484866142273,
      "num_tokens": 105820482.0,
      "step": 3041
    },
    {
      "epoch": 0.5649025069637883,
      "grad_norm": 1.5880991775653273,
      "learning_rate": 1e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8495601415634155,
      "num_tokens": 105852951.0,
      "step": 3042
    },
    {
      "epoch": 0.5650882079851439,
      "grad_norm": 1.502495860011552,
      "learning_rate": 1e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8491477966308594,
      "num_tokens": 105891219.0,
      "step": 3043
    },
    {
      "epoch": 0.5652739090064995,
      "grad_norm": 1.496795459972521,
      "learning_rate": 1e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8612135648727417,
      "num_tokens": 105927982.0,
      "step": 3044
    },
    {
      "epoch": 0.5654596100278552,
      "grad_norm": 1.4683988916357213,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8676337003707886,
      "num_tokens": 105961471.0,
      "step": 3045
    },
    {
      "epoch": 0.5656453110492108,
      "grad_norm": 1.3908707296170302,
      "learning_rate": 1e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.854593813419342,
      "num_tokens": 106004457.0,
      "step": 3046
    },
    {
      "epoch": 0.5658310120705664,
      "grad_norm": 1.3708518513776133,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8635493516921997,
      "num_tokens": 106042293.0,
      "step": 3047
    },
    {
      "epoch": 0.566016713091922,
      "grad_norm": 1.3718757135436108,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8813780546188354,
      "num_tokens": 106079230.0,
      "step": 3048
    },
    {
      "epoch": 0.5662024141132777,
      "grad_norm": 1.463084469991656,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8665852546691895,
      "num_tokens": 106115492.0,
      "step": 3049
    },
    {
      "epoch": 0.5663881151346333,
      "grad_norm": 1.565409451341033,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8608063459396362,
      "num_tokens": 106146786.0,
      "step": 3050
    },
    {
      "epoch": 0.5665738161559889,
      "grad_norm": 1.4979724539873298,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8773736953735352,
      "num_tokens": 106181980.0,
      "step": 3051
    },
    {
      "epoch": 0.5667595171773445,
      "grad_norm": 1.348763387771195,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8755171298980713,
      "num_tokens": 106220559.0,
      "step": 3052
    },
    {
      "epoch": 0.5669452181987001,
      "grad_norm": 1.545347823891838,
      "learning_rate": 1e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8583249449729919,
      "num_tokens": 106256892.0,
      "step": 3053
    },
    {
      "epoch": 0.5671309192200558,
      "grad_norm": 1.4865136589169619,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8666278123855591,
      "num_tokens": 106290780.0,
      "step": 3054
    },
    {
      "epoch": 0.5673166202414113,
      "grad_norm": 1.305153215433133,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8745290040969849,
      "num_tokens": 106330407.0,
      "step": 3055
    },
    {
      "epoch": 0.5675023212627669,
      "grad_norm": 1.4656329881791186,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8679249882698059,
      "num_tokens": 106363391.0,
      "step": 3056
    },
    {
      "epoch": 0.5676880222841225,
      "grad_norm": 1.4077938064708055,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8760748505592346,
      "num_tokens": 106397527.0,
      "step": 3057
    },
    {
      "epoch": 0.5678737233054781,
      "grad_norm": 1.618552658204293,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8702637553215027,
      "num_tokens": 106426701.0,
      "step": 3058
    },
    {
      "epoch": 0.5680594243268338,
      "grad_norm": 1.4500816263396659,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8735389709472656,
      "num_tokens": 106461679.0,
      "step": 3059
    },
    {
      "epoch": 0.5682451253481894,
      "grad_norm": 1.3499976641273845,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8642489910125732,
      "num_tokens": 106502601.0,
      "step": 3060
    },
    {
      "epoch": 0.568430826369545,
      "grad_norm": 1.4793699185138915,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8749222755432129,
      "num_tokens": 106535983.0,
      "step": 3061
    },
    {
      "epoch": 0.5686165273909006,
      "grad_norm": 1.6643067449168363,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8587049245834351,
      "num_tokens": 106569181.0,
      "step": 3062
    },
    {
      "epoch": 0.5688022284122562,
      "grad_norm": 1.5632686304809553,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8695759773254395,
      "num_tokens": 106599236.0,
      "step": 3063
    },
    {
      "epoch": 0.5689879294336119,
      "grad_norm": 1.578381431603595,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8755886554718018,
      "num_tokens": 106626979.0,
      "step": 3064
    },
    {
      "epoch": 0.5691736304549675,
      "grad_norm": 1.5385036334458695,
      "learning_rate": 1e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8589386343955994,
      "num_tokens": 106661524.0,
      "step": 3065
    },
    {
      "epoch": 0.5693593314763231,
      "grad_norm": 1.4160751321841627,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8844336271286011,
      "num_tokens": 106694015.0,
      "step": 3066
    },
    {
      "epoch": 0.5695450324976787,
      "grad_norm": 1.4754215316778274,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8548322319984436,
      "num_tokens": 106732750.0,
      "step": 3067
    },
    {
      "epoch": 0.5697307335190344,
      "grad_norm": 1.5347524554832228,
      "learning_rate": 1e-06,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8666175007820129,
      "num_tokens": 106765725.0,
      "step": 3068
    },
    {
      "epoch": 0.56991643454039,
      "grad_norm": 1.549882158104744,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8562541604042053,
      "num_tokens": 106806990.0,
      "step": 3069
    },
    {
      "epoch": 0.5701021355617456,
      "grad_norm": 1.548373487803244,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8665460348129272,
      "num_tokens": 106840703.0,
      "step": 3070
    },
    {
      "epoch": 0.5702878365831012,
      "grad_norm": 1.5086323384600349,
      "learning_rate": 1e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8548856973648071,
      "num_tokens": 106872008.0,
      "step": 3071
    },
    {
      "epoch": 0.5704735376044568,
      "grad_norm": 1.3938911401919227,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8804842233657837,
      "num_tokens": 106907404.0,
      "step": 3072
    },
    {
      "epoch": 0.5706592386258125,
      "grad_norm": 1.4092653413914102,
      "learning_rate": 1e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.864336371421814,
      "num_tokens": 106943527.0,
      "step": 3073
    },
    {
      "epoch": 0.5708449396471681,
      "grad_norm": 1.479164950511097,
      "learning_rate": 1e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8654434680938721,
      "num_tokens": 106978332.0,
      "step": 3074
    },
    {
      "epoch": 0.5710306406685237,
      "grad_norm": 1.5350390828952054,
      "learning_rate": 1e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8553485870361328,
      "num_tokens": 107013852.0,
      "step": 3075
    },
    {
      "epoch": 0.5712163416898793,
      "grad_norm": 1.5364581081180222,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8717767596244812,
      "num_tokens": 107049142.0,
      "step": 3076
    },
    {
      "epoch": 0.571402042711235,
      "grad_norm": 1.3739875454881114,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8761137127876282,
      "num_tokens": 107083307.0,
      "step": 3077
    },
    {
      "epoch": 0.5715877437325906,
      "grad_norm": 1.3418872303279696,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8657957315444946,
      "num_tokens": 107125694.0,
      "step": 3078
    },
    {
      "epoch": 0.5717734447539461,
      "grad_norm": 1.424788918985943,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8657060861587524,
      "num_tokens": 107162003.0,
      "step": 3079
    },
    {
      "epoch": 0.5719591457753017,
      "grad_norm": 1.4528361945094534,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8722444772720337,
      "num_tokens": 107198314.0,
      "step": 3080
    },
    {
      "epoch": 0.5721448467966573,
      "grad_norm": 1.459306536012174,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8652503490447998,
      "num_tokens": 107235387.0,
      "step": 3081
    },
    {
      "epoch": 0.572330547818013,
      "grad_norm": 1.554314175327879,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8663516044616699,
      "num_tokens": 107267929.0,
      "step": 3082
    },
    {
      "epoch": 0.5725162488393686,
      "grad_norm": 1.3825648468043885,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.883086085319519,
      "num_tokens": 107307194.0,
      "step": 3083
    },
    {
      "epoch": 0.5727019498607242,
      "grad_norm": 1.5918964991765825,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8620948791503906,
      "num_tokens": 107338886.0,
      "step": 3084
    },
    {
      "epoch": 0.5728876508820798,
      "grad_norm": 1.5002436524627332,
      "learning_rate": 1e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8575789928436279,
      "num_tokens": 107374806.0,
      "step": 3085
    },
    {
      "epoch": 0.5730733519034354,
      "grad_norm": 1.6134410721851662,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.865683913230896,
      "num_tokens": 107405994.0,
      "step": 3086
    },
    {
      "epoch": 0.5732590529247911,
      "grad_norm": 1.7847609778037914,
      "learning_rate": 1e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8580417633056641,
      "num_tokens": 107431327.0,
      "step": 3087
    },
    {
      "epoch": 0.5734447539461467,
      "grad_norm": 1.4801881352642898,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8766569495201111,
      "num_tokens": 107467018.0,
      "step": 3088
    },
    {
      "epoch": 0.5736304549675023,
      "grad_norm": 1.3851923044293468,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8631242513656616,
      "num_tokens": 107504653.0,
      "step": 3089
    },
    {
      "epoch": 0.5738161559888579,
      "grad_norm": 1.5160210322916388,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8667333722114563,
      "num_tokens": 107534154.0,
      "step": 3090
    },
    {
      "epoch": 0.5740018570102136,
      "grad_norm": 1.6094782718765206,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8708841800689697,
      "num_tokens": 107563213.0,
      "step": 3091
    },
    {
      "epoch": 0.5741875580315692,
      "grad_norm": 1.447636032919081,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8545387387275696,
      "num_tokens": 107598435.0,
      "step": 3092
    },
    {
      "epoch": 0.5743732590529248,
      "grad_norm": 1.3755715782688487,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8699972033500671,
      "num_tokens": 107635386.0,
      "step": 3093
    },
    {
      "epoch": 0.5745589600742804,
      "grad_norm": 1.4838401601876696,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8692429661750793,
      "num_tokens": 107666934.0,
      "step": 3094
    },
    {
      "epoch": 0.574744661095636,
      "grad_norm": 1.3807845547840087,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8795523047447205,
      "num_tokens": 107703014.0,
      "step": 3095
    },
    {
      "epoch": 0.5749303621169917,
      "grad_norm": 1.5601453223385815,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8599755167961121,
      "num_tokens": 107733585.0,
      "step": 3096
    },
    {
      "epoch": 0.5751160631383473,
      "grad_norm": 1.5523906155536458,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8571712970733643,
      "num_tokens": 107769618.0,
      "step": 3097
    },
    {
      "epoch": 0.5753017641597029,
      "grad_norm": 1.4382938033807855,
      "learning_rate": 1e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8498560786247253,
      "num_tokens": 107811873.0,
      "step": 3098
    },
    {
      "epoch": 0.5754874651810585,
      "grad_norm": 1.633552276810943,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.870396614074707,
      "num_tokens": 107840731.0,
      "step": 3099
    },
    {
      "epoch": 0.5756731662024142,
      "grad_norm": 1.7563838766792121,
      "learning_rate": 1e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8372960686683655,
      "num_tokens": 107874148.0,
      "step": 3100
    },
    {
      "epoch": 0.5758588672237698,
      "grad_norm": 1.4276783380248974,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8698575496673584,
      "num_tokens": 107910092.0,
      "step": 3101
    },
    {
      "epoch": 0.5760445682451254,
      "grad_norm": 1.4445694510473561,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8745924830436707,
      "num_tokens": 107944040.0,
      "step": 3102
    },
    {
      "epoch": 0.576230269266481,
      "grad_norm": 1.676771515452603,
      "learning_rate": 1e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8506768345832825,
      "num_tokens": 107973018.0,
      "step": 3103
    },
    {
      "epoch": 0.5764159702878365,
      "grad_norm": 1.572394005514476,
      "learning_rate": 1e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8584704995155334,
      "num_tokens": 108008142.0,
      "step": 3104
    },
    {
      "epoch": 0.5766016713091922,
      "grad_norm": 1.4957813705435334,
      "learning_rate": 1e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8552176356315613,
      "num_tokens": 108042215.0,
      "step": 3105
    },
    {
      "epoch": 0.5767873723305478,
      "grad_norm": 1.5254471314552693,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8617925643920898,
      "num_tokens": 108078321.0,
      "step": 3106
    },
    {
      "epoch": 0.5769730733519034,
      "grad_norm": 1.4457702621930466,
      "learning_rate": 1e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8614834547042847,
      "num_tokens": 108114192.0,
      "step": 3107
    },
    {
      "epoch": 0.577158774373259,
      "grad_norm": 1.4819677111788725,
      "learning_rate": 1e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8559786677360535,
      "num_tokens": 108154224.0,
      "step": 3108
    },
    {
      "epoch": 0.5773444753946146,
      "grad_norm": 1.443111114229524,
      "learning_rate": 1e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8633042573928833,
      "num_tokens": 108192399.0,
      "step": 3109
    },
    {
      "epoch": 0.5775301764159703,
      "grad_norm": 1.508987343398233,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8628246188163757,
      "num_tokens": 108225430.0,
      "step": 3110
    },
    {
      "epoch": 0.5777158774373259,
      "grad_norm": 1.5226525381006584,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8654994964599609,
      "num_tokens": 108258935.0,
      "step": 3111
    },
    {
      "epoch": 0.5779015784586815,
      "grad_norm": 1.412615249045313,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8719308376312256,
      "num_tokens": 108299211.0,
      "step": 3112
    },
    {
      "epoch": 0.5780872794800371,
      "grad_norm": 1.3530074874273594,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8670922517776489,
      "num_tokens": 108341774.0,
      "step": 3113
    },
    {
      "epoch": 0.5782729805013928,
      "grad_norm": 1.6134051467153345,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8586240410804749,
      "num_tokens": 108372799.0,
      "step": 3114
    },
    {
      "epoch": 0.5784586815227484,
      "grad_norm": 1.5332696893150657,
      "learning_rate": 1e-06,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8662579655647278,
      "num_tokens": 108404426.0,
      "step": 3115
    },
    {
      "epoch": 0.578644382544104,
      "grad_norm": 1.4137631586535875,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.879859447479248,
      "num_tokens": 108437479.0,
      "step": 3116
    },
    {
      "epoch": 0.5788300835654596,
      "grad_norm": 1.384769917578883,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8789792060852051,
      "num_tokens": 108474498.0,
      "step": 3117
    },
    {
      "epoch": 0.5790157845868152,
      "grad_norm": 1.4653038291507783,
      "learning_rate": 1e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8559737801551819,
      "num_tokens": 108507859.0,
      "step": 3118
    },
    {
      "epoch": 0.5792014856081709,
      "grad_norm": 1.4567438124475114,
      "learning_rate": 1e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.844527006149292,
      "num_tokens": 108545898.0,
      "step": 3119
    },
    {
      "epoch": 0.5793871866295265,
      "grad_norm": 1.5115758276298144,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8638838529586792,
      "num_tokens": 108578353.0,
      "step": 3120
    },
    {
      "epoch": 0.5795728876508821,
      "grad_norm": 1.5104425716914205,
      "learning_rate": 1e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8543314933776855,
      "num_tokens": 108613239.0,
      "step": 3121
    },
    {
      "epoch": 0.5797585886722377,
      "grad_norm": 1.5874627881533,
      "learning_rate": 1e-06,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8598906993865967,
      "num_tokens": 108642957.0,
      "step": 3122
    },
    {
      "epoch": 0.5799442896935934,
      "grad_norm": 1.4878032463217568,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8770961761474609,
      "num_tokens": 108676620.0,
      "step": 3123
    },
    {
      "epoch": 0.580129990714949,
      "grad_norm": 1.7641012716600555,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8572795987129211,
      "num_tokens": 108707901.0,
      "step": 3124
    },
    {
      "epoch": 0.5803156917363046,
      "grad_norm": 1.4692092739620093,
      "learning_rate": 1e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8412283062934875,
      "num_tokens": 108746868.0,
      "step": 3125
    },
    {
      "epoch": 0.5805013927576602,
      "grad_norm": 1.4055536101073267,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8736226558685303,
      "num_tokens": 108782777.0,
      "step": 3126
    },
    {
      "epoch": 0.5806870937790158,
      "grad_norm": 1.5197536565059764,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8774183988571167,
      "num_tokens": 108810780.0,
      "step": 3127
    },
    {
      "epoch": 0.5808727948003714,
      "grad_norm": 1.5553904497624276,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8612798452377319,
      "num_tokens": 108845424.0,
      "step": 3128
    },
    {
      "epoch": 0.581058495821727,
      "grad_norm": 1.4950239196376225,
      "learning_rate": 1e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8410753011703491,
      "num_tokens": 108883648.0,
      "step": 3129
    },
    {
      "epoch": 0.5812441968430826,
      "grad_norm": 1.5064170092921119,
      "learning_rate": 1e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8595465421676636,
      "num_tokens": 108920480.0,
      "step": 3130
    },
    {
      "epoch": 0.5814298978644382,
      "grad_norm": 1.514808841234284,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.878115177154541,
      "num_tokens": 108948320.0,
      "step": 3131
    },
    {
      "epoch": 0.5816155988857938,
      "grad_norm": 1.5993645935505074,
      "learning_rate": 1e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8544814586639404,
      "num_tokens": 108977890.0,
      "step": 3132
    },
    {
      "epoch": 0.5818012999071495,
      "grad_norm": 1.4471435301128357,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8663904666900635,
      "num_tokens": 109018132.0,
      "step": 3133
    },
    {
      "epoch": 0.5819870009285051,
      "grad_norm": 1.4265575332429852,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8788886070251465,
      "num_tokens": 109054725.0,
      "step": 3134
    },
    {
      "epoch": 0.5821727019498607,
      "grad_norm": 1.430361009833404,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8713563680648804,
      "num_tokens": 109091063.0,
      "step": 3135
    },
    {
      "epoch": 0.5823584029712163,
      "grad_norm": 1.40062602287018,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8542946577072144,
      "num_tokens": 109130348.0,
      "step": 3136
    },
    {
      "epoch": 0.582544103992572,
      "grad_norm": 1.4396841848770896,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8671690821647644,
      "num_tokens": 109165214.0,
      "step": 3137
    },
    {
      "epoch": 0.5827298050139276,
      "grad_norm": 1.6328856243184977,
      "learning_rate": 1e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8508051037788391,
      "num_tokens": 109202034.0,
      "step": 3138
    },
    {
      "epoch": 0.5829155060352832,
      "grad_norm": 1.56771469657705,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8573094606399536,
      "num_tokens": 109234639.0,
      "step": 3139
    },
    {
      "epoch": 0.5831012070566388,
      "grad_norm": 1.3777854821489486,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8702403903007507,
      "num_tokens": 109271311.0,
      "step": 3140
    },
    {
      "epoch": 0.5832869080779944,
      "grad_norm": 1.3407661377438544,
      "learning_rate": 1e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.854903519153595,
      "num_tokens": 109316356.0,
      "step": 3141
    },
    {
      "epoch": 0.5834726090993501,
      "grad_norm": 1.535467565796327,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8590377569198608,
      "num_tokens": 109350414.0,
      "step": 3142
    },
    {
      "epoch": 0.5836583101207057,
      "grad_norm": 1.4796971742870497,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8753651976585388,
      "num_tokens": 109384588.0,
      "step": 3143
    },
    {
      "epoch": 0.5838440111420613,
      "grad_norm": 1.3420644708830365,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8708974719047546,
      "num_tokens": 109425125.0,
      "step": 3144
    },
    {
      "epoch": 0.5840297121634169,
      "grad_norm": 1.4592405484503341,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8694427609443665,
      "num_tokens": 109461210.0,
      "step": 3145
    },
    {
      "epoch": 0.5842154131847725,
      "grad_norm": 1.4305349984621185,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8731756806373596,
      "num_tokens": 109493280.0,
      "step": 3146
    },
    {
      "epoch": 0.5844011142061282,
      "grad_norm": 1.5551682171314603,
      "learning_rate": 1e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8584884405136108,
      "num_tokens": 109525475.0,
      "step": 3147
    },
    {
      "epoch": 0.5845868152274838,
      "grad_norm": 1.43249966301935,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.874910831451416,
      "num_tokens": 109561479.0,
      "step": 3148
    },
    {
      "epoch": 0.5847725162488394,
      "grad_norm": 1.712450582652551,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8748882412910461,
      "num_tokens": 109592875.0,
      "step": 3149
    },
    {
      "epoch": 0.584958217270195,
      "grad_norm": 1.4302659854669468,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.863758385181427,
      "num_tokens": 109630976.0,
      "step": 3150
    },
    {
      "epoch": 0.5851439182915507,
      "grad_norm": 1.4213132839860494,
      "learning_rate": 1e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8501296043395996,
      "num_tokens": 109672032.0,
      "step": 3151
    },
    {
      "epoch": 0.5853296193129062,
      "grad_norm": 1.5329662213368187,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8604112267494202,
      "num_tokens": 109705299.0,
      "step": 3152
    },
    {
      "epoch": 0.5855153203342618,
      "grad_norm": 1.5367720211854399,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.867457389831543,
      "num_tokens": 109735056.0,
      "step": 3153
    },
    {
      "epoch": 0.5857010213556174,
      "grad_norm": 1.6242256148192333,
      "learning_rate": 1e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8574360013008118,
      "num_tokens": 109763289.0,
      "step": 3154
    },
    {
      "epoch": 0.585886722376973,
      "grad_norm": 1.403549891626624,
      "learning_rate": 1e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.847070574760437,
      "num_tokens": 109804744.0,
      "step": 3155
    },
    {
      "epoch": 0.5860724233983287,
      "grad_norm": 1.4977085525067027,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8621212244033813,
      "num_tokens": 109838603.0,
      "step": 3156
    },
    {
      "epoch": 0.5862581244196843,
      "grad_norm": 1.3861234543685483,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8797081708908081,
      "num_tokens": 109879090.0,
      "step": 3157
    },
    {
      "epoch": 0.5864438254410399,
      "grad_norm": 1.4455248114784265,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8691490888595581,
      "num_tokens": 109914109.0,
      "step": 3158
    },
    {
      "epoch": 0.5866295264623955,
      "grad_norm": 1.376755875192373,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8666079044342041,
      "num_tokens": 109951285.0,
      "step": 3159
    },
    {
      "epoch": 0.5868152274837511,
      "grad_norm": 1.6437184094841064,
      "learning_rate": 1e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8535138964653015,
      "num_tokens": 109986942.0,
      "step": 3160
    },
    {
      "epoch": 0.5870009285051068,
      "grad_norm": 1.489834598838489,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8648629188537598,
      "num_tokens": 110021779.0,
      "step": 3161
    },
    {
      "epoch": 0.5871866295264624,
      "grad_norm": 1.446766141668018,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8563058972358704,
      "num_tokens": 110056832.0,
      "step": 3162
    },
    {
      "epoch": 0.587372330547818,
      "grad_norm": 1.5080909092296888,
      "learning_rate": 1e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8597005605697632,
      "num_tokens": 110089405.0,
      "step": 3163
    },
    {
      "epoch": 0.5875580315691736,
      "grad_norm": 1.5198128406335543,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8622610569000244,
      "num_tokens": 110123386.0,
      "step": 3164
    },
    {
      "epoch": 0.5877437325905293,
      "grad_norm": 1.3760711987731258,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8622143268585205,
      "num_tokens": 110163445.0,
      "step": 3165
    },
    {
      "epoch": 0.5879294336118849,
      "grad_norm": 1.5324681707078123,
      "learning_rate": 1e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8625544309616089,
      "num_tokens": 110196737.0,
      "step": 3166
    },
    {
      "epoch": 0.5881151346332405,
      "grad_norm": 1.526836136837001,
      "learning_rate": 1e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8583239912986755,
      "num_tokens": 110231631.0,
      "step": 3167
    },
    {
      "epoch": 0.5883008356545961,
      "grad_norm": 1.425168875275374,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8730844259262085,
      "num_tokens": 110269391.0,
      "step": 3168
    },
    {
      "epoch": 0.5884865366759517,
      "grad_norm": 1.535289392904888,
      "learning_rate": 1e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.860410749912262,
      "num_tokens": 110300232.0,
      "step": 3169
    },
    {
      "epoch": 0.5886722376973074,
      "grad_norm": 1.5078681890209082,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8629231452941895,
      "num_tokens": 110335248.0,
      "step": 3170
    },
    {
      "epoch": 0.588857938718663,
      "grad_norm": 1.4601497169324704,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.862990140914917,
      "num_tokens": 110371524.0,
      "step": 3171
    },
    {
      "epoch": 0.5890436397400186,
      "grad_norm": 1.6726018080783807,
      "learning_rate": 1e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8598343133926392,
      "num_tokens": 110402387.0,
      "step": 3172
    },
    {
      "epoch": 0.5892293407613742,
      "grad_norm": 1.6162861857118316,
      "learning_rate": 1e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8636471033096313,
      "num_tokens": 110432715.0,
      "step": 3173
    },
    {
      "epoch": 0.5894150417827299,
      "grad_norm": 1.4959430994599163,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8653457760810852,
      "num_tokens": 110471742.0,
      "step": 3174
    },
    {
      "epoch": 0.5896007428040855,
      "grad_norm": 1.3376736852909201,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8682523965835571,
      "num_tokens": 110516116.0,
      "step": 3175
    },
    {
      "epoch": 0.589786443825441,
      "grad_norm": 1.4549354970550374,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8672159910202026,
      "num_tokens": 110550264.0,
      "step": 3176
    },
    {
      "epoch": 0.5899721448467966,
      "grad_norm": 1.4955115351082082,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8642399311065674,
      "num_tokens": 110587488.0,
      "step": 3177
    },
    {
      "epoch": 0.5901578458681522,
      "grad_norm": 1.3317693229288798,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8733514547348022,
      "num_tokens": 110626123.0,
      "step": 3178
    },
    {
      "epoch": 0.5903435468895079,
      "grad_norm": 1.5993520008922497,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8729041814804077,
      "num_tokens": 110655264.0,
      "step": 3179
    },
    {
      "epoch": 0.5905292479108635,
      "grad_norm": 1.4294854986142511,
      "learning_rate": 1e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8490883111953735,
      "num_tokens": 110695717.0,
      "step": 3180
    },
    {
      "epoch": 0.5907149489322191,
      "grad_norm": 1.4762631405990378,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8696141242980957,
      "num_tokens": 110729351.0,
      "step": 3181
    },
    {
      "epoch": 0.5909006499535747,
      "grad_norm": 1.4812553837381315,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8769911527633667,
      "num_tokens": 110763052.0,
      "step": 3182
    },
    {
      "epoch": 0.5910863509749303,
      "grad_norm": 1.3728419156550213,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8716607093811035,
      "num_tokens": 110799473.0,
      "step": 3183
    },
    {
      "epoch": 0.591272051996286,
      "grad_norm": 1.5111723799404715,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8701024055480957,
      "num_tokens": 110832717.0,
      "step": 3184
    },
    {
      "epoch": 0.5914577530176416,
      "grad_norm": 1.5036219722095023,
      "learning_rate": 1e-06,
      "loss": 0.3871,
      "mean_token_accuracy": 0.8686761260032654,
      "num_tokens": 110867075.0,
      "step": 3185
    },
    {
      "epoch": 0.5916434540389972,
      "grad_norm": 1.6219229562422286,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.86009681224823,
      "num_tokens": 110894968.0,
      "step": 3186
    },
    {
      "epoch": 0.5918291550603528,
      "grad_norm": 1.634617424550415,
      "learning_rate": 1e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.85263592004776,
      "num_tokens": 110924420.0,
      "step": 3187
    },
    {
      "epoch": 0.5920148560817085,
      "grad_norm": 1.4510260885038972,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8631064891815186,
      "num_tokens": 110962864.0,
      "step": 3188
    },
    {
      "epoch": 0.5922005571030641,
      "grad_norm": 2.2981045629175862,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8589390516281128,
      "num_tokens": 110997162.0,
      "step": 3189
    },
    {
      "epoch": 0.5923862581244197,
      "grad_norm": 1.6117631426540757,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8697130680084229,
      "num_tokens": 111024915.0,
      "step": 3190
    },
    {
      "epoch": 0.5925719591457753,
      "grad_norm": 1.5593458710326487,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8593274354934692,
      "num_tokens": 111058254.0,
      "step": 3191
    },
    {
      "epoch": 0.5927576601671309,
      "grad_norm": 1.4498529518334424,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8634833097457886,
      "num_tokens": 111093144.0,
      "step": 3192
    },
    {
      "epoch": 0.5929433611884866,
      "grad_norm": 1.467392455280141,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8634018301963806,
      "num_tokens": 111129747.0,
      "step": 3193
    },
    {
      "epoch": 0.5931290622098422,
      "grad_norm": 1.5487535275118869,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8571038842201233,
      "num_tokens": 111164922.0,
      "step": 3194
    },
    {
      "epoch": 0.5933147632311978,
      "grad_norm": 1.5301811469068658,
      "learning_rate": 1e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8517298698425293,
      "num_tokens": 111199808.0,
      "step": 3195
    },
    {
      "epoch": 0.5935004642525534,
      "grad_norm": 1.528435524471905,
      "learning_rate": 1e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8482000231742859,
      "num_tokens": 111234997.0,
      "step": 3196
    },
    {
      "epoch": 0.593686165273909,
      "grad_norm": 1.571915494826322,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8678958415985107,
      "num_tokens": 111264672.0,
      "step": 3197
    },
    {
      "epoch": 0.5938718662952647,
      "grad_norm": 1.6560270683194704,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8652592897415161,
      "num_tokens": 111293627.0,
      "step": 3198
    },
    {
      "epoch": 0.5940575673166203,
      "grad_norm": 1.4966900964021794,
      "learning_rate": 1e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8642292022705078,
      "num_tokens": 111325936.0,
      "step": 3199
    },
    {
      "epoch": 0.5942432683379758,
      "grad_norm": 1.531223785442496,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8616056442260742,
      "num_tokens": 111364173.0,
      "step": 3200
    },
    {
      "epoch": 0.5944289693593314,
      "grad_norm": 1.431821485763386,
      "learning_rate": 1e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8573936223983765,
      "num_tokens": 111401868.0,
      "step": 3201
    },
    {
      "epoch": 0.594614670380687,
      "grad_norm": 1.4364677852690462,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8597654104232788,
      "num_tokens": 111438151.0,
      "step": 3202
    },
    {
      "epoch": 0.5948003714020427,
      "grad_norm": 1.5308001041902288,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8698970079421997,
      "num_tokens": 111468111.0,
      "step": 3203
    },
    {
      "epoch": 0.5949860724233983,
      "grad_norm": 1.4307271130880514,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8730687499046326,
      "num_tokens": 111499124.0,
      "step": 3204
    },
    {
      "epoch": 0.5951717734447539,
      "grad_norm": 1.435158922163642,
      "learning_rate": 1e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.860627293586731,
      "num_tokens": 111538900.0,
      "step": 3205
    },
    {
      "epoch": 0.5953574744661095,
      "grad_norm": 1.5014465847365839,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8712793588638306,
      "num_tokens": 111569668.0,
      "step": 3206
    },
    {
      "epoch": 0.5955431754874652,
      "grad_norm": 1.684309656101671,
      "learning_rate": 1e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8576995134353638,
      "num_tokens": 111601194.0,
      "step": 3207
    },
    {
      "epoch": 0.5957288765088208,
      "grad_norm": 1.5894905616656132,
      "learning_rate": 1e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8390839099884033,
      "num_tokens": 111632663.0,
      "step": 3208
    },
    {
      "epoch": 0.5959145775301764,
      "grad_norm": 1.7169470397638513,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8551540970802307,
      "num_tokens": 111659450.0,
      "step": 3209
    },
    {
      "epoch": 0.596100278551532,
      "grad_norm": 1.5187451933929603,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8774075508117676,
      "num_tokens": 111689769.0,
      "step": 3210
    },
    {
      "epoch": 0.5962859795728876,
      "grad_norm": 1.5718447452463093,
      "learning_rate": 1e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8618758320808411,
      "num_tokens": 111720294.0,
      "step": 3211
    },
    {
      "epoch": 0.5964716805942433,
      "grad_norm": 1.4991315650123291,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.871417224407196,
      "num_tokens": 111752594.0,
      "step": 3212
    },
    {
      "epoch": 0.5966573816155989,
      "grad_norm": 1.500882939753187,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8703514933586121,
      "num_tokens": 111787407.0,
      "step": 3213
    },
    {
      "epoch": 0.5968430826369545,
      "grad_norm": 1.6006464510742489,
      "learning_rate": 1e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.859952986240387,
      "num_tokens": 111816767.0,
      "step": 3214
    },
    {
      "epoch": 0.5970287836583101,
      "grad_norm": 1.5048364139703527,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8660593032836914,
      "num_tokens": 111849529.0,
      "step": 3215
    },
    {
      "epoch": 0.5972144846796658,
      "grad_norm": 1.5234753650477908,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8747245073318481,
      "num_tokens": 111881930.0,
      "step": 3216
    },
    {
      "epoch": 0.5974001857010214,
      "grad_norm": 1.4509907507759117,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.857361912727356,
      "num_tokens": 111922057.0,
      "step": 3217
    },
    {
      "epoch": 0.597585886722377,
      "grad_norm": 1.5563890638911557,
      "learning_rate": 1e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8644832372665405,
      "num_tokens": 111960121.0,
      "step": 3218
    },
    {
      "epoch": 0.5977715877437326,
      "grad_norm": 1.4942319165617042,
      "learning_rate": 1e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8611010313034058,
      "num_tokens": 111996714.0,
      "step": 3219
    },
    {
      "epoch": 0.5979572887650882,
      "grad_norm": 1.6166727768897493,
      "learning_rate": 1e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8551405668258667,
      "num_tokens": 112031036.0,
      "step": 3220
    },
    {
      "epoch": 0.5981429897864439,
      "grad_norm": 1.5810953665418972,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.866270899772644,
      "num_tokens": 112063144.0,
      "step": 3221
    },
    {
      "epoch": 0.5983286908077995,
      "grad_norm": 1.602059081035959,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8656913638114929,
      "num_tokens": 112092997.0,
      "step": 3222
    },
    {
      "epoch": 0.5985143918291551,
      "grad_norm": 1.5552150238277365,
      "learning_rate": 1e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8527178168296814,
      "num_tokens": 112130696.0,
      "step": 3223
    },
    {
      "epoch": 0.5987000928505106,
      "grad_norm": 1.4586732455113856,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8548520803451538,
      "num_tokens": 112171816.0,
      "step": 3224
    },
    {
      "epoch": 0.5988857938718662,
      "grad_norm": 1.6073463045840362,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8575257062911987,
      "num_tokens": 112204253.0,
      "step": 3225
    },
    {
      "epoch": 0.5990714948932219,
      "grad_norm": 1.655304277266917,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8635554909706116,
      "num_tokens": 112234541.0,
      "step": 3226
    },
    {
      "epoch": 0.5992571959145775,
      "grad_norm": 1.4107366044666714,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8616483211517334,
      "num_tokens": 112273266.0,
      "step": 3227
    },
    {
      "epoch": 0.5994428969359331,
      "grad_norm": 1.4728202641342634,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8652702569961548,
      "num_tokens": 112312987.0,
      "step": 3228
    },
    {
      "epoch": 0.5996285979572887,
      "grad_norm": 1.5188750022717379,
      "learning_rate": 1e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8551198840141296,
      "num_tokens": 112344699.0,
      "step": 3229
    },
    {
      "epoch": 0.5998142989786444,
      "grad_norm": 1.4819311092905412,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8635032176971436,
      "num_tokens": 112379608.0,
      "step": 3230
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4477937098480547,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8612732887268066,
      "num_tokens": 112417934.0,
      "step": 3231
    },
    {
      "epoch": 0.6001857010213556,
      "grad_norm": 1.4085420922025103,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8704894185066223,
      "num_tokens": 112453914.0,
      "step": 3232
    },
    {
      "epoch": 0.6003714020427112,
      "grad_norm": 1.5508910415064496,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8606537580490112,
      "num_tokens": 112486120.0,
      "step": 3233
    },
    {
      "epoch": 0.6005571030640668,
      "grad_norm": 1.575434540459157,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8643361330032349,
      "num_tokens": 112519789.0,
      "step": 3234
    },
    {
      "epoch": 0.6007428040854225,
      "grad_norm": 1.6014872289981334,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8773829936981201,
      "num_tokens": 112551073.0,
      "step": 3235
    },
    {
      "epoch": 0.6009285051067781,
      "grad_norm": 1.6763941161633455,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8622063398361206,
      "num_tokens": 112582301.0,
      "step": 3236
    },
    {
      "epoch": 0.6011142061281337,
      "grad_norm": 1.5484360584867676,
      "learning_rate": 1e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8527728319168091,
      "num_tokens": 112619306.0,
      "step": 3237
    },
    {
      "epoch": 0.6012999071494893,
      "grad_norm": 1.49858505048832,
      "learning_rate": 1e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8596702814102173,
      "num_tokens": 112654707.0,
      "step": 3238
    },
    {
      "epoch": 0.601485608170845,
      "grad_norm": 1.6485948205001613,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8727983236312866,
      "num_tokens": 112681218.0,
      "step": 3239
    },
    {
      "epoch": 0.6016713091922006,
      "grad_norm": 1.5394475869321134,
      "learning_rate": 1e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8419519662857056,
      "num_tokens": 112718611.0,
      "step": 3240
    },
    {
      "epoch": 0.6018570102135562,
      "grad_norm": 1.4231141690995601,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8747241497039795,
      "num_tokens": 112757108.0,
      "step": 3241
    },
    {
      "epoch": 0.6020427112349118,
      "grad_norm": 1.5533423350668925,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8712603449821472,
      "num_tokens": 112791853.0,
      "step": 3242
    },
    {
      "epoch": 0.6022284122562674,
      "grad_norm": 1.6540026037644509,
      "learning_rate": 1e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8452908396720886,
      "num_tokens": 112824801.0,
      "step": 3243
    },
    {
      "epoch": 0.6024141132776231,
      "grad_norm": 1.4983555383809752,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8725860714912415,
      "num_tokens": 112856858.0,
      "step": 3244
    },
    {
      "epoch": 0.6025998142989787,
      "grad_norm": 1.5482073725597274,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8623684644699097,
      "num_tokens": 112889845.0,
      "step": 3245
    },
    {
      "epoch": 0.6027855153203343,
      "grad_norm": 1.5662167791600945,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8633409738540649,
      "num_tokens": 112922684.0,
      "step": 3246
    },
    {
      "epoch": 0.6029712163416899,
      "grad_norm": 1.4894366941510804,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8724461793899536,
      "num_tokens": 112954621.0,
      "step": 3247
    },
    {
      "epoch": 0.6031569173630454,
      "grad_norm": 1.4914046703874875,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8675975799560547,
      "num_tokens": 112988245.0,
      "step": 3248
    },
    {
      "epoch": 0.6033426183844011,
      "grad_norm": 1.5218021983030723,
      "learning_rate": 1e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8623977303504944,
      "num_tokens": 113019701.0,
      "step": 3249
    },
    {
      "epoch": 0.6035283194057567,
      "grad_norm": 1.515152972972093,
      "learning_rate": 1e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.840114951133728,
      "num_tokens": 113059955.0,
      "step": 3250
    },
    {
      "epoch": 0.6037140204271123,
      "grad_norm": 1.592366906748567,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8680105805397034,
      "num_tokens": 113092692.0,
      "step": 3251
    },
    {
      "epoch": 0.6038997214484679,
      "grad_norm": 1.3414792466157373,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8743754625320435,
      "num_tokens": 113131051.0,
      "step": 3252
    },
    {
      "epoch": 0.6040854224698236,
      "grad_norm": 1.4559820024806098,
      "learning_rate": 1e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8464387655258179,
      "num_tokens": 113170649.0,
      "step": 3253
    },
    {
      "epoch": 0.6042711234911792,
      "grad_norm": 1.5613495187500135,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8676779270172119,
      "num_tokens": 113202014.0,
      "step": 3254
    },
    {
      "epoch": 0.6044568245125348,
      "grad_norm": 1.4249604536252036,
      "learning_rate": 1e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8586670160293579,
      "num_tokens": 113244458.0,
      "step": 3255
    },
    {
      "epoch": 0.6046425255338904,
      "grad_norm": 1.4716834391298048,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8681355714797974,
      "num_tokens": 113282161.0,
      "step": 3256
    },
    {
      "epoch": 0.604828226555246,
      "grad_norm": 1.5595707959723517,
      "learning_rate": 1e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8537721633911133,
      "num_tokens": 113317086.0,
      "step": 3257
    },
    {
      "epoch": 0.6050139275766017,
      "grad_norm": 1.517248252653416,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8690658807754517,
      "num_tokens": 113349937.0,
      "step": 3258
    },
    {
      "epoch": 0.6051996285979573,
      "grad_norm": 1.5722331139396408,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8650419116020203,
      "num_tokens": 113384901.0,
      "step": 3259
    },
    {
      "epoch": 0.6053853296193129,
      "grad_norm": 1.4899458645273274,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8609373569488525,
      "num_tokens": 113420612.0,
      "step": 3260
    },
    {
      "epoch": 0.6055710306406685,
      "grad_norm": 1.5227095959780743,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.874920129776001,
      "num_tokens": 113455459.0,
      "step": 3261
    },
    {
      "epoch": 0.6057567316620242,
      "grad_norm": 1.4016679994189176,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8774614930152893,
      "num_tokens": 113493794.0,
      "step": 3262
    },
    {
      "epoch": 0.6059424326833798,
      "grad_norm": 1.6331840551954817,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.862491250038147,
      "num_tokens": 113521100.0,
      "step": 3263
    },
    {
      "epoch": 0.6061281337047354,
      "grad_norm": 1.4914830221230038,
      "learning_rate": 1e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.861824631690979,
      "num_tokens": 113553825.0,
      "step": 3264
    },
    {
      "epoch": 0.606313834726091,
      "grad_norm": 1.4643741315995302,
      "learning_rate": 1e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8622441291809082,
      "num_tokens": 113588338.0,
      "step": 3265
    },
    {
      "epoch": 0.6064995357474466,
      "grad_norm": 1.5823830897356457,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8627310991287231,
      "num_tokens": 113621962.0,
      "step": 3266
    },
    {
      "epoch": 0.6066852367688023,
      "grad_norm": 1.495726606131881,
      "learning_rate": 1e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8687117099761963,
      "num_tokens": 113653985.0,
      "step": 3267
    },
    {
      "epoch": 0.6068709377901579,
      "grad_norm": 1.4576291968285893,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8614126443862915,
      "num_tokens": 113694416.0,
      "step": 3268
    },
    {
      "epoch": 0.6070566388115135,
      "grad_norm": 1.3104834350921357,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8601369857788086,
      "num_tokens": 113733524.0,
      "step": 3269
    },
    {
      "epoch": 0.6072423398328691,
      "grad_norm": 1.4841549952903632,
      "learning_rate": 1e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8499757051467896,
      "num_tokens": 113774302.0,
      "step": 3270
    },
    {
      "epoch": 0.6074280408542247,
      "grad_norm": 1.6400156316070063,
      "learning_rate": 1e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8571639657020569,
      "num_tokens": 113805731.0,
      "step": 3271
    },
    {
      "epoch": 0.6076137418755804,
      "grad_norm": 1.4691424893042153,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8699119687080383,
      "num_tokens": 113839711.0,
      "step": 3272
    },
    {
      "epoch": 0.6077994428969359,
      "grad_norm": 1.5957146644750946,
      "learning_rate": 1e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.850541353225708,
      "num_tokens": 113873493.0,
      "step": 3273
    },
    {
      "epoch": 0.6079851439182915,
      "grad_norm": 1.4926626329080817,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8650929927825928,
      "num_tokens": 113906955.0,
      "step": 3274
    },
    {
      "epoch": 0.6081708449396471,
      "grad_norm": 1.4270501494297902,
      "learning_rate": 1e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8636868000030518,
      "num_tokens": 113942428.0,
      "step": 3275
    },
    {
      "epoch": 0.6083565459610027,
      "grad_norm": 1.3753506128556163,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.878021240234375,
      "num_tokens": 113981714.0,
      "step": 3276
    },
    {
      "epoch": 0.6085422469823584,
      "grad_norm": 1.3775254197253768,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.87722247838974,
      "num_tokens": 114019045.0,
      "step": 3277
    },
    {
      "epoch": 0.608727948003714,
      "grad_norm": 1.4777852694090405,
      "learning_rate": 1e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8575628995895386,
      "num_tokens": 114059952.0,
      "step": 3278
    },
    {
      "epoch": 0.6089136490250696,
      "grad_norm": 1.3937665388806733,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8654392957687378,
      "num_tokens": 114099742.0,
      "step": 3279
    },
    {
      "epoch": 0.6090993500464252,
      "grad_norm": 1.6844818755890687,
      "learning_rate": 1e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8400914669036865,
      "num_tokens": 114129219.0,
      "step": 3280
    },
    {
      "epoch": 0.6092850510677809,
      "grad_norm": 1.49138044242712,
      "learning_rate": 1e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.831612229347229,
      "num_tokens": 114166634.0,
      "step": 3281
    },
    {
      "epoch": 0.6094707520891365,
      "grad_norm": 1.5063920686281156,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8675410747528076,
      "num_tokens": 114202387.0,
      "step": 3282
    },
    {
      "epoch": 0.6096564531104921,
      "grad_norm": 1.4252869220229962,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8685557246208191,
      "num_tokens": 114238052.0,
      "step": 3283
    },
    {
      "epoch": 0.6098421541318477,
      "grad_norm": 1.3529783464295302,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8693877458572388,
      "num_tokens": 114277135.0,
      "step": 3284
    },
    {
      "epoch": 0.6100278551532033,
      "grad_norm": 1.44563705344308,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8672024011611938,
      "num_tokens": 114309307.0,
      "step": 3285
    },
    {
      "epoch": 0.610213556174559,
      "grad_norm": 1.6092244774100706,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8766986131668091,
      "num_tokens": 114337795.0,
      "step": 3286
    },
    {
      "epoch": 0.6103992571959146,
      "grad_norm": 1.4247825296312107,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8627557754516602,
      "num_tokens": 114380178.0,
      "step": 3287
    },
    {
      "epoch": 0.6105849582172702,
      "grad_norm": 1.59242374779113,
      "learning_rate": 1e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.858482837677002,
      "num_tokens": 114413630.0,
      "step": 3288
    },
    {
      "epoch": 0.6107706592386258,
      "grad_norm": 1.569354026753245,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8568115234375,
      "num_tokens": 114446654.0,
      "step": 3289
    },
    {
      "epoch": 0.6109563602599815,
      "grad_norm": 1.4708911661489057,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8660264015197754,
      "num_tokens": 114480671.0,
      "step": 3290
    },
    {
      "epoch": 0.6111420612813371,
      "grad_norm": 1.6403459645396175,
      "learning_rate": 1e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8437505960464478,
      "num_tokens": 114509265.0,
      "step": 3291
    },
    {
      "epoch": 0.6113277623026927,
      "grad_norm": 1.4472193113331493,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8675936460494995,
      "num_tokens": 114547014.0,
      "step": 3292
    },
    {
      "epoch": 0.6115134633240483,
      "grad_norm": 1.4730121793080653,
      "learning_rate": 1e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.853983461856842,
      "num_tokens": 114582240.0,
      "step": 3293
    },
    {
      "epoch": 0.611699164345404,
      "grad_norm": 1.556406457549085,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8712372779846191,
      "num_tokens": 114611357.0,
      "step": 3294
    },
    {
      "epoch": 0.6118848653667596,
      "grad_norm": 1.6095994002993737,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8650655150413513,
      "num_tokens": 114641609.0,
      "step": 3295
    },
    {
      "epoch": 0.6120705663881152,
      "grad_norm": 1.635323116679454,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8666624426841736,
      "num_tokens": 114671036.0,
      "step": 3296
    },
    {
      "epoch": 0.6122562674094707,
      "grad_norm": 1.5677613433365813,
      "learning_rate": 1e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8559199571609497,
      "num_tokens": 114704479.0,
      "step": 3297
    },
    {
      "epoch": 0.6124419684308263,
      "grad_norm": 1.4493732566755066,
      "learning_rate": 1e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.868316113948822,
      "num_tokens": 114742051.0,
      "step": 3298
    },
    {
      "epoch": 0.6126276694521819,
      "grad_norm": 1.5512679171640653,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8575710654258728,
      "num_tokens": 114779490.0,
      "step": 3299
    },
    {
      "epoch": 0.6128133704735376,
      "grad_norm": 1.4077803902026909,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8661696314811707,
      "num_tokens": 114817776.0,
      "step": 3300
    },
    {
      "epoch": 0.6129990714948932,
      "grad_norm": 1.4795325727214528,
      "learning_rate": 1e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8605707287788391,
      "num_tokens": 114855801.0,
      "step": 3301
    },
    {
      "epoch": 0.6131847725162488,
      "grad_norm": 1.6082751564476099,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8676825165748596,
      "num_tokens": 114883235.0,
      "step": 3302
    },
    {
      "epoch": 0.6133704735376044,
      "grad_norm": 1.3637548519760574,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8686528205871582,
      "num_tokens": 114922642.0,
      "step": 3303
    },
    {
      "epoch": 0.61355617455896,
      "grad_norm": 1.450705557830267,
      "learning_rate": 1e-06,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8481088280677795,
      "num_tokens": 114959862.0,
      "step": 3304
    },
    {
      "epoch": 0.6137418755803157,
      "grad_norm": 1.4643284836082955,
      "learning_rate": 1e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8583032488822937,
      "num_tokens": 114998503.0,
      "step": 3305
    },
    {
      "epoch": 0.6139275766016713,
      "grad_norm": 1.3882318333613304,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8732246160507202,
      "num_tokens": 115036301.0,
      "step": 3306
    },
    {
      "epoch": 0.6141132776230269,
      "grad_norm": 1.5111974387239335,
      "learning_rate": 1e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.859342098236084,
      "num_tokens": 115074940.0,
      "step": 3307
    },
    {
      "epoch": 0.6142989786443825,
      "grad_norm": 1.5260881111609148,
      "learning_rate": 1e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8631753921508789,
      "num_tokens": 115107808.0,
      "step": 3308
    },
    {
      "epoch": 0.6144846796657382,
      "grad_norm": 1.5333409397998614,
      "learning_rate": 1e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8591836094856262,
      "num_tokens": 115143031.0,
      "step": 3309
    },
    {
      "epoch": 0.6146703806870938,
      "grad_norm": 1.4831470341237158,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8760911226272583,
      "num_tokens": 115176852.0,
      "step": 3310
    },
    {
      "epoch": 0.6148560817084494,
      "grad_norm": 1.5619422060798505,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8741381168365479,
      "num_tokens": 115208089.0,
      "step": 3311
    },
    {
      "epoch": 0.615041782729805,
      "grad_norm": 1.6424186025105798,
      "learning_rate": 1e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8519555330276489,
      "num_tokens": 115238761.0,
      "step": 3312
    },
    {
      "epoch": 0.6152274837511607,
      "grad_norm": 1.4650972506401245,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8644295334815979,
      "num_tokens": 115272456.0,
      "step": 3313
    },
    {
      "epoch": 0.6154131847725163,
      "grad_norm": 1.4889769327858529,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8621464371681213,
      "num_tokens": 115307958.0,
      "step": 3314
    },
    {
      "epoch": 0.6155988857938719,
      "grad_norm": 1.567896341928522,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.877633810043335,
      "num_tokens": 115338539.0,
      "step": 3315
    },
    {
      "epoch": 0.6157845868152275,
      "grad_norm": 1.5704031399880312,
      "learning_rate": 1e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8581876754760742,
      "num_tokens": 115371189.0,
      "step": 3316
    },
    {
      "epoch": 0.6159702878365831,
      "grad_norm": 1.4934603604903645,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8673352599143982,
      "num_tokens": 115403825.0,
      "step": 3317
    },
    {
      "epoch": 0.6161559888579388,
      "grad_norm": 1.3792742709389239,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.872222900390625,
      "num_tokens": 115441675.0,
      "step": 3318
    },
    {
      "epoch": 0.6163416898792944,
      "grad_norm": 1.4886134455054525,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8561038970947266,
      "num_tokens": 115474083.0,
      "step": 3319
    },
    {
      "epoch": 0.61652739090065,
      "grad_norm": 1.515215480014164,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8593926429748535,
      "num_tokens": 115511530.0,
      "step": 3320
    },
    {
      "epoch": 0.6167130919220055,
      "grad_norm": 1.343228143650648,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8731467127799988,
      "num_tokens": 115549529.0,
      "step": 3321
    },
    {
      "epoch": 0.6168987929433611,
      "grad_norm": 1.526487714606072,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8619060516357422,
      "num_tokens": 115585860.0,
      "step": 3322
    },
    {
      "epoch": 0.6170844939647168,
      "grad_norm": 1.5495730237233838,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8614351153373718,
      "num_tokens": 115620952.0,
      "step": 3323
    },
    {
      "epoch": 0.6172701949860724,
      "grad_norm": 1.5441466396063548,
      "learning_rate": 1e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8481582403182983,
      "num_tokens": 115657687.0,
      "step": 3324
    },
    {
      "epoch": 0.617455896007428,
      "grad_norm": 1.607209933102543,
      "learning_rate": 1e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8499637842178345,
      "num_tokens": 115689832.0,
      "step": 3325
    },
    {
      "epoch": 0.6176415970287836,
      "grad_norm": 1.655083964901644,
      "learning_rate": 1e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8573668599128723,
      "num_tokens": 115718981.0,
      "step": 3326
    },
    {
      "epoch": 0.6178272980501393,
      "grad_norm": 1.5112167510339123,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8606569170951843,
      "num_tokens": 115753741.0,
      "step": 3327
    },
    {
      "epoch": 0.6180129990714949,
      "grad_norm": 1.4894004037402064,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8570639491081238,
      "num_tokens": 115791601.0,
      "step": 3328
    },
    {
      "epoch": 0.6181987000928505,
      "grad_norm": 1.475552434474608,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8756366968154907,
      "num_tokens": 115824274.0,
      "step": 3329
    },
    {
      "epoch": 0.6183844011142061,
      "grad_norm": 1.5599788116913058,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8603702783584595,
      "num_tokens": 115858515.0,
      "step": 3330
    },
    {
      "epoch": 0.6185701021355617,
      "grad_norm": 1.567921494080326,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8625651597976685,
      "num_tokens": 115890190.0,
      "step": 3331
    },
    {
      "epoch": 0.6187558031569174,
      "grad_norm": 1.4303035915715023,
      "learning_rate": 1e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8548635244369507,
      "num_tokens": 115928868.0,
      "step": 3332
    },
    {
      "epoch": 0.618941504178273,
      "grad_norm": 1.4158862381714605,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8596609830856323,
      "num_tokens": 115965330.0,
      "step": 3333
    },
    {
      "epoch": 0.6191272051996286,
      "grad_norm": 1.6909187372159253,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8651142716407776,
      "num_tokens": 115993480.0,
      "step": 3334
    },
    {
      "epoch": 0.6193129062209842,
      "grad_norm": 1.6415537559136695,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.880946159362793,
      "num_tokens": 116026263.0,
      "step": 3335
    },
    {
      "epoch": 0.6194986072423398,
      "grad_norm": 1.5500921586134708,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8786660432815552,
      "num_tokens": 116056671.0,
      "step": 3336
    },
    {
      "epoch": 0.6196843082636955,
      "grad_norm": 1.5811983593958148,
      "learning_rate": 1e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8489576578140259,
      "num_tokens": 116089188.0,
      "step": 3337
    },
    {
      "epoch": 0.6198700092850511,
      "grad_norm": 1.5217587541721176,
      "learning_rate": 1e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8541231155395508,
      "num_tokens": 116126596.0,
      "step": 3338
    },
    {
      "epoch": 0.6200557103064067,
      "grad_norm": 1.5020634649205773,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8596333861351013,
      "num_tokens": 116161056.0,
      "step": 3339
    },
    {
      "epoch": 0.6202414113277623,
      "grad_norm": 1.6761558188944774,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8615416288375854,
      "num_tokens": 116188727.0,
      "step": 3340
    },
    {
      "epoch": 0.620427112349118,
      "grad_norm": 1.6566639706323842,
      "learning_rate": 1e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8415789604187012,
      "num_tokens": 116224612.0,
      "step": 3341
    },
    {
      "epoch": 0.6206128133704736,
      "grad_norm": 1.535008735414603,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8669413328170776,
      "num_tokens": 116259633.0,
      "step": 3342
    },
    {
      "epoch": 0.6207985143918292,
      "grad_norm": 1.6635385535061833,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8619179129600525,
      "num_tokens": 116285929.0,
      "step": 3343
    },
    {
      "epoch": 0.6209842154131848,
      "grad_norm": 1.527430838198438,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8608516454696655,
      "num_tokens": 116320827.0,
      "step": 3344
    },
    {
      "epoch": 0.6211699164345403,
      "grad_norm": 1.3531223052833872,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8686016798019409,
      "num_tokens": 116361440.0,
      "step": 3345
    },
    {
      "epoch": 0.621355617455896,
      "grad_norm": 1.4964772459966482,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8654820919036865,
      "num_tokens": 116395854.0,
      "step": 3346
    },
    {
      "epoch": 0.6215413184772516,
      "grad_norm": 1.3345960930849716,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.866392970085144,
      "num_tokens": 116436529.0,
      "step": 3347
    },
    {
      "epoch": 0.6217270194986072,
      "grad_norm": 1.5384751563359311,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.881318211555481,
      "num_tokens": 116467950.0,
      "step": 3348
    },
    {
      "epoch": 0.6219127205199628,
      "grad_norm": 1.3973874651294096,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8834506273269653,
      "num_tokens": 116505844.0,
      "step": 3349
    },
    {
      "epoch": 0.6220984215413184,
      "grad_norm": 1.3381784564015138,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8718311190605164,
      "num_tokens": 116544622.0,
      "step": 3350
    },
    {
      "epoch": 0.6222841225626741,
      "grad_norm": 1.5193629580836927,
      "learning_rate": 1e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8548656702041626,
      "num_tokens": 116579405.0,
      "step": 3351
    },
    {
      "epoch": 0.6224698235840297,
      "grad_norm": 1.5912893891131645,
      "learning_rate": 1e-06,
      "loss": 0.3866,
      "mean_token_accuracy": 0.8688538670539856,
      "num_tokens": 116610714.0,
      "step": 3352
    },
    {
      "epoch": 0.6226555246053853,
      "grad_norm": 1.419664244014176,
      "learning_rate": 1e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8675456047058105,
      "num_tokens": 116647101.0,
      "step": 3353
    },
    {
      "epoch": 0.6228412256267409,
      "grad_norm": 1.5655366423725565,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.864105224609375,
      "num_tokens": 116679022.0,
      "step": 3354
    },
    {
      "epoch": 0.6230269266480966,
      "grad_norm": 1.5048009616453841,
      "learning_rate": 1e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8526002764701843,
      "num_tokens": 116719301.0,
      "step": 3355
    },
    {
      "epoch": 0.6232126276694522,
      "grad_norm": 1.411241761509794,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8758693933486938,
      "num_tokens": 116757667.0,
      "step": 3356
    },
    {
      "epoch": 0.6233983286908078,
      "grad_norm": 1.4465396321047401,
      "learning_rate": 1e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8599629402160645,
      "num_tokens": 116797428.0,
      "step": 3357
    },
    {
      "epoch": 0.6235840297121634,
      "grad_norm": 1.544137891075209,
      "learning_rate": 1e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8426976203918457,
      "num_tokens": 116835555.0,
      "step": 3358
    },
    {
      "epoch": 0.623769730733519,
      "grad_norm": 1.4637878105580817,
      "learning_rate": 1e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8554924726486206,
      "num_tokens": 116874020.0,
      "step": 3359
    },
    {
      "epoch": 0.6239554317548747,
      "grad_norm": 1.4063452463583017,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8732208013534546,
      "num_tokens": 116905024.0,
      "step": 3360
    },
    {
      "epoch": 0.6241411327762303,
      "grad_norm": 1.482293324754893,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8686299324035645,
      "num_tokens": 116941237.0,
      "step": 3361
    },
    {
      "epoch": 0.6243268337975859,
      "grad_norm": 1.6241539340472102,
      "learning_rate": 1e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8612427711486816,
      "num_tokens": 116968712.0,
      "step": 3362
    },
    {
      "epoch": 0.6245125348189415,
      "grad_norm": 1.5502649297638853,
      "learning_rate": 1e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8562389612197876,
      "num_tokens": 117006606.0,
      "step": 3363
    },
    {
      "epoch": 0.6246982358402972,
      "grad_norm": 1.509498178409976,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8672233819961548,
      "num_tokens": 117041772.0,
      "step": 3364
    },
    {
      "epoch": 0.6248839368616528,
      "grad_norm": 1.3547840930577884,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8638077974319458,
      "num_tokens": 117084053.0,
      "step": 3365
    },
    {
      "epoch": 0.6250696378830084,
      "grad_norm": 1.4085701576910834,
      "learning_rate": 1e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8560152053833008,
      "num_tokens": 117129041.0,
      "step": 3366
    },
    {
      "epoch": 0.625255338904364,
      "grad_norm": 1.672710535241587,
      "learning_rate": 1e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8522293567657471,
      "num_tokens": 117158019.0,
      "step": 3367
    },
    {
      "epoch": 0.6254410399257196,
      "grad_norm": 1.5189544604394987,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8673353791236877,
      "num_tokens": 117190633.0,
      "step": 3368
    },
    {
      "epoch": 0.6256267409470752,
      "grad_norm": 1.4623346886480184,
      "learning_rate": 1e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8559662699699402,
      "num_tokens": 117230166.0,
      "step": 3369
    },
    {
      "epoch": 0.6258124419684308,
      "grad_norm": 1.5230240229389849,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8705854415893555,
      "num_tokens": 117268765.0,
      "step": 3370
    },
    {
      "epoch": 0.6259981429897864,
      "grad_norm": 1.5917610309334365,
      "learning_rate": 1e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8621076345443726,
      "num_tokens": 117301310.0,
      "step": 3371
    },
    {
      "epoch": 0.626183844011142,
      "grad_norm": 1.4554435838273942,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8755462765693665,
      "num_tokens": 117336314.0,
      "step": 3372
    },
    {
      "epoch": 0.6263695450324976,
      "grad_norm": 1.5584222822996097,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8607724905014038,
      "num_tokens": 117371335.0,
      "step": 3373
    },
    {
      "epoch": 0.6265552460538533,
      "grad_norm": 1.3938125992040984,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8742941617965698,
      "num_tokens": 117409850.0,
      "step": 3374
    },
    {
      "epoch": 0.6267409470752089,
      "grad_norm": 1.495363788767012,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8684356212615967,
      "num_tokens": 117443877.0,
      "step": 3375
    },
    {
      "epoch": 0.6269266480965645,
      "grad_norm": 1.3730172622169259,
      "learning_rate": 1e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8636745810508728,
      "num_tokens": 117483825.0,
      "step": 3376
    },
    {
      "epoch": 0.6271123491179201,
      "grad_norm": 1.4571808341821615,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8780658841133118,
      "num_tokens": 117519943.0,
      "step": 3377
    },
    {
      "epoch": 0.6272980501392758,
      "grad_norm": 1.4997683943408877,
      "learning_rate": 1e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8488345146179199,
      "num_tokens": 117555277.0,
      "step": 3378
    },
    {
      "epoch": 0.6274837511606314,
      "grad_norm": 1.6356520190761734,
      "learning_rate": 1e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8564862012863159,
      "num_tokens": 117588482.0,
      "step": 3379
    },
    {
      "epoch": 0.627669452181987,
      "grad_norm": 1.4227415934970402,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8790549039840698,
      "num_tokens": 117626426.0,
      "step": 3380
    },
    {
      "epoch": 0.6278551532033426,
      "grad_norm": 1.6689164741176559,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8748733401298523,
      "num_tokens": 117657474.0,
      "step": 3381
    },
    {
      "epoch": 0.6280408542246982,
      "grad_norm": 1.4940493257384633,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8733541369438171,
      "num_tokens": 117691020.0,
      "step": 3382
    },
    {
      "epoch": 0.6282265552460539,
      "grad_norm": 1.5620179794932416,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8650170564651489,
      "num_tokens": 117722164.0,
      "step": 3383
    },
    {
      "epoch": 0.6284122562674095,
      "grad_norm": 1.6152679080059875,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8684728741645813,
      "num_tokens": 117754086.0,
      "step": 3384
    },
    {
      "epoch": 0.6285979572887651,
      "grad_norm": 1.405164286423374,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8671422004699707,
      "num_tokens": 117790890.0,
      "step": 3385
    },
    {
      "epoch": 0.6287836583101207,
      "grad_norm": 1.7866754045715876,
      "learning_rate": 1e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8571735620498657,
      "num_tokens": 117821754.0,
      "step": 3386
    },
    {
      "epoch": 0.6289693593314764,
      "grad_norm": 1.5717471562354346,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8594222068786621,
      "num_tokens": 117856018.0,
      "step": 3387
    },
    {
      "epoch": 0.629155060352832,
      "grad_norm": 1.424605747846141,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.875886857509613,
      "num_tokens": 117893302.0,
      "step": 3388
    },
    {
      "epoch": 0.6293407613741876,
      "grad_norm": 1.5639874625638215,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8726677894592285,
      "num_tokens": 117924015.0,
      "step": 3389
    },
    {
      "epoch": 0.6295264623955432,
      "grad_norm": 1.4697766869331974,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8722304701805115,
      "num_tokens": 117959464.0,
      "step": 3390
    },
    {
      "epoch": 0.6297121634168988,
      "grad_norm": 1.6263969881882223,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8628765344619751,
      "num_tokens": 117991912.0,
      "step": 3391
    },
    {
      "epoch": 0.6298978644382545,
      "grad_norm": 1.5272331521634497,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8760004043579102,
      "num_tokens": 118024984.0,
      "step": 3392
    },
    {
      "epoch": 0.63008356545961,
      "grad_norm": 1.7736734720274445,
      "learning_rate": 1e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8560670614242554,
      "num_tokens": 118050817.0,
      "step": 3393
    },
    {
      "epoch": 0.6302692664809656,
      "grad_norm": 1.3415150222102563,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.867682158946991,
      "num_tokens": 118095253.0,
      "step": 3394
    },
    {
      "epoch": 0.6304549675023212,
      "grad_norm": 1.486096619479568,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8694337010383606,
      "num_tokens": 118129318.0,
      "step": 3395
    },
    {
      "epoch": 0.6306406685236768,
      "grad_norm": 1.4873394246724674,
      "learning_rate": 1e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8636516332626343,
      "num_tokens": 118164396.0,
      "step": 3396
    },
    {
      "epoch": 0.6308263695450325,
      "grad_norm": 1.5055197560902884,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8598015308380127,
      "num_tokens": 118196869.0,
      "step": 3397
    },
    {
      "epoch": 0.6310120705663881,
      "grad_norm": 1.4061575660852665,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8655619025230408,
      "num_tokens": 118233478.0,
      "step": 3398
    },
    {
      "epoch": 0.6311977715877437,
      "grad_norm": 1.347041892299358,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8678163290023804,
      "num_tokens": 118273399.0,
      "step": 3399
    },
    {
      "epoch": 0.6313834726090993,
      "grad_norm": 1.5299123842100857,
      "learning_rate": 1e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8601846694946289,
      "num_tokens": 118309934.0,
      "step": 3400
    },
    {
      "epoch": 0.631569173630455,
      "grad_norm": 1.4677227963379254,
      "learning_rate": 1e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8507775664329529,
      "num_tokens": 118348637.0,
      "step": 3401
    },
    {
      "epoch": 0.6317548746518106,
      "grad_norm": 1.7740967819206728,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8543981313705444,
      "num_tokens": 118378204.0,
      "step": 3402
    },
    {
      "epoch": 0.6319405756731662,
      "grad_norm": 1.4573517360393429,
      "learning_rate": 1e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8581254482269287,
      "num_tokens": 118414522.0,
      "step": 3403
    },
    {
      "epoch": 0.6321262766945218,
      "grad_norm": 1.4568407839316526,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8586568832397461,
      "num_tokens": 118450940.0,
      "step": 3404
    },
    {
      "epoch": 0.6323119777158774,
      "grad_norm": 1.406890708203634,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8572218418121338,
      "num_tokens": 118491464.0,
      "step": 3405
    },
    {
      "epoch": 0.6324976787372331,
      "grad_norm": 1.4874345024512803,
      "learning_rate": 1e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8498984575271606,
      "num_tokens": 118528859.0,
      "step": 3406
    },
    {
      "epoch": 0.6326833797585887,
      "grad_norm": 1.5176824176842056,
      "learning_rate": 1e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8562489748001099,
      "num_tokens": 118561567.0,
      "step": 3407
    },
    {
      "epoch": 0.6328690807799443,
      "grad_norm": 1.6144999222151648,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8763672709465027,
      "num_tokens": 118594028.0,
      "step": 3408
    },
    {
      "epoch": 0.6330547818012999,
      "grad_norm": 1.3183456587616498,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8731117248535156,
      "num_tokens": 118636410.0,
      "step": 3409
    },
    {
      "epoch": 0.6332404828226555,
      "grad_norm": 1.6090421405463327,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8693026900291443,
      "num_tokens": 118667278.0,
      "step": 3410
    },
    {
      "epoch": 0.6334261838440112,
      "grad_norm": 1.454279203830764,
      "learning_rate": 1e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8530293703079224,
      "num_tokens": 118704658.0,
      "step": 3411
    },
    {
      "epoch": 0.6336118848653668,
      "grad_norm": 1.4298150704700892,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.857408881187439,
      "num_tokens": 118749129.0,
      "step": 3412
    },
    {
      "epoch": 0.6337975858867224,
      "grad_norm": 1.4345242644870013,
      "learning_rate": 1e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8676337003707886,
      "num_tokens": 118785930.0,
      "step": 3413
    },
    {
      "epoch": 0.633983286908078,
      "grad_norm": 1.4131780082687837,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8622836470603943,
      "num_tokens": 118829998.0,
      "step": 3414
    },
    {
      "epoch": 0.6341689879294337,
      "grad_norm": 1.359388242950248,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8812706470489502,
      "num_tokens": 118866406.0,
      "step": 3415
    },
    {
      "epoch": 0.6343546889507893,
      "grad_norm": 1.6049052329546114,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8734594583511353,
      "num_tokens": 118893754.0,
      "step": 3416
    },
    {
      "epoch": 0.6345403899721448,
      "grad_norm": 1.5276774974082303,
      "learning_rate": 1e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8633013963699341,
      "num_tokens": 118927819.0,
      "step": 3417
    },
    {
      "epoch": 0.6347260909935004,
      "grad_norm": 1.535851273374873,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8774638175964355,
      "num_tokens": 118957555.0,
      "step": 3418
    },
    {
      "epoch": 0.634911792014856,
      "grad_norm": 1.598079820784967,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8617913722991943,
      "num_tokens": 118987560.0,
      "step": 3419
    },
    {
      "epoch": 0.6350974930362117,
      "grad_norm": 1.3923529381431785,
      "learning_rate": 1e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8552034497261047,
      "num_tokens": 119031265.0,
      "step": 3420
    },
    {
      "epoch": 0.6352831940575673,
      "grad_norm": 1.4911491319936512,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8563709259033203,
      "num_tokens": 119066020.0,
      "step": 3421
    },
    {
      "epoch": 0.6354688950789229,
      "grad_norm": 1.4531941771286767,
      "learning_rate": 1e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.859227180480957,
      "num_tokens": 119102307.0,
      "step": 3422
    },
    {
      "epoch": 0.6356545961002785,
      "grad_norm": 1.352732155604984,
      "learning_rate": 1e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.872237503528595,
      "num_tokens": 119140724.0,
      "step": 3423
    },
    {
      "epoch": 0.6358402971216341,
      "grad_norm": 1.415870848891765,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8748663663864136,
      "num_tokens": 119176289.0,
      "step": 3424
    },
    {
      "epoch": 0.6360259981429898,
      "grad_norm": 1.3138704024705068,
      "learning_rate": 1e-06,
      "loss": 0.3564,
      "mean_token_accuracy": 0.8795671463012695,
      "num_tokens": 119216023.0,
      "step": 3425
    },
    {
      "epoch": 0.6362116991643454,
      "grad_norm": 1.5735285506777934,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8662801384925842,
      "num_tokens": 119244891.0,
      "step": 3426
    },
    {
      "epoch": 0.636397400185701,
      "grad_norm": 1.5931322695887455,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8638241291046143,
      "num_tokens": 119280782.0,
      "step": 3427
    },
    {
      "epoch": 0.6365831012070566,
      "grad_norm": 1.6855004064972567,
      "learning_rate": 1e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8616905212402344,
      "num_tokens": 119307971.0,
      "step": 3428
    },
    {
      "epoch": 0.6367688022284123,
      "grad_norm": 1.4696089454216439,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8632295727729797,
      "num_tokens": 119344203.0,
      "step": 3429
    },
    {
      "epoch": 0.6369545032497679,
      "grad_norm": 1.4549211609218304,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.873353123664856,
      "num_tokens": 119378571.0,
      "step": 3430
    },
    {
      "epoch": 0.6371402042711235,
      "grad_norm": 1.550787166018662,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8658486604690552,
      "num_tokens": 119411377.0,
      "step": 3431
    },
    {
      "epoch": 0.6373259052924791,
      "grad_norm": 1.4990407150069784,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8717852830886841,
      "num_tokens": 119444907.0,
      "step": 3432
    },
    {
      "epoch": 0.6375116063138347,
      "grad_norm": 1.6413317685169087,
      "learning_rate": 1e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.841871976852417,
      "num_tokens": 119480065.0,
      "step": 3433
    },
    {
      "epoch": 0.6376973073351904,
      "grad_norm": 1.666135427782453,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8760099411010742,
      "num_tokens": 119511994.0,
      "step": 3434
    },
    {
      "epoch": 0.637883008356546,
      "grad_norm": 1.4766406852558884,
      "learning_rate": 1e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8709903955459595,
      "num_tokens": 119550130.0,
      "step": 3435
    },
    {
      "epoch": 0.6380687093779016,
      "grad_norm": 1.406672646363182,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8727165460586548,
      "num_tokens": 119587427.0,
      "step": 3436
    },
    {
      "epoch": 0.6382544103992572,
      "grad_norm": 1.4319847612628802,
      "learning_rate": 1e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8573681116104126,
      "num_tokens": 119628871.0,
      "step": 3437
    },
    {
      "epoch": 0.6384401114206129,
      "grad_norm": 1.5816901881736896,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8735572099685669,
      "num_tokens": 119662994.0,
      "step": 3438
    },
    {
      "epoch": 0.6386258124419685,
      "grad_norm": 1.487931553542303,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8663836717605591,
      "num_tokens": 119699501.0,
      "step": 3439
    },
    {
      "epoch": 0.6388115134633241,
      "grad_norm": 1.5149087610850052,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8719877004623413,
      "num_tokens": 119732506.0,
      "step": 3440
    },
    {
      "epoch": 0.6389972144846797,
      "grad_norm": 1.3946765427780654,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8681219816207886,
      "num_tokens": 119772231.0,
      "step": 3441
    },
    {
      "epoch": 0.6391829155060352,
      "grad_norm": 1.3900051848895567,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8775613307952881,
      "num_tokens": 119811247.0,
      "step": 3442
    },
    {
      "epoch": 0.6393686165273909,
      "grad_norm": 1.6098919102415299,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.860823929309845,
      "num_tokens": 119844597.0,
      "step": 3443
    },
    {
      "epoch": 0.6395543175487465,
      "grad_norm": 1.409906874355417,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8726941347122192,
      "num_tokens": 119883359.0,
      "step": 3444
    },
    {
      "epoch": 0.6397400185701021,
      "grad_norm": 1.5677934265287243,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8671213388442993,
      "num_tokens": 119914025.0,
      "step": 3445
    },
    {
      "epoch": 0.6399257195914577,
      "grad_norm": 1.5801512451172157,
      "learning_rate": 1e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8581124544143677,
      "num_tokens": 119945208.0,
      "step": 3446
    },
    {
      "epoch": 0.6401114206128133,
      "grad_norm": 1.46971757470439,
      "learning_rate": 1e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8536176681518555,
      "num_tokens": 119979727.0,
      "step": 3447
    },
    {
      "epoch": 0.640297121634169,
      "grad_norm": 1.424549787669767,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8776649236679077,
      "num_tokens": 120011911.0,
      "step": 3448
    },
    {
      "epoch": 0.6404828226555246,
      "grad_norm": 1.599710256350085,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8628717064857483,
      "num_tokens": 120047416.0,
      "step": 3449
    },
    {
      "epoch": 0.6406685236768802,
      "grad_norm": 1.472635224145227,
      "learning_rate": 1e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8466659784317017,
      "num_tokens": 120087189.0,
      "step": 3450
    },
    {
      "epoch": 0.6408542246982358,
      "grad_norm": 1.585684811271524,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8601049184799194,
      "num_tokens": 120116924.0,
      "step": 3451
    },
    {
      "epoch": 0.6410399257195915,
      "grad_norm": 1.551424637273529,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8632705211639404,
      "num_tokens": 120153126.0,
      "step": 3452
    },
    {
      "epoch": 0.6412256267409471,
      "grad_norm": 1.7643786535795736,
      "learning_rate": 1e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8506777882575989,
      "num_tokens": 120182919.0,
      "step": 3453
    },
    {
      "epoch": 0.6414113277623027,
      "grad_norm": 1.5026863682759404,
      "learning_rate": 1e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8524484634399414,
      "num_tokens": 120217646.0,
      "step": 3454
    },
    {
      "epoch": 0.6415970287836583,
      "grad_norm": 1.5445267251380612,
      "learning_rate": 1e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8561052083969116,
      "num_tokens": 120250708.0,
      "step": 3455
    },
    {
      "epoch": 0.6417827298050139,
      "grad_norm": 1.5826729327343843,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8707777857780457,
      "num_tokens": 120283876.0,
      "step": 3456
    },
    {
      "epoch": 0.6419684308263696,
      "grad_norm": 1.559430935554399,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8761084079742432,
      "num_tokens": 120315293.0,
      "step": 3457
    },
    {
      "epoch": 0.6421541318477252,
      "grad_norm": 1.4562954479285,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8615814447402954,
      "num_tokens": 120350808.0,
      "step": 3458
    },
    {
      "epoch": 0.6423398328690808,
      "grad_norm": 1.4494243060590837,
      "learning_rate": 1e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8509453535079956,
      "num_tokens": 120389198.0,
      "step": 3459
    },
    {
      "epoch": 0.6425255338904364,
      "grad_norm": 1.4705249325947645,
      "learning_rate": 1e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8534536361694336,
      "num_tokens": 120424765.0,
      "step": 3460
    },
    {
      "epoch": 0.642711234911792,
      "grad_norm": 1.3757061428179889,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8677487373352051,
      "num_tokens": 120464048.0,
      "step": 3461
    },
    {
      "epoch": 0.6428969359331477,
      "grad_norm": 1.5207798263366508,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8673004508018494,
      "num_tokens": 120496201.0,
      "step": 3462
    },
    {
      "epoch": 0.6430826369545033,
      "grad_norm": 1.49677076817242,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.871705174446106,
      "num_tokens": 120530715.0,
      "step": 3463
    },
    {
      "epoch": 0.6432683379758589,
      "grad_norm": 1.3227293692656064,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.867316722869873,
      "num_tokens": 120569974.0,
      "step": 3464
    },
    {
      "epoch": 0.6434540389972145,
      "grad_norm": 1.4272434879658864,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8619165420532227,
      "num_tokens": 120607391.0,
      "step": 3465
    },
    {
      "epoch": 0.64363974001857,
      "grad_norm": 1.4609223946646082,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8621468544006348,
      "num_tokens": 120643474.0,
      "step": 3466
    },
    {
      "epoch": 0.6438254410399257,
      "grad_norm": 1.505757688934773,
      "learning_rate": 1e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8536376953125,
      "num_tokens": 120682329.0,
      "step": 3467
    },
    {
      "epoch": 0.6440111420612813,
      "grad_norm": 1.583277013937275,
      "learning_rate": 1e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8476443290710449,
      "num_tokens": 120715706.0,
      "step": 3468
    },
    {
      "epoch": 0.6441968430826369,
      "grad_norm": 1.3825101213418869,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8701077699661255,
      "num_tokens": 120753970.0,
      "step": 3469
    },
    {
      "epoch": 0.6443825441039925,
      "grad_norm": 1.5660240387528064,
      "learning_rate": 1e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8486226201057434,
      "num_tokens": 120790251.0,
      "step": 3470
    },
    {
      "epoch": 0.6445682451253482,
      "grad_norm": 1.484295530273021,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.871097207069397,
      "num_tokens": 120821594.0,
      "step": 3471
    },
    {
      "epoch": 0.6447539461467038,
      "grad_norm": 1.401473573140341,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8714411854743958,
      "num_tokens": 120859304.0,
      "step": 3472
    },
    {
      "epoch": 0.6449396471680594,
      "grad_norm": 1.4366808269056015,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8717186450958252,
      "num_tokens": 120897874.0,
      "step": 3473
    },
    {
      "epoch": 0.645125348189415,
      "grad_norm": 1.527506291315228,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8732640743255615,
      "num_tokens": 120929521.0,
      "step": 3474
    },
    {
      "epoch": 0.6453110492107706,
      "grad_norm": 1.2975614855489679,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8797339200973511,
      "num_tokens": 120967644.0,
      "step": 3475
    },
    {
      "epoch": 0.6454967502321263,
      "grad_norm": 1.3546883683588853,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8597948551177979,
      "num_tokens": 121009295.0,
      "step": 3476
    },
    {
      "epoch": 0.6456824512534819,
      "grad_norm": 1.4754265979154295,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.873946487903595,
      "num_tokens": 121042171.0,
      "step": 3477
    },
    {
      "epoch": 0.6458681522748375,
      "grad_norm": 1.4224695797509457,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.876844584941864,
      "num_tokens": 121077732.0,
      "step": 3478
    },
    {
      "epoch": 0.6460538532961931,
      "grad_norm": 1.6319220260358909,
      "learning_rate": 1e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.84156334400177,
      "num_tokens": 121110362.0,
      "step": 3479
    },
    {
      "epoch": 0.6462395543175488,
      "grad_norm": 1.423887103813932,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.8717042207717896,
      "num_tokens": 121147079.0,
      "step": 3480
    },
    {
      "epoch": 0.6464252553389044,
      "grad_norm": 1.5400399132373435,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8647090792655945,
      "num_tokens": 121178022.0,
      "step": 3481
    },
    {
      "epoch": 0.64661095636026,
      "grad_norm": 1.576282558161475,
      "learning_rate": 1e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8489648103713989,
      "num_tokens": 121215098.0,
      "step": 3482
    },
    {
      "epoch": 0.6467966573816156,
      "grad_norm": 1.6098993745334114,
      "learning_rate": 1e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8665115833282471,
      "num_tokens": 121249080.0,
      "step": 3483
    },
    {
      "epoch": 0.6469823584029712,
      "grad_norm": 1.4435859571368943,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8713176846504211,
      "num_tokens": 121284212.0,
      "step": 3484
    },
    {
      "epoch": 0.6471680594243269,
      "grad_norm": 1.460686145832849,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8651117086410522,
      "num_tokens": 121322253.0,
      "step": 3485
    },
    {
      "epoch": 0.6473537604456825,
      "grad_norm": 1.594850393440382,
      "learning_rate": 1e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.849307656288147,
      "num_tokens": 121355584.0,
      "step": 3486
    },
    {
      "epoch": 0.6475394614670381,
      "grad_norm": 1.5106473628375705,
      "learning_rate": 1e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8573589324951172,
      "num_tokens": 121389562.0,
      "step": 3487
    },
    {
      "epoch": 0.6477251624883937,
      "grad_norm": 1.5394483713856006,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.864920973777771,
      "num_tokens": 121422435.0,
      "step": 3488
    },
    {
      "epoch": 0.6479108635097494,
      "grad_norm": 1.5437914706997231,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8561487197875977,
      "num_tokens": 121457290.0,
      "step": 3489
    },
    {
      "epoch": 0.6480965645311049,
      "grad_norm": 1.5066521526403522,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.87221759557724,
      "num_tokens": 121489309.0,
      "step": 3490
    },
    {
      "epoch": 0.6482822655524605,
      "grad_norm": 1.3139577448066508,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8639164566993713,
      "num_tokens": 121530512.0,
      "step": 3491
    },
    {
      "epoch": 0.6484679665738161,
      "grad_norm": 1.3870237463605604,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8652663230895996,
      "num_tokens": 121569623.0,
      "step": 3492
    },
    {
      "epoch": 0.6486536675951717,
      "grad_norm": 1.4677989164099325,
      "learning_rate": 1e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8527398109436035,
      "num_tokens": 121604861.0,
      "step": 3493
    },
    {
      "epoch": 0.6488393686165274,
      "grad_norm": 1.4516602964553935,
      "learning_rate": 1e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8556285500526428,
      "num_tokens": 121646503.0,
      "step": 3494
    },
    {
      "epoch": 0.649025069637883,
      "grad_norm": 1.5274283504426593,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8690117001533508,
      "num_tokens": 121679206.0,
      "step": 3495
    },
    {
      "epoch": 0.6492107706592386,
      "grad_norm": 1.4629188732955547,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8843166828155518,
      "num_tokens": 121712991.0,
      "step": 3496
    },
    {
      "epoch": 0.6493964716805942,
      "grad_norm": 1.5152902436019913,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8753511905670166,
      "num_tokens": 121745876.0,
      "step": 3497
    },
    {
      "epoch": 0.6495821727019498,
      "grad_norm": 1.6174715899983414,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8767725229263306,
      "num_tokens": 121773160.0,
      "step": 3498
    },
    {
      "epoch": 0.6497678737233055,
      "grad_norm": 1.3364456330052217,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8702104091644287,
      "num_tokens": 121815991.0,
      "step": 3499
    },
    {
      "epoch": 0.6499535747446611,
      "grad_norm": 1.497404830642948,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8643022179603577,
      "num_tokens": 121848768.0,
      "step": 3500
    },
    {
      "epoch": 0.6501392757660167,
      "grad_norm": 1.3830694117536626,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8718411326408386,
      "num_tokens": 121887559.0,
      "step": 3501
    },
    {
      "epoch": 0.6503249767873723,
      "grad_norm": 1.5286191242078642,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8736149668693542,
      "num_tokens": 121920066.0,
      "step": 3502
    },
    {
      "epoch": 0.650510677808728,
      "grad_norm": 1.3965633299080653,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8713802099227905,
      "num_tokens": 121955762.0,
      "step": 3503
    },
    {
      "epoch": 0.6506963788300836,
      "grad_norm": 1.7086932502936052,
      "learning_rate": 1e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8491830825805664,
      "num_tokens": 121986070.0,
      "step": 3504
    },
    {
      "epoch": 0.6508820798514392,
      "grad_norm": 1.4507287435011296,
      "learning_rate": 1e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8486945033073425,
      "num_tokens": 122024013.0,
      "step": 3505
    },
    {
      "epoch": 0.6510677808727948,
      "grad_norm": 1.3230940351685105,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8734139800071716,
      "num_tokens": 122063079.0,
      "step": 3506
    },
    {
      "epoch": 0.6512534818941504,
      "grad_norm": 1.390671016804537,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8651370406150818,
      "num_tokens": 122100901.0,
      "step": 3507
    },
    {
      "epoch": 0.6514391829155061,
      "grad_norm": 1.3176736405511649,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8728569149971008,
      "num_tokens": 122143128.0,
      "step": 3508
    },
    {
      "epoch": 0.6516248839368617,
      "grad_norm": 1.4068488887579573,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8732762932777405,
      "num_tokens": 122179968.0,
      "step": 3509
    },
    {
      "epoch": 0.6518105849582173,
      "grad_norm": 1.3965241995424216,
      "learning_rate": 1e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.863959550857544,
      "num_tokens": 122219363.0,
      "step": 3510
    },
    {
      "epoch": 0.6519962859795729,
      "grad_norm": 1.5330391152552756,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8654280304908752,
      "num_tokens": 122253334.0,
      "step": 3511
    },
    {
      "epoch": 0.6521819870009286,
      "grad_norm": 1.5756343192051143,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8629355430603027,
      "num_tokens": 122282582.0,
      "step": 3512
    },
    {
      "epoch": 0.6523676880222842,
      "grad_norm": 1.342315383642202,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.881594181060791,
      "num_tokens": 122320538.0,
      "step": 3513
    },
    {
      "epoch": 0.6525533890436397,
      "grad_norm": 1.5383741408528278,
      "learning_rate": 1e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8473207950592041,
      "num_tokens": 122353250.0,
      "step": 3514
    },
    {
      "epoch": 0.6527390900649953,
      "grad_norm": 1.4620916791327774,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8647538423538208,
      "num_tokens": 122387016.0,
      "step": 3515
    },
    {
      "epoch": 0.6529247910863509,
      "grad_norm": 1.4691501109591913,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8789399266242981,
      "num_tokens": 122417155.0,
      "step": 3516
    },
    {
      "epoch": 0.6531104921077066,
      "grad_norm": 1.5721828156611757,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8641468286514282,
      "num_tokens": 122448331.0,
      "step": 3517
    },
    {
      "epoch": 0.6532961931290622,
      "grad_norm": 1.5821176188859458,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8635895252227783,
      "num_tokens": 122478654.0,
      "step": 3518
    },
    {
      "epoch": 0.6534818941504178,
      "grad_norm": 1.3968972006172249,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8653204441070557,
      "num_tokens": 122519330.0,
      "step": 3519
    },
    {
      "epoch": 0.6536675951717734,
      "grad_norm": 1.457301476283862,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8723819851875305,
      "num_tokens": 122555730.0,
      "step": 3520
    },
    {
      "epoch": 0.653853296193129,
      "grad_norm": 1.7217540761613443,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.867849588394165,
      "num_tokens": 122586831.0,
      "step": 3521
    },
    {
      "epoch": 0.6540389972144847,
      "grad_norm": 1.5992068956276537,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.868440568447113,
      "num_tokens": 122614202.0,
      "step": 3522
    },
    {
      "epoch": 0.6542246982358403,
      "grad_norm": 1.5064192584180494,
      "learning_rate": 1e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8597944974899292,
      "num_tokens": 122650110.0,
      "step": 3523
    },
    {
      "epoch": 0.6544103992571959,
      "grad_norm": 1.4110380775454476,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8693955540657043,
      "num_tokens": 122688032.0,
      "step": 3524
    },
    {
      "epoch": 0.6545961002785515,
      "grad_norm": 1.6190621242776522,
      "learning_rate": 1e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8613326549530029,
      "num_tokens": 122717174.0,
      "step": 3525
    },
    {
      "epoch": 0.6547818012999072,
      "grad_norm": 1.483947631773908,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8769748210906982,
      "num_tokens": 122753241.0,
      "step": 3526
    },
    {
      "epoch": 0.6549675023212628,
      "grad_norm": 1.3818914764600545,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8660817742347717,
      "num_tokens": 122789968.0,
      "step": 3527
    },
    {
      "epoch": 0.6551532033426184,
      "grad_norm": 1.4633029031219469,
      "learning_rate": 1e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.8745222687721252,
      "num_tokens": 122822256.0,
      "step": 3528
    },
    {
      "epoch": 0.655338904363974,
      "grad_norm": 1.5364413006549664,
      "learning_rate": 1e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8640108704566956,
      "num_tokens": 122857509.0,
      "step": 3529
    },
    {
      "epoch": 0.6555246053853296,
      "grad_norm": 1.4962267135947462,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8625092506408691,
      "num_tokens": 122891118.0,
      "step": 3530
    },
    {
      "epoch": 0.6557103064066853,
      "grad_norm": 1.4459545241471063,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8673217296600342,
      "num_tokens": 122925045.0,
      "step": 3531
    },
    {
      "epoch": 0.6558960074280409,
      "grad_norm": 1.2858437179009885,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8705486059188843,
      "num_tokens": 122966513.0,
      "step": 3532
    },
    {
      "epoch": 0.6560817084493965,
      "grad_norm": 1.5131275669557682,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8757189512252808,
      "num_tokens": 122998570.0,
      "step": 3533
    },
    {
      "epoch": 0.6562674094707521,
      "grad_norm": 1.362197280036606,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8769139647483826,
      "num_tokens": 123037189.0,
      "step": 3534
    },
    {
      "epoch": 0.6564531104921077,
      "grad_norm": 1.5683376900306956,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8733439445495605,
      "num_tokens": 123068101.0,
      "step": 3535
    },
    {
      "epoch": 0.6566388115134634,
      "grad_norm": 1.5030352225958703,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8882949352264404,
      "num_tokens": 123100219.0,
      "step": 3536
    },
    {
      "epoch": 0.656824512534819,
      "grad_norm": 1.3819843121358826,
      "learning_rate": 1e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8571815490722656,
      "num_tokens": 123139800.0,
      "step": 3537
    },
    {
      "epoch": 0.6570102135561745,
      "grad_norm": 1.5961536119654325,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8699142932891846,
      "num_tokens": 123167633.0,
      "step": 3538
    },
    {
      "epoch": 0.6571959145775301,
      "grad_norm": 1.5737492636314077,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8664004802703857,
      "num_tokens": 123197376.0,
      "step": 3539
    },
    {
      "epoch": 0.6573816155988857,
      "grad_norm": 1.4707354012498257,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8723165988922119,
      "num_tokens": 123233241.0,
      "step": 3540
    },
    {
      "epoch": 0.6575673166202414,
      "grad_norm": 1.4927573969017722,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8715845346450806,
      "num_tokens": 123266334.0,
      "step": 3541
    },
    {
      "epoch": 0.657753017641597,
      "grad_norm": 1.5390851822391638,
      "learning_rate": 1e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.861873209476471,
      "num_tokens": 123305061.0,
      "step": 3542
    },
    {
      "epoch": 0.6579387186629526,
      "grad_norm": 1.5442230012971279,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8593554496765137,
      "num_tokens": 123338738.0,
      "step": 3543
    },
    {
      "epoch": 0.6581244196843082,
      "grad_norm": 1.4592670774305432,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8701853156089783,
      "num_tokens": 123374808.0,
      "step": 3544
    },
    {
      "epoch": 0.6583101207056639,
      "grad_norm": 1.5104590466198795,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8666156530380249,
      "num_tokens": 123405404.0,
      "step": 3545
    },
    {
      "epoch": 0.6584958217270195,
      "grad_norm": 1.764742246399013,
      "learning_rate": 1e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8599830865859985,
      "num_tokens": 123432177.0,
      "step": 3546
    },
    {
      "epoch": 0.6586815227483751,
      "grad_norm": 1.4755224604755857,
      "learning_rate": 1e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8557838797569275,
      "num_tokens": 123468169.0,
      "step": 3547
    },
    {
      "epoch": 0.6588672237697307,
      "grad_norm": 1.4745310007014263,
      "learning_rate": 1e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8604567050933838,
      "num_tokens": 123508043.0,
      "step": 3548
    },
    {
      "epoch": 0.6590529247910863,
      "grad_norm": 1.3381479279213802,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8730490207672119,
      "num_tokens": 123547288.0,
      "step": 3549
    },
    {
      "epoch": 0.659238625812442,
      "grad_norm": 1.479618233651526,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.865662693977356,
      "num_tokens": 123581425.0,
      "step": 3550
    },
    {
      "epoch": 0.6594243268337976,
      "grad_norm": 1.4993211172120466,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8757221698760986,
      "num_tokens": 123612945.0,
      "step": 3551
    },
    {
      "epoch": 0.6596100278551532,
      "grad_norm": 1.5568128083209873,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8656303286552429,
      "num_tokens": 123644868.0,
      "step": 3552
    },
    {
      "epoch": 0.6597957288765088,
      "grad_norm": 1.5959139307683627,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8547642230987549,
      "num_tokens": 123677273.0,
      "step": 3553
    },
    {
      "epoch": 0.6599814298978645,
      "grad_norm": 1.4304366894146217,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8673180341720581,
      "num_tokens": 123717702.0,
      "step": 3554
    },
    {
      "epoch": 0.6601671309192201,
      "grad_norm": 1.4544495180458623,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8739590644836426,
      "num_tokens": 123756159.0,
      "step": 3555
    },
    {
      "epoch": 0.6603528319405757,
      "grad_norm": 1.411492815804304,
      "learning_rate": 1e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8487591743469238,
      "num_tokens": 123797557.0,
      "step": 3556
    },
    {
      "epoch": 0.6605385329619313,
      "grad_norm": 1.4435911987131,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8729095458984375,
      "num_tokens": 123833024.0,
      "step": 3557
    },
    {
      "epoch": 0.660724233983287,
      "grad_norm": 1.4911770709680452,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8670017123222351,
      "num_tokens": 123867000.0,
      "step": 3558
    },
    {
      "epoch": 0.6609099350046426,
      "grad_norm": 1.4711201614421652,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8675133585929871,
      "num_tokens": 123900639.0,
      "step": 3559
    },
    {
      "epoch": 0.6610956360259982,
      "grad_norm": 1.4527614731117973,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8722275495529175,
      "num_tokens": 123939256.0,
      "step": 3560
    },
    {
      "epoch": 0.6612813370473538,
      "grad_norm": 1.518043730687403,
      "learning_rate": 1e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8531079888343811,
      "num_tokens": 123972170.0,
      "step": 3561
    },
    {
      "epoch": 0.6614670380687093,
      "grad_norm": 1.2882982551607591,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8617439270019531,
      "num_tokens": 124014919.0,
      "step": 3562
    },
    {
      "epoch": 0.661652739090065,
      "grad_norm": 1.6140763570777652,
      "learning_rate": 1e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8465416431427002,
      "num_tokens": 124048943.0,
      "step": 3563
    },
    {
      "epoch": 0.6618384401114206,
      "grad_norm": 1.4820162329523148,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8641382455825806,
      "num_tokens": 124086208.0,
      "step": 3564
    },
    {
      "epoch": 0.6620241411327762,
      "grad_norm": 1.405991781635903,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8680896162986755,
      "num_tokens": 124121546.0,
      "step": 3565
    },
    {
      "epoch": 0.6622098421541318,
      "grad_norm": 1.5120269122901284,
      "learning_rate": 1e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8506417274475098,
      "num_tokens": 124158455.0,
      "step": 3566
    },
    {
      "epoch": 0.6623955431754874,
      "grad_norm": 1.412067903863941,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8797019720077515,
      "num_tokens": 124191279.0,
      "step": 3567
    },
    {
      "epoch": 0.662581244196843,
      "grad_norm": 1.411566512115301,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.871049165725708,
      "num_tokens": 124228838.0,
      "step": 3568
    },
    {
      "epoch": 0.6627669452181987,
      "grad_norm": 1.5678667796360537,
      "learning_rate": 1e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8515454530715942,
      "num_tokens": 124263966.0,
      "step": 3569
    },
    {
      "epoch": 0.6629526462395543,
      "grad_norm": 1.493171302272809,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8735859394073486,
      "num_tokens": 124294815.0,
      "step": 3570
    },
    {
      "epoch": 0.6631383472609099,
      "grad_norm": 1.5418199151310863,
      "learning_rate": 1e-06,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8590059280395508,
      "num_tokens": 124323451.0,
      "step": 3571
    },
    {
      "epoch": 0.6633240482822655,
      "grad_norm": 1.4033169421914156,
      "learning_rate": 1e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8547838926315308,
      "num_tokens": 124358578.0,
      "step": 3572
    },
    {
      "epoch": 0.6635097493036212,
      "grad_norm": 1.4142205985968612,
      "learning_rate": 1e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8587356209754944,
      "num_tokens": 124396817.0,
      "step": 3573
    },
    {
      "epoch": 0.6636954503249768,
      "grad_norm": 1.5974192668655798,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8838971853256226,
      "num_tokens": 124426071.0,
      "step": 3574
    },
    {
      "epoch": 0.6638811513463324,
      "grad_norm": 1.4931487531978855,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8667107820510864,
      "num_tokens": 124459516.0,
      "step": 3575
    },
    {
      "epoch": 0.664066852367688,
      "grad_norm": 1.313862476031701,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.8922187089920044,
      "num_tokens": 124497290.0,
      "step": 3576
    },
    {
      "epoch": 0.6642525533890437,
      "grad_norm": 1.5228209991147745,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8555779457092285,
      "num_tokens": 124536386.0,
      "step": 3577
    },
    {
      "epoch": 0.6644382544103993,
      "grad_norm": 1.4683081529478275,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8796387314796448,
      "num_tokens": 124570573.0,
      "step": 3578
    },
    {
      "epoch": 0.6646239554317549,
      "grad_norm": 1.6436559958876482,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8660173416137695,
      "num_tokens": 124597337.0,
      "step": 3579
    },
    {
      "epoch": 0.6648096564531105,
      "grad_norm": 1.3940700450074526,
      "learning_rate": 1e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8596417903900146,
      "num_tokens": 124637252.0,
      "step": 3580
    },
    {
      "epoch": 0.6649953574744661,
      "grad_norm": 1.3998426809179745,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8639994859695435,
      "num_tokens": 124676830.0,
      "step": 3581
    },
    {
      "epoch": 0.6651810584958218,
      "grad_norm": 1.4247859241849603,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8740004897117615,
      "num_tokens": 124713658.0,
      "step": 3582
    },
    {
      "epoch": 0.6653667595171774,
      "grad_norm": 1.4553323364740152,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.855648398399353,
      "num_tokens": 124755751.0,
      "step": 3583
    },
    {
      "epoch": 0.665552460538533,
      "grad_norm": 1.4039916190696957,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8748096227645874,
      "num_tokens": 124792940.0,
      "step": 3584
    },
    {
      "epoch": 0.6657381615598886,
      "grad_norm": 1.596161395503918,
      "learning_rate": 1e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8529902696609497,
      "num_tokens": 124823945.0,
      "step": 3585
    },
    {
      "epoch": 0.6659238625812441,
      "grad_norm": 1.7653887622018267,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8643701672554016,
      "num_tokens": 124861064.0,
      "step": 3586
    },
    {
      "epoch": 0.6661095636025998,
      "grad_norm": 1.4705570187585726,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8591588735580444,
      "num_tokens": 124895797.0,
      "step": 3587
    },
    {
      "epoch": 0.6662952646239554,
      "grad_norm": 1.6701492257728163,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8574744462966919,
      "num_tokens": 124927307.0,
      "step": 3588
    },
    {
      "epoch": 0.666480965645311,
      "grad_norm": 1.4363500896206804,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.875544548034668,
      "num_tokens": 124959253.0,
      "step": 3589
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.5392787675915065,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8755825757980347,
      "num_tokens": 124987116.0,
      "step": 3590
    },
    {
      "epoch": 0.6668523676880223,
      "grad_norm": 1.4928166789655046,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8686991930007935,
      "num_tokens": 125022483.0,
      "step": 3591
    },
    {
      "epoch": 0.6670380687093779,
      "grad_norm": 1.3989773598428625,
      "learning_rate": 1e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8553173542022705,
      "num_tokens": 125058744.0,
      "step": 3592
    },
    {
      "epoch": 0.6672237697307335,
      "grad_norm": 1.5274087376600227,
      "learning_rate": 1e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8677467107772827,
      "num_tokens": 125087009.0,
      "step": 3593
    },
    {
      "epoch": 0.6674094707520891,
      "grad_norm": 1.4862587423659561,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8682634234428406,
      "num_tokens": 125121930.0,
      "step": 3594
    },
    {
      "epoch": 0.6675951717734447,
      "grad_norm": 1.5732351732391803,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8627011775970459,
      "num_tokens": 125154367.0,
      "step": 3595
    },
    {
      "epoch": 0.6677808727948004,
      "grad_norm": 1.5272683493500179,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8752383589744568,
      "num_tokens": 125186656.0,
      "step": 3596
    },
    {
      "epoch": 0.667966573816156,
      "grad_norm": 1.64336712697127,
      "learning_rate": 1e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8497793078422546,
      "num_tokens": 125215548.0,
      "step": 3597
    },
    {
      "epoch": 0.6681522748375116,
      "grad_norm": 1.5743150620523938,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8656376004219055,
      "num_tokens": 125254108.0,
      "step": 3598
    },
    {
      "epoch": 0.6683379758588672,
      "grad_norm": 1.4714953515296043,
      "learning_rate": 1e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8612456917762756,
      "num_tokens": 125290465.0,
      "step": 3599
    },
    {
      "epoch": 0.6685236768802229,
      "grad_norm": 1.47481007165811,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8615579605102539,
      "num_tokens": 125323898.0,
      "step": 3600
    },
    {
      "epoch": 0.6687093779015785,
      "grad_norm": 1.546393716964227,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8705176115036011,
      "num_tokens": 125354163.0,
      "step": 3601
    },
    {
      "epoch": 0.6688950789229341,
      "grad_norm": 1.40869483266945,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8650932908058167,
      "num_tokens": 125395766.0,
      "step": 3602
    },
    {
      "epoch": 0.6690807799442897,
      "grad_norm": 1.425314535212538,
      "learning_rate": 1e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.858777642250061,
      "num_tokens": 125429747.0,
      "step": 3603
    },
    {
      "epoch": 0.6692664809656453,
      "grad_norm": 1.4378147747461236,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8661206364631653,
      "num_tokens": 125470727.0,
      "step": 3604
    },
    {
      "epoch": 0.669452181987001,
      "grad_norm": 1.480307367782705,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.856404185295105,
      "num_tokens": 125508256.0,
      "step": 3605
    },
    {
      "epoch": 0.6696378830083566,
      "grad_norm": 1.4668921636784638,
      "learning_rate": 1e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8548224568367004,
      "num_tokens": 125547803.0,
      "step": 3606
    },
    {
      "epoch": 0.6698235840297122,
      "grad_norm": 1.6088659149838518,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8667028546333313,
      "num_tokens": 125576218.0,
      "step": 3607
    },
    {
      "epoch": 0.6700092850510678,
      "grad_norm": 1.4659952141683912,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8752115964889526,
      "num_tokens": 125608660.0,
      "step": 3608
    },
    {
      "epoch": 0.6701949860724234,
      "grad_norm": 1.6870362770763885,
      "learning_rate": 1e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8621726036071777,
      "num_tokens": 125635442.0,
      "step": 3609
    },
    {
      "epoch": 0.6703806870937791,
      "grad_norm": 1.629465425113808,
      "learning_rate": 1e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.861699104309082,
      "num_tokens": 125668661.0,
      "step": 3610
    },
    {
      "epoch": 0.6705663881151346,
      "grad_norm": 1.7612260047650539,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8668867349624634,
      "num_tokens": 125704669.0,
      "step": 3611
    },
    {
      "epoch": 0.6707520891364902,
      "grad_norm": 1.9581459671822408,
      "learning_rate": 1e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8538576364517212,
      "num_tokens": 125741358.0,
      "step": 3612
    },
    {
      "epoch": 0.6709377901578458,
      "grad_norm": 1.4166929979959315,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8749881982803345,
      "num_tokens": 125779190.0,
      "step": 3613
    },
    {
      "epoch": 0.6711234911792014,
      "grad_norm": 1.4211313630057314,
      "learning_rate": 1e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8544266223907471,
      "num_tokens": 125819240.0,
      "step": 3614
    },
    {
      "epoch": 0.6713091922005571,
      "grad_norm": 1.534069210767437,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8640356659889221,
      "num_tokens": 125853220.0,
      "step": 3615
    },
    {
      "epoch": 0.6714948932219127,
      "grad_norm": 1.510066444740479,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8709326982498169,
      "num_tokens": 125889491.0,
      "step": 3616
    },
    {
      "epoch": 0.6716805942432683,
      "grad_norm": 1.4265715836994215,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.867842972278595,
      "num_tokens": 125929360.0,
      "step": 3617
    },
    {
      "epoch": 0.6718662952646239,
      "grad_norm": 1.857845930436034,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.86940997838974,
      "num_tokens": 125961326.0,
      "step": 3618
    },
    {
      "epoch": 0.6720519962859796,
      "grad_norm": 1.7355518485876125,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8642070293426514,
      "num_tokens": 125985694.0,
      "step": 3619
    },
    {
      "epoch": 0.6722376973073352,
      "grad_norm": 1.4424211354077199,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8756144046783447,
      "num_tokens": 126022569.0,
      "step": 3620
    },
    {
      "epoch": 0.6724233983286908,
      "grad_norm": 1.4572706187935172,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8665964603424072,
      "num_tokens": 126059408.0,
      "step": 3621
    },
    {
      "epoch": 0.6726090993500464,
      "grad_norm": 1.3610557663891305,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.878655195236206,
      "num_tokens": 126099189.0,
      "step": 3622
    },
    {
      "epoch": 0.672794800371402,
      "grad_norm": 1.5029924804278554,
      "learning_rate": 1e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8546247482299805,
      "num_tokens": 126136354.0,
      "step": 3623
    },
    {
      "epoch": 0.6729805013927577,
      "grad_norm": 1.537422638200414,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8799808025360107,
      "num_tokens": 126170847.0,
      "step": 3624
    },
    {
      "epoch": 0.6731662024141133,
      "grad_norm": 1.7232572626131932,
      "learning_rate": 1e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.862583339214325,
      "num_tokens": 126200857.0,
      "step": 3625
    },
    {
      "epoch": 0.6733519034354689,
      "grad_norm": 1.5438978896467943,
      "learning_rate": 1e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8589462041854858,
      "num_tokens": 126235247.0,
      "step": 3626
    },
    {
      "epoch": 0.6735376044568245,
      "grad_norm": 1.4666206717586652,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8772754073143005,
      "num_tokens": 126271257.0,
      "step": 3627
    },
    {
      "epoch": 0.6737233054781802,
      "grad_norm": 1.6466066249307745,
      "learning_rate": 1e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8569519519805908,
      "num_tokens": 126299268.0,
      "step": 3628
    },
    {
      "epoch": 0.6739090064995358,
      "grad_norm": 1.5034595582256414,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8674152493476868,
      "num_tokens": 126332165.0,
      "step": 3629
    },
    {
      "epoch": 0.6740947075208914,
      "grad_norm": 1.508266887374568,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8795669078826904,
      "num_tokens": 126370254.0,
      "step": 3630
    },
    {
      "epoch": 0.674280408542247,
      "grad_norm": 1.6798100240386622,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.867265522480011,
      "num_tokens": 126399012.0,
      "step": 3631
    },
    {
      "epoch": 0.6744661095636026,
      "grad_norm": 1.570504612618712,
      "learning_rate": 1e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.843413233757019,
      "num_tokens": 126437686.0,
      "step": 3632
    },
    {
      "epoch": 0.6746518105849583,
      "grad_norm": 1.5399132418914603,
      "learning_rate": 1e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8555426597595215,
      "num_tokens": 126473391.0,
      "step": 3633
    },
    {
      "epoch": 0.6748375116063139,
      "grad_norm": 1.5061429915765538,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8764581084251404,
      "num_tokens": 126505189.0,
      "step": 3634
    },
    {
      "epoch": 0.6750232126276694,
      "grad_norm": 1.4837452906714315,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8663421869277954,
      "num_tokens": 126544483.0,
      "step": 3635
    },
    {
      "epoch": 0.675208913649025,
      "grad_norm": 1.4733135948995724,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8566796779632568,
      "num_tokens": 126584338.0,
      "step": 3636
    },
    {
      "epoch": 0.6753946146703806,
      "grad_norm": 1.5252592920963508,
      "learning_rate": 1e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8699962496757507,
      "num_tokens": 126616838.0,
      "step": 3637
    },
    {
      "epoch": 0.6755803156917363,
      "grad_norm": 1.3596955017028165,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8701746463775635,
      "num_tokens": 126653727.0,
      "step": 3638
    },
    {
      "epoch": 0.6757660167130919,
      "grad_norm": 1.4243216930604783,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8662554025650024,
      "num_tokens": 126690432.0,
      "step": 3639
    },
    {
      "epoch": 0.6759517177344475,
      "grad_norm": 1.4873990002865605,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8600766062736511,
      "num_tokens": 126725188.0,
      "step": 3640
    },
    {
      "epoch": 0.6761374187558031,
      "grad_norm": 1.5671502734749612,
      "learning_rate": 1e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8535345196723938,
      "num_tokens": 126757508.0,
      "step": 3641
    },
    {
      "epoch": 0.6763231197771588,
      "grad_norm": 1.5604720218177415,
      "learning_rate": 1e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8521857261657715,
      "num_tokens": 126800652.0,
      "step": 3642
    },
    {
      "epoch": 0.6765088207985144,
      "grad_norm": 1.4065841056979163,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8754309415817261,
      "num_tokens": 126834729.0,
      "step": 3643
    },
    {
      "epoch": 0.67669452181987,
      "grad_norm": 1.4363880265196247,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8600465655326843,
      "num_tokens": 126871112.0,
      "step": 3644
    },
    {
      "epoch": 0.6768802228412256,
      "grad_norm": 1.6396272881113316,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8774356842041016,
      "num_tokens": 126898929.0,
      "step": 3645
    },
    {
      "epoch": 0.6770659238625812,
      "grad_norm": 1.5068687608522526,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8605737090110779,
      "num_tokens": 126934021.0,
      "step": 3646
    },
    {
      "epoch": 0.6772516248839369,
      "grad_norm": 1.5917898381463298,
      "learning_rate": 1e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8560513257980347,
      "num_tokens": 126967975.0,
      "step": 3647
    },
    {
      "epoch": 0.6774373259052925,
      "grad_norm": 1.4377678902744808,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8658490180969238,
      "num_tokens": 127004676.0,
      "step": 3648
    },
    {
      "epoch": 0.6776230269266481,
      "grad_norm": 1.5056303935593163,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8738149404525757,
      "num_tokens": 127038544.0,
      "step": 3649
    },
    {
      "epoch": 0.6778087279480037,
      "grad_norm": 1.3497810882112884,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8715335726737976,
      "num_tokens": 127081305.0,
      "step": 3650
    },
    {
      "epoch": 0.6779944289693594,
      "grad_norm": 1.5387316758847884,
      "learning_rate": 1e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8447370529174805,
      "num_tokens": 127116653.0,
      "step": 3651
    },
    {
      "epoch": 0.678180129990715,
      "grad_norm": 1.4578532891491816,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8706927299499512,
      "num_tokens": 127150284.0,
      "step": 3652
    },
    {
      "epoch": 0.6783658310120706,
      "grad_norm": 1.4418261722995491,
      "learning_rate": 1e-06,
      "loss": 0.4192,
      "mean_token_accuracy": 0.8592381477355957,
      "num_tokens": 127186340.0,
      "step": 3653
    },
    {
      "epoch": 0.6785515320334262,
      "grad_norm": 1.4421578674067683,
      "learning_rate": 1e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8625447750091553,
      "num_tokens": 127219614.0,
      "step": 3654
    },
    {
      "epoch": 0.6787372330547818,
      "grad_norm": 1.5092089836798575,
      "learning_rate": 1e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8490118980407715,
      "num_tokens": 127256697.0,
      "step": 3655
    },
    {
      "epoch": 0.6789229340761375,
      "grad_norm": 1.4526760694879133,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8811870813369751,
      "num_tokens": 127289356.0,
      "step": 3656
    },
    {
      "epoch": 0.6791086350974931,
      "grad_norm": 1.440762138787626,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8701953887939453,
      "num_tokens": 127323608.0,
      "step": 3657
    },
    {
      "epoch": 0.6792943361188487,
      "grad_norm": 1.447353452860945,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8663496971130371,
      "num_tokens": 127360701.0,
      "step": 3658
    },
    {
      "epoch": 0.6794800371402042,
      "grad_norm": 1.3700013469396177,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8656143546104431,
      "num_tokens": 127399324.0,
      "step": 3659
    },
    {
      "epoch": 0.6796657381615598,
      "grad_norm": 1.3671234106197399,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8622385859489441,
      "num_tokens": 127438361.0,
      "step": 3660
    },
    {
      "epoch": 0.6798514391829155,
      "grad_norm": 1.5581178564034663,
      "learning_rate": 1e-06,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8664400577545166,
      "num_tokens": 127467783.0,
      "step": 3661
    },
    {
      "epoch": 0.6800371402042711,
      "grad_norm": 1.4038728600477244,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8732113838195801,
      "num_tokens": 127506901.0,
      "step": 3662
    },
    {
      "epoch": 0.6802228412256267,
      "grad_norm": 1.4381052494510742,
      "learning_rate": 1e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8545354604721069,
      "num_tokens": 127543676.0,
      "step": 3663
    },
    {
      "epoch": 0.6804085422469823,
      "grad_norm": 1.5342347157817111,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8673012852668762,
      "num_tokens": 127574508.0,
      "step": 3664
    },
    {
      "epoch": 0.680594243268338,
      "grad_norm": 1.5268400742636643,
      "learning_rate": 1e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8570886850357056,
      "num_tokens": 127610340.0,
      "step": 3665
    },
    {
      "epoch": 0.6807799442896936,
      "grad_norm": 1.6069499185057856,
      "learning_rate": 1e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.859857439994812,
      "num_tokens": 127642462.0,
      "step": 3666
    },
    {
      "epoch": 0.6809656453110492,
      "grad_norm": 1.42100917521759,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8679261803627014,
      "num_tokens": 127679085.0,
      "step": 3667
    },
    {
      "epoch": 0.6811513463324048,
      "grad_norm": 1.362771187929635,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8778888583183289,
      "num_tokens": 127718202.0,
      "step": 3668
    },
    {
      "epoch": 0.6813370473537604,
      "grad_norm": 1.6930314823305168,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.86217200756073,
      "num_tokens": 127749933.0,
      "step": 3669
    },
    {
      "epoch": 0.6815227483751161,
      "grad_norm": 1.429271983103965,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8639935255050659,
      "num_tokens": 127786193.0,
      "step": 3670
    },
    {
      "epoch": 0.6817084493964717,
      "grad_norm": 1.3811929234892337,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8747854232788086,
      "num_tokens": 127824115.0,
      "step": 3671
    },
    {
      "epoch": 0.6818941504178273,
      "grad_norm": 1.4667194374118522,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.868439257144928,
      "num_tokens": 127859491.0,
      "step": 3672
    },
    {
      "epoch": 0.6820798514391829,
      "grad_norm": 1.5810738210912896,
      "learning_rate": 1e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8575718998908997,
      "num_tokens": 127894552.0,
      "step": 3673
    },
    {
      "epoch": 0.6822655524605385,
      "grad_norm": 1.4097173646620094,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8753424882888794,
      "num_tokens": 127931425.0,
      "step": 3674
    },
    {
      "epoch": 0.6824512534818942,
      "grad_norm": 1.4550456755704404,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8750380277633667,
      "num_tokens": 127963681.0,
      "step": 3675
    },
    {
      "epoch": 0.6826369545032498,
      "grad_norm": 1.3792155134380468,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8732868432998657,
      "num_tokens": 128004100.0,
      "step": 3676
    },
    {
      "epoch": 0.6828226555246054,
      "grad_norm": 1.461647221779741,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8726860284805298,
      "num_tokens": 128039605.0,
      "step": 3677
    },
    {
      "epoch": 0.683008356545961,
      "grad_norm": 1.4985893580458407,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8718909621238708,
      "num_tokens": 128069962.0,
      "step": 3678
    },
    {
      "epoch": 0.6831940575673167,
      "grad_norm": 1.4393724709123108,
      "learning_rate": 1e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8614473342895508,
      "num_tokens": 128109434.0,
      "step": 3679
    },
    {
      "epoch": 0.6833797585886723,
      "grad_norm": 1.3195842964285514,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8811007738113403,
      "num_tokens": 128149486.0,
      "step": 3680
    },
    {
      "epoch": 0.6835654596100279,
      "grad_norm": 1.595413031679805,
      "learning_rate": 1e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8585445880889893,
      "num_tokens": 128179133.0,
      "step": 3681
    },
    {
      "epoch": 0.6837511606313835,
      "grad_norm": 1.4394583115459054,
      "learning_rate": 1e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8466354012489319,
      "num_tokens": 128215485.0,
      "step": 3682
    },
    {
      "epoch": 0.683936861652739,
      "grad_norm": 1.4592704541345842,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8739657998085022,
      "num_tokens": 128248163.0,
      "step": 3683
    },
    {
      "epoch": 0.6841225626740947,
      "grad_norm": 1.4369685419366225,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8625891208648682,
      "num_tokens": 128285354.0,
      "step": 3684
    },
    {
      "epoch": 0.6843082636954503,
      "grad_norm": 1.3698929275933776,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8687962293624878,
      "num_tokens": 128326309.0,
      "step": 3685
    },
    {
      "epoch": 0.6844939647168059,
      "grad_norm": 1.3738486181041345,
      "learning_rate": 1e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.866095244884491,
      "num_tokens": 128367662.0,
      "step": 3686
    },
    {
      "epoch": 0.6846796657381615,
      "grad_norm": 1.620442029538698,
      "learning_rate": 1e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.855024516582489,
      "num_tokens": 128403386.0,
      "step": 3687
    },
    {
      "epoch": 0.6848653667595171,
      "grad_norm": 1.5444917110842136,
      "learning_rate": 1e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8540700674057007,
      "num_tokens": 128439482.0,
      "step": 3688
    },
    {
      "epoch": 0.6850510677808728,
      "grad_norm": 1.4993529045564868,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8661295771598816,
      "num_tokens": 128475756.0,
      "step": 3689
    },
    {
      "epoch": 0.6852367688022284,
      "grad_norm": 1.416795489727141,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8717910051345825,
      "num_tokens": 128509377.0,
      "step": 3690
    },
    {
      "epoch": 0.685422469823584,
      "grad_norm": 1.5383373009957602,
      "learning_rate": 1e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8564440011978149,
      "num_tokens": 128541005.0,
      "step": 3691
    },
    {
      "epoch": 0.6856081708449396,
      "grad_norm": 1.3801780703683557,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8795328736305237,
      "num_tokens": 128576361.0,
      "step": 3692
    },
    {
      "epoch": 0.6857938718662953,
      "grad_norm": 1.4188778002370168,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8728055953979492,
      "num_tokens": 128615971.0,
      "step": 3693
    },
    {
      "epoch": 0.6859795728876509,
      "grad_norm": 1.5429674047818605,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8673326969146729,
      "num_tokens": 128648185.0,
      "step": 3694
    },
    {
      "epoch": 0.6861652739090065,
      "grad_norm": 1.5152495877854781,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8566400408744812,
      "num_tokens": 128683880.0,
      "step": 3695
    },
    {
      "epoch": 0.6863509749303621,
      "grad_norm": 1.389591763375107,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8676093816757202,
      "num_tokens": 128721940.0,
      "step": 3696
    },
    {
      "epoch": 0.6865366759517177,
      "grad_norm": 1.5236106831113334,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8711718320846558,
      "num_tokens": 128758083.0,
      "step": 3697
    },
    {
      "epoch": 0.6867223769730734,
      "grad_norm": 1.4327442607481946,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8737242221832275,
      "num_tokens": 128794453.0,
      "step": 3698
    },
    {
      "epoch": 0.686908077994429,
      "grad_norm": 1.5538033763869927,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8512647151947021,
      "num_tokens": 128827182.0,
      "step": 3699
    },
    {
      "epoch": 0.6870937790157846,
      "grad_norm": 1.398284763866062,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8719417452812195,
      "num_tokens": 128868299.0,
      "step": 3700
    },
    {
      "epoch": 0.6872794800371402,
      "grad_norm": 1.538452107289011,
      "learning_rate": 1e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8570683002471924,
      "num_tokens": 128899663.0,
      "step": 3701
    },
    {
      "epoch": 0.6874651810584959,
      "grad_norm": 1.477714799764391,
      "learning_rate": 1e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.851047158241272,
      "num_tokens": 128937154.0,
      "step": 3702
    },
    {
      "epoch": 0.6876508820798515,
      "grad_norm": 1.3176883489145923,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8726751804351807,
      "num_tokens": 128978139.0,
      "step": 3703
    },
    {
      "epoch": 0.6878365831012071,
      "grad_norm": 1.3850610330834732,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8647549748420715,
      "num_tokens": 129017117.0,
      "step": 3704
    },
    {
      "epoch": 0.6880222841225627,
      "grad_norm": 1.4531550098303412,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8721815943717957,
      "num_tokens": 129050683.0,
      "step": 3705
    },
    {
      "epoch": 0.6882079851439183,
      "grad_norm": 1.5017026754352505,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8612786531448364,
      "num_tokens": 129084178.0,
      "step": 3706
    },
    {
      "epoch": 0.6883936861652739,
      "grad_norm": 1.5046417518932085,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8630771040916443,
      "num_tokens": 129121287.0,
      "step": 3707
    },
    {
      "epoch": 0.6885793871866295,
      "grad_norm": 1.4016866092987406,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8663697242736816,
      "num_tokens": 129162716.0,
      "step": 3708
    },
    {
      "epoch": 0.6887650882079851,
      "grad_norm": 1.468999096385385,
      "learning_rate": 1e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.868998646736145,
      "num_tokens": 129197917.0,
      "step": 3709
    },
    {
      "epoch": 0.6889507892293407,
      "grad_norm": 1.4892475856356209,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8711987137794495,
      "num_tokens": 129233747.0,
      "step": 3710
    },
    {
      "epoch": 0.6891364902506963,
      "grad_norm": 1.539297637923181,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8704246282577515,
      "num_tokens": 129265205.0,
      "step": 3711
    },
    {
      "epoch": 0.689322191272052,
      "grad_norm": 1.5526235329511182,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8552570939064026,
      "num_tokens": 129303323.0,
      "step": 3712
    },
    {
      "epoch": 0.6895078922934076,
      "grad_norm": 1.8044496891471076,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8745800256729126,
      "num_tokens": 129326758.0,
      "step": 3713
    },
    {
      "epoch": 0.6896935933147632,
      "grad_norm": 1.584085672114724,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8576167821884155,
      "num_tokens": 129358043.0,
      "step": 3714
    },
    {
      "epoch": 0.6898792943361188,
      "grad_norm": 1.558263986744297,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8687459230422974,
      "num_tokens": 129388600.0,
      "step": 3715
    },
    {
      "epoch": 0.6900649953574745,
      "grad_norm": 1.6457719672734064,
      "learning_rate": 1e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8507726192474365,
      "num_tokens": 129420039.0,
      "step": 3716
    },
    {
      "epoch": 0.6902506963788301,
      "grad_norm": 1.5324154046594147,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8664711713790894,
      "num_tokens": 129450212.0,
      "step": 3717
    },
    {
      "epoch": 0.6904363974001857,
      "grad_norm": 1.4979534296240207,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8748940825462341,
      "num_tokens": 129480652.0,
      "step": 3718
    },
    {
      "epoch": 0.6906220984215413,
      "grad_norm": 1.4659640985953948,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8815852403640747,
      "num_tokens": 129510905.0,
      "step": 3719
    },
    {
      "epoch": 0.6908077994428969,
      "grad_norm": 1.452755167732793,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8709205389022827,
      "num_tokens": 129544946.0,
      "step": 3720
    },
    {
      "epoch": 0.6909935004642526,
      "grad_norm": 1.4358169250997233,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8773573637008667,
      "num_tokens": 129579082.0,
      "step": 3721
    },
    {
      "epoch": 0.6911792014856082,
      "grad_norm": 1.580571756388531,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8671554327011108,
      "num_tokens": 129611759.0,
      "step": 3722
    },
    {
      "epoch": 0.6913649025069638,
      "grad_norm": 1.4164020732501619,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8766875267028809,
      "num_tokens": 129647922.0,
      "step": 3723
    },
    {
      "epoch": 0.6915506035283194,
      "grad_norm": 1.4296019888480855,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8629298210144043,
      "num_tokens": 129683494.0,
      "step": 3724
    },
    {
      "epoch": 0.691736304549675,
      "grad_norm": 1.5992876918579957,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8635114431381226,
      "num_tokens": 129711122.0,
      "step": 3725
    },
    {
      "epoch": 0.6919220055710307,
      "grad_norm": 1.5386616694556055,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8659704327583313,
      "num_tokens": 129740162.0,
      "step": 3726
    },
    {
      "epoch": 0.6921077065923863,
      "grad_norm": 1.4880163932419763,
      "learning_rate": 1e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8511832356452942,
      "num_tokens": 129776747.0,
      "step": 3727
    },
    {
      "epoch": 0.6922934076137419,
      "grad_norm": 1.524705948869598,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8745297193527222,
      "num_tokens": 129804045.0,
      "step": 3728
    },
    {
      "epoch": 0.6924791086350975,
      "grad_norm": 1.5265497493634494,
      "learning_rate": 1e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.855619490146637,
      "num_tokens": 129838093.0,
      "step": 3729
    },
    {
      "epoch": 0.6926648096564532,
      "grad_norm": 1.874145121596016,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8636342883110046,
      "num_tokens": 129872613.0,
      "step": 3730
    },
    {
      "epoch": 0.6928505106778087,
      "grad_norm": 1.5941999170154157,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8612869381904602,
      "num_tokens": 129900909.0,
      "step": 3731
    },
    {
      "epoch": 0.6930362116991643,
      "grad_norm": 1.558455630792658,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8649919033050537,
      "num_tokens": 129930953.0,
      "step": 3732
    },
    {
      "epoch": 0.6932219127205199,
      "grad_norm": 1.5196643846426137,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8632267117500305,
      "num_tokens": 129965136.0,
      "step": 3733
    },
    {
      "epoch": 0.6934076137418755,
      "grad_norm": 1.538700632689354,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8599531054496765,
      "num_tokens": 129999781.0,
      "step": 3734
    },
    {
      "epoch": 0.6935933147632312,
      "grad_norm": 1.546116750679038,
      "learning_rate": 1e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8566508293151855,
      "num_tokens": 130032860.0,
      "step": 3735
    },
    {
      "epoch": 0.6937790157845868,
      "grad_norm": 1.4415829794896016,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8664520978927612,
      "num_tokens": 130071395.0,
      "step": 3736
    },
    {
      "epoch": 0.6939647168059424,
      "grad_norm": 1.5331429978660138,
      "learning_rate": 1e-06,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8483699560165405,
      "num_tokens": 130108996.0,
      "step": 3737
    },
    {
      "epoch": 0.694150417827298,
      "grad_norm": 1.4570836654138268,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8628557920455933,
      "num_tokens": 130145512.0,
      "step": 3738
    },
    {
      "epoch": 0.6943361188486536,
      "grad_norm": 1.4400268039957198,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8741085529327393,
      "num_tokens": 130180509.0,
      "step": 3739
    },
    {
      "epoch": 0.6945218198700093,
      "grad_norm": 1.3872200345834231,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8787845969200134,
      "num_tokens": 130216352.0,
      "step": 3740
    },
    {
      "epoch": 0.6947075208913649,
      "grad_norm": 1.6330925576809407,
      "learning_rate": 1e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8604955077171326,
      "num_tokens": 130246156.0,
      "step": 3741
    },
    {
      "epoch": 0.6948932219127205,
      "grad_norm": 1.5010213291794705,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8660562038421631,
      "num_tokens": 130281198.0,
      "step": 3742
    },
    {
      "epoch": 0.6950789229340761,
      "grad_norm": 1.493461769509548,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8583674430847168,
      "num_tokens": 130315656.0,
      "step": 3743
    },
    {
      "epoch": 0.6952646239554318,
      "grad_norm": 1.4095988327917661,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8651114702224731,
      "num_tokens": 130350979.0,
      "step": 3744
    },
    {
      "epoch": 0.6954503249767874,
      "grad_norm": 1.4255732313246148,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.873616635799408,
      "num_tokens": 130386911.0,
      "step": 3745
    },
    {
      "epoch": 0.695636025998143,
      "grad_norm": 1.614185147716774,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8758448362350464,
      "num_tokens": 130416551.0,
      "step": 3746
    },
    {
      "epoch": 0.6958217270194986,
      "grad_norm": 1.4112466748840828,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8599713444709778,
      "num_tokens": 130455106.0,
      "step": 3747
    },
    {
      "epoch": 0.6960074280408542,
      "grad_norm": 1.523125604645327,
      "learning_rate": 1e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8594731688499451,
      "num_tokens": 130489361.0,
      "step": 3748
    },
    {
      "epoch": 0.6961931290622099,
      "grad_norm": 1.4412280544495035,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8772052526473999,
      "num_tokens": 130523952.0,
      "step": 3749
    },
    {
      "epoch": 0.6963788300835655,
      "grad_norm": 1.3446982238123968,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8664472103118896,
      "num_tokens": 130563384.0,
      "step": 3750
    },
    {
      "epoch": 0.6965645311049211,
      "grad_norm": 1.3661095281833486,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8763463497161865,
      "num_tokens": 130600565.0,
      "step": 3751
    },
    {
      "epoch": 0.6967502321262767,
      "grad_norm": 1.5530991668842054,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8615984916687012,
      "num_tokens": 130635424.0,
      "step": 3752
    },
    {
      "epoch": 0.6969359331476324,
      "grad_norm": 1.5407303128427463,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8762091398239136,
      "num_tokens": 130666741.0,
      "step": 3753
    },
    {
      "epoch": 0.697121634168988,
      "grad_norm": 1.5010810237154055,
      "learning_rate": 1e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8564198017120361,
      "num_tokens": 130704519.0,
      "step": 3754
    },
    {
      "epoch": 0.6973073351903436,
      "grad_norm": 1.5540481278237899,
      "learning_rate": 1e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8519300222396851,
      "num_tokens": 130736686.0,
      "step": 3755
    },
    {
      "epoch": 0.6974930362116991,
      "grad_norm": 1.4445788728288984,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8632174730300903,
      "num_tokens": 130773815.0,
      "step": 3756
    },
    {
      "epoch": 0.6976787372330547,
      "grad_norm": 1.487447932108581,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8625317215919495,
      "num_tokens": 130808358.0,
      "step": 3757
    },
    {
      "epoch": 0.6978644382544104,
      "grad_norm": 1.32852546854158,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8775883316993713,
      "num_tokens": 130848075.0,
      "step": 3758
    },
    {
      "epoch": 0.698050139275766,
      "grad_norm": 1.4464539192332726,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8726064562797546,
      "num_tokens": 130882038.0,
      "step": 3759
    },
    {
      "epoch": 0.6982358402971216,
      "grad_norm": 1.4875220835753944,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8564361333847046,
      "num_tokens": 130916413.0,
      "step": 3760
    },
    {
      "epoch": 0.6984215413184772,
      "grad_norm": 1.476629319358137,
      "learning_rate": 1e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8689165115356445,
      "num_tokens": 130950067.0,
      "step": 3761
    },
    {
      "epoch": 0.6986072423398328,
      "grad_norm": 1.4524259914710171,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.86164790391922,
      "num_tokens": 130984473.0,
      "step": 3762
    },
    {
      "epoch": 0.6987929433611885,
      "grad_norm": 1.4231140031191516,
      "learning_rate": 1e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8500628471374512,
      "num_tokens": 131026022.0,
      "step": 3763
    },
    {
      "epoch": 0.6989786443825441,
      "grad_norm": 1.4672586428343797,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8599084615707397,
      "num_tokens": 131058960.0,
      "step": 3764
    },
    {
      "epoch": 0.6991643454038997,
      "grad_norm": 1.3759443364962096,
      "learning_rate": 1e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8638532757759094,
      "num_tokens": 131098697.0,
      "step": 3765
    },
    {
      "epoch": 0.6993500464252553,
      "grad_norm": 1.4203110755384183,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8709251284599304,
      "num_tokens": 131136139.0,
      "step": 3766
    },
    {
      "epoch": 0.699535747446611,
      "grad_norm": 1.5848493732454567,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8651912212371826,
      "num_tokens": 131169309.0,
      "step": 3767
    },
    {
      "epoch": 0.6997214484679666,
      "grad_norm": 1.4063695169407766,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8751899003982544,
      "num_tokens": 131207872.0,
      "step": 3768
    },
    {
      "epoch": 0.6999071494893222,
      "grad_norm": 1.4319011974513827,
      "learning_rate": 1e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8553979992866516,
      "num_tokens": 131248002.0,
      "step": 3769
    },
    {
      "epoch": 0.7000928505106778,
      "grad_norm": 1.432139448369025,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8744151592254639,
      "num_tokens": 131286464.0,
      "step": 3770
    },
    {
      "epoch": 0.7002785515320334,
      "grad_norm": 1.5426351857329217,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.862969160079956,
      "num_tokens": 131320198.0,
      "step": 3771
    },
    {
      "epoch": 0.7004642525533891,
      "grad_norm": 1.6090295865754343,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8686838150024414,
      "num_tokens": 131348042.0,
      "step": 3772
    },
    {
      "epoch": 0.7006499535747447,
      "grad_norm": 1.4782014202145544,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8694243431091309,
      "num_tokens": 131382204.0,
      "step": 3773
    },
    {
      "epoch": 0.7008356545961003,
      "grad_norm": 1.426683897749191,
      "learning_rate": 1e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8728216886520386,
      "num_tokens": 131417845.0,
      "step": 3774
    },
    {
      "epoch": 0.7010213556174559,
      "grad_norm": 1.6099226125524524,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8727253675460815,
      "num_tokens": 131447196.0,
      "step": 3775
    },
    {
      "epoch": 0.7012070566388116,
      "grad_norm": 1.518935571993932,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8726745247840881,
      "num_tokens": 131480690.0,
      "step": 3776
    },
    {
      "epoch": 0.7013927576601672,
      "grad_norm": 1.4939942788945462,
      "learning_rate": 1e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8609439134597778,
      "num_tokens": 131517474.0,
      "step": 3777
    },
    {
      "epoch": 0.7015784586815228,
      "grad_norm": 1.367887058761495,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8664383292198181,
      "num_tokens": 131555971.0,
      "step": 3778
    },
    {
      "epoch": 0.7017641597028784,
      "grad_norm": 1.708015911291509,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8747763633728027,
      "num_tokens": 131583924.0,
      "step": 3779
    },
    {
      "epoch": 0.7019498607242339,
      "grad_norm": 1.6549775494390468,
      "learning_rate": 1e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8621764183044434,
      "num_tokens": 131611612.0,
      "step": 3780
    },
    {
      "epoch": 0.7021355617455896,
      "grad_norm": 1.3820571858028565,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.867860734462738,
      "num_tokens": 131651097.0,
      "step": 3781
    },
    {
      "epoch": 0.7023212627669452,
      "grad_norm": 1.4696476405054746,
      "learning_rate": 1e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8389274477958679,
      "num_tokens": 131689880.0,
      "step": 3782
    },
    {
      "epoch": 0.7025069637883008,
      "grad_norm": 1.611687548922337,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8664109706878662,
      "num_tokens": 131722994.0,
      "step": 3783
    },
    {
      "epoch": 0.7026926648096564,
      "grad_norm": 1.5020690186474785,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8673389554023743,
      "num_tokens": 131757617.0,
      "step": 3784
    },
    {
      "epoch": 0.702878365831012,
      "grad_norm": 1.5319220560130205,
      "learning_rate": 1e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8664901256561279,
      "num_tokens": 131789690.0,
      "step": 3785
    },
    {
      "epoch": 0.7030640668523677,
      "grad_norm": 1.5064745888801498,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8732576370239258,
      "num_tokens": 131821686.0,
      "step": 3786
    },
    {
      "epoch": 0.7032497678737233,
      "grad_norm": 1.5588919492663957,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.874051570892334,
      "num_tokens": 131853700.0,
      "step": 3787
    },
    {
      "epoch": 0.7034354688950789,
      "grad_norm": 1.4416549569719546,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8806988000869751,
      "num_tokens": 131889230.0,
      "step": 3788
    },
    {
      "epoch": 0.7036211699164345,
      "grad_norm": 1.5013151220397047,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8659858703613281,
      "num_tokens": 131930477.0,
      "step": 3789
    },
    {
      "epoch": 0.7038068709377902,
      "grad_norm": 1.4820087010356955,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8655262589454651,
      "num_tokens": 131965081.0,
      "step": 3790
    },
    {
      "epoch": 0.7039925719591458,
      "grad_norm": 1.497912140424882,
      "learning_rate": 1e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8629704713821411,
      "num_tokens": 131998232.0,
      "step": 3791
    },
    {
      "epoch": 0.7041782729805014,
      "grad_norm": 1.3657932746383574,
      "learning_rate": 1e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.858761191368103,
      "num_tokens": 132038746.0,
      "step": 3792
    },
    {
      "epoch": 0.704363974001857,
      "grad_norm": 1.5281301880578244,
      "learning_rate": 1e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8640713691711426,
      "num_tokens": 132075183.0,
      "step": 3793
    },
    {
      "epoch": 0.7045496750232126,
      "grad_norm": 1.5064638639284675,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8612158298492432,
      "num_tokens": 132109590.0,
      "step": 3794
    },
    {
      "epoch": 0.7047353760445683,
      "grad_norm": 1.6871051910901242,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8617891073226929,
      "num_tokens": 132136437.0,
      "step": 3795
    },
    {
      "epoch": 0.7049210770659239,
      "grad_norm": 1.612620760175506,
      "learning_rate": 1e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8528779745101929,
      "num_tokens": 132169824.0,
      "step": 3796
    },
    {
      "epoch": 0.7051067780872795,
      "grad_norm": 1.6108839262113264,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8793908357620239,
      "num_tokens": 132197043.0,
      "step": 3797
    },
    {
      "epoch": 0.7052924791086351,
      "grad_norm": 1.3948929154835272,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.882640540599823,
      "num_tokens": 132235781.0,
      "step": 3798
    },
    {
      "epoch": 0.7054781801299908,
      "grad_norm": 1.5530492890322622,
      "learning_rate": 1e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8665182590484619,
      "num_tokens": 132271265.0,
      "step": 3799
    },
    {
      "epoch": 0.7056638811513464,
      "grad_norm": 1.3020920029963532,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8780314326286316,
      "num_tokens": 132311359.0,
      "step": 3800
    },
    {
      "epoch": 0.705849582172702,
      "grad_norm": 1.4961621955627373,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8650302886962891,
      "num_tokens": 132345931.0,
      "step": 3801
    },
    {
      "epoch": 0.7060352831940576,
      "grad_norm": 1.520759224840939,
      "learning_rate": 1e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8655787706375122,
      "num_tokens": 132384410.0,
      "step": 3802
    },
    {
      "epoch": 0.7062209842154132,
      "grad_norm": 1.5352149310487453,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8575182557106018,
      "num_tokens": 132420178.0,
      "step": 3803
    },
    {
      "epoch": 0.7064066852367687,
      "grad_norm": 1.5718861393254866,
      "learning_rate": 1e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8607918620109558,
      "num_tokens": 132455213.0,
      "step": 3804
    },
    {
      "epoch": 0.7065923862581244,
      "grad_norm": 1.4199180966345437,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8695220351219177,
      "num_tokens": 132489452.0,
      "step": 3805
    },
    {
      "epoch": 0.70677808727948,
      "grad_norm": 1.352141668705093,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8690480589866638,
      "num_tokens": 132529259.0,
      "step": 3806
    },
    {
      "epoch": 0.7069637883008356,
      "grad_norm": 1.7122554364046076,
      "learning_rate": 1e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8562933206558228,
      "num_tokens": 132557573.0,
      "step": 3807
    },
    {
      "epoch": 0.7071494893221912,
      "grad_norm": 1.6402732357069532,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8769412636756897,
      "num_tokens": 132585862.0,
      "step": 3808
    },
    {
      "epoch": 0.7073351903435469,
      "grad_norm": 1.5945885326574551,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8732089996337891,
      "num_tokens": 132619492.0,
      "step": 3809
    },
    {
      "epoch": 0.7075208913649025,
      "grad_norm": 1.468359245143516,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8806026577949524,
      "num_tokens": 132654301.0,
      "step": 3810
    },
    {
      "epoch": 0.7077065923862581,
      "grad_norm": 1.5070469976686043,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8712065815925598,
      "num_tokens": 132685546.0,
      "step": 3811
    },
    {
      "epoch": 0.7078922934076137,
      "grad_norm": 1.755693321121543,
      "learning_rate": 1e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.862041711807251,
      "num_tokens": 132714002.0,
      "step": 3812
    },
    {
      "epoch": 0.7080779944289693,
      "grad_norm": 1.4641903348606098,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8721415996551514,
      "num_tokens": 132746696.0,
      "step": 3813
    },
    {
      "epoch": 0.708263695450325,
      "grad_norm": 1.5681918363612848,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8613759279251099,
      "num_tokens": 132781146.0,
      "step": 3814
    },
    {
      "epoch": 0.7084493964716806,
      "grad_norm": 1.4382184656522252,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8645594716072083,
      "num_tokens": 132820449.0,
      "step": 3815
    },
    {
      "epoch": 0.7086350974930362,
      "grad_norm": 1.5491502360186546,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8602601289749146,
      "num_tokens": 132857939.0,
      "step": 3816
    },
    {
      "epoch": 0.7088207985143918,
      "grad_norm": 1.5632565232142495,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8804414868354797,
      "num_tokens": 132891387.0,
      "step": 3817
    },
    {
      "epoch": 0.7090064995357475,
      "grad_norm": 1.3938519072196498,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.859637975692749,
      "num_tokens": 132929105.0,
      "step": 3818
    },
    {
      "epoch": 0.7091922005571031,
      "grad_norm": 1.4878857352909471,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.859144926071167,
      "num_tokens": 132963932.0,
      "step": 3819
    },
    {
      "epoch": 0.7093779015784587,
      "grad_norm": 1.396932671802062,
      "learning_rate": 1e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8620941042900085,
      "num_tokens": 133005420.0,
      "step": 3820
    },
    {
      "epoch": 0.7095636025998143,
      "grad_norm": 1.4691065790855216,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8756324648857117,
      "num_tokens": 133038817.0,
      "step": 3821
    },
    {
      "epoch": 0.70974930362117,
      "grad_norm": 1.6171487417820114,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8662940859794617,
      "num_tokens": 133070417.0,
      "step": 3822
    },
    {
      "epoch": 0.7099350046425256,
      "grad_norm": 1.4170379802157842,
      "learning_rate": 1e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8535283803939819,
      "num_tokens": 133106022.0,
      "step": 3823
    },
    {
      "epoch": 0.7101207056638812,
      "grad_norm": 1.419836955699654,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8755241632461548,
      "num_tokens": 133146757.0,
      "step": 3824
    },
    {
      "epoch": 0.7103064066852368,
      "grad_norm": 1.4129279419247742,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8707395792007446,
      "num_tokens": 133181819.0,
      "step": 3825
    },
    {
      "epoch": 0.7104921077065924,
      "grad_norm": 1.7476449206294733,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8665421009063721,
      "num_tokens": 133207886.0,
      "step": 3826
    },
    {
      "epoch": 0.7106778087279481,
      "grad_norm": 1.4997572119822185,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8643606305122375,
      "num_tokens": 133239706.0,
      "step": 3827
    },
    {
      "epoch": 0.7108635097493036,
      "grad_norm": 1.5897371930584343,
      "learning_rate": 1e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8590751886367798,
      "num_tokens": 133272621.0,
      "step": 3828
    },
    {
      "epoch": 0.7110492107706592,
      "grad_norm": 1.581737693987201,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8703203797340393,
      "num_tokens": 133305091.0,
      "step": 3829
    },
    {
      "epoch": 0.7112349117920148,
      "grad_norm": 1.5516767353003211,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8643677830696106,
      "num_tokens": 133340530.0,
      "step": 3830
    },
    {
      "epoch": 0.7114206128133704,
      "grad_norm": 1.462719976994822,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8668073415756226,
      "num_tokens": 133372631.0,
      "step": 3831
    },
    {
      "epoch": 0.7116063138347261,
      "grad_norm": 1.523351542074211,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8635696768760681,
      "num_tokens": 133403921.0,
      "step": 3832
    },
    {
      "epoch": 0.7117920148560817,
      "grad_norm": 1.2947023986206059,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.870988130569458,
      "num_tokens": 133447098.0,
      "step": 3833
    },
    {
      "epoch": 0.7119777158774373,
      "grad_norm": 1.6378507169478669,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8568753600120544,
      "num_tokens": 133477687.0,
      "step": 3834
    },
    {
      "epoch": 0.7121634168987929,
      "grad_norm": 1.5570210303902028,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8650723099708557,
      "num_tokens": 133508209.0,
      "step": 3835
    },
    {
      "epoch": 0.7123491179201485,
      "grad_norm": 1.423774325064588,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8760458827018738,
      "num_tokens": 133544637.0,
      "step": 3836
    },
    {
      "epoch": 0.7125348189415042,
      "grad_norm": 1.533550744989521,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.87183678150177,
      "num_tokens": 133578014.0,
      "step": 3837
    },
    {
      "epoch": 0.7127205199628598,
      "grad_norm": 1.5109887647165632,
      "learning_rate": 1e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8503341674804688,
      "num_tokens": 133611718.0,
      "step": 3838
    },
    {
      "epoch": 0.7129062209842154,
      "grad_norm": 1.6201747565651197,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8609100580215454,
      "num_tokens": 133645390.0,
      "step": 3839
    },
    {
      "epoch": 0.713091922005571,
      "grad_norm": 1.6440027011259726,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8563845157623291,
      "num_tokens": 133676414.0,
      "step": 3840
    },
    {
      "epoch": 0.7132776230269267,
      "grad_norm": 1.3891662934248596,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.87626713514328,
      "num_tokens": 133717105.0,
      "step": 3841
    },
    {
      "epoch": 0.7134633240482823,
      "grad_norm": 1.6220749270875894,
      "learning_rate": 1e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8312106132507324,
      "num_tokens": 133747573.0,
      "step": 3842
    },
    {
      "epoch": 0.7136490250696379,
      "grad_norm": 1.5059409149853784,
      "learning_rate": 1e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8616451025009155,
      "num_tokens": 133782000.0,
      "step": 3843
    },
    {
      "epoch": 0.7138347260909935,
      "grad_norm": 1.529778392584661,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8682074546813965,
      "num_tokens": 133813575.0,
      "step": 3844
    },
    {
      "epoch": 0.7140204271123491,
      "grad_norm": 1.4432734273907484,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.872408390045166,
      "num_tokens": 133847609.0,
      "step": 3845
    },
    {
      "epoch": 0.7142061281337048,
      "grad_norm": 1.5445723858767197,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8623235821723938,
      "num_tokens": 133885943.0,
      "step": 3846
    },
    {
      "epoch": 0.7143918291550604,
      "grad_norm": 1.4029294742654794,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8745338320732117,
      "num_tokens": 133920549.0,
      "step": 3847
    },
    {
      "epoch": 0.714577530176416,
      "grad_norm": 1.5460923944404783,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8784860968589783,
      "num_tokens": 133950547.0,
      "step": 3848
    },
    {
      "epoch": 0.7147632311977716,
      "grad_norm": 1.4518359271090604,
      "learning_rate": 1e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8653385043144226,
      "num_tokens": 133984028.0,
      "step": 3849
    },
    {
      "epoch": 0.7149489322191273,
      "grad_norm": 1.5616707725856926,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.86626797914505,
      "num_tokens": 134016510.0,
      "step": 3850
    },
    {
      "epoch": 0.7151346332404829,
      "grad_norm": 1.6114899550659907,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8691336512565613,
      "num_tokens": 134047151.0,
      "step": 3851
    },
    {
      "epoch": 0.7153203342618384,
      "grad_norm": 1.524024957913287,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8729918003082275,
      "num_tokens": 134077416.0,
      "step": 3852
    },
    {
      "epoch": 0.715506035283194,
      "grad_norm": 1.5569244473910087,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8644994497299194,
      "num_tokens": 134111452.0,
      "step": 3853
    },
    {
      "epoch": 0.7156917363045496,
      "grad_norm": 1.5083296532069617,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8815218210220337,
      "num_tokens": 134147538.0,
      "step": 3854
    },
    {
      "epoch": 0.7158774373259053,
      "grad_norm": 1.4722689536863873,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8687078952789307,
      "num_tokens": 134179073.0,
      "step": 3855
    },
    {
      "epoch": 0.7160631383472609,
      "grad_norm": 1.4766681066710063,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.873826801776886,
      "num_tokens": 134212801.0,
      "step": 3856
    },
    {
      "epoch": 0.7162488393686165,
      "grad_norm": 1.5191922015687158,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8677048087120056,
      "num_tokens": 134245507.0,
      "step": 3857
    },
    {
      "epoch": 0.7164345403899721,
      "grad_norm": 1.3924876092558218,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8760585784912109,
      "num_tokens": 134281634.0,
      "step": 3858
    },
    {
      "epoch": 0.7166202414113277,
      "grad_norm": 1.5417845933002925,
      "learning_rate": 1e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8450987339019775,
      "num_tokens": 134320171.0,
      "step": 3859
    },
    {
      "epoch": 0.7168059424326834,
      "grad_norm": 1.404516885499783,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8648682832717896,
      "num_tokens": 134361838.0,
      "step": 3860
    },
    {
      "epoch": 0.716991643454039,
      "grad_norm": 1.4791819249205416,
      "learning_rate": 1e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8598411083221436,
      "num_tokens": 134397940.0,
      "step": 3861
    },
    {
      "epoch": 0.7171773444753946,
      "grad_norm": 1.487480712865877,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.876090407371521,
      "num_tokens": 134429106.0,
      "step": 3862
    },
    {
      "epoch": 0.7173630454967502,
      "grad_norm": 1.4642608187833028,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8712584972381592,
      "num_tokens": 134461717.0,
      "step": 3863
    },
    {
      "epoch": 0.7175487465181059,
      "grad_norm": 1.4762917438993923,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8674992322921753,
      "num_tokens": 134494507.0,
      "step": 3864
    },
    {
      "epoch": 0.7177344475394615,
      "grad_norm": 1.5118691853960022,
      "learning_rate": 1e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8659610748291016,
      "num_tokens": 134529458.0,
      "step": 3865
    },
    {
      "epoch": 0.7179201485608171,
      "grad_norm": 1.3735193406250539,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8595523834228516,
      "num_tokens": 134569463.0,
      "step": 3866
    },
    {
      "epoch": 0.7181058495821727,
      "grad_norm": 1.6035850922259665,
      "learning_rate": 1e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8596785068511963,
      "num_tokens": 134601702.0,
      "step": 3867
    },
    {
      "epoch": 0.7182915506035283,
      "grad_norm": 1.610182229642125,
      "learning_rate": 1e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8616110682487488,
      "num_tokens": 134631966.0,
      "step": 3868
    },
    {
      "epoch": 0.718477251624884,
      "grad_norm": 1.5691928627423182,
      "learning_rate": 1e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8517130613327026,
      "num_tokens": 134665062.0,
      "step": 3869
    },
    {
      "epoch": 0.7186629526462396,
      "grad_norm": 1.4793783113230452,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8767428398132324,
      "num_tokens": 134697490.0,
      "step": 3870
    },
    {
      "epoch": 0.7188486536675952,
      "grad_norm": 1.455334352568527,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8820726871490479,
      "num_tokens": 134728091.0,
      "step": 3871
    },
    {
      "epoch": 0.7190343546889508,
      "grad_norm": 1.4793372231314086,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8734590411186218,
      "num_tokens": 134761668.0,
      "step": 3872
    },
    {
      "epoch": 0.7192200557103064,
      "grad_norm": 1.5556779010009114,
      "learning_rate": 1e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8589813709259033,
      "num_tokens": 134791787.0,
      "step": 3873
    },
    {
      "epoch": 0.7194057567316621,
      "grad_norm": 1.4491006182544839,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8701493740081787,
      "num_tokens": 134827863.0,
      "step": 3874
    },
    {
      "epoch": 0.7195914577530177,
      "grad_norm": 1.3435520721800327,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8688628673553467,
      "num_tokens": 134868089.0,
      "step": 3875
    },
    {
      "epoch": 0.7197771587743732,
      "grad_norm": 1.4864759572926565,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8831557035446167,
      "num_tokens": 134901920.0,
      "step": 3876
    },
    {
      "epoch": 0.7199628597957288,
      "grad_norm": 1.5030586617048391,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8678253889083862,
      "num_tokens": 134935003.0,
      "step": 3877
    },
    {
      "epoch": 0.7201485608170844,
      "grad_norm": 1.5872273959227012,
      "learning_rate": 1e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8564727306365967,
      "num_tokens": 134966762.0,
      "step": 3878
    },
    {
      "epoch": 0.7203342618384401,
      "grad_norm": 1.4385064878975504,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.873951256275177,
      "num_tokens": 135002890.0,
      "step": 3879
    },
    {
      "epoch": 0.7205199628597957,
      "grad_norm": 1.4991966811988795,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8750893473625183,
      "num_tokens": 135035421.0,
      "step": 3880
    },
    {
      "epoch": 0.7207056638811513,
      "grad_norm": 1.5851001779545022,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.861278772354126,
      "num_tokens": 135066177.0,
      "step": 3881
    },
    {
      "epoch": 0.7208913649025069,
      "grad_norm": 1.4746940507319775,
      "learning_rate": 1e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8541345000267029,
      "num_tokens": 135105596.0,
      "step": 3882
    },
    {
      "epoch": 0.7210770659238626,
      "grad_norm": 1.6782509978621627,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8772354125976562,
      "num_tokens": 135132415.0,
      "step": 3883
    },
    {
      "epoch": 0.7212627669452182,
      "grad_norm": 1.3018096521911622,
      "learning_rate": 1e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8554667234420776,
      "num_tokens": 135179064.0,
      "step": 3884
    },
    {
      "epoch": 0.7214484679665738,
      "grad_norm": 1.5764248466057216,
      "learning_rate": 1e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8406083583831787,
      "num_tokens": 135215268.0,
      "step": 3885
    },
    {
      "epoch": 0.7216341689879294,
      "grad_norm": 1.4927664475772353,
      "learning_rate": 1e-06,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8631387948989868,
      "num_tokens": 135251759.0,
      "step": 3886
    },
    {
      "epoch": 0.721819870009285,
      "grad_norm": 1.4962780574736323,
      "learning_rate": 1e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.864509105682373,
      "num_tokens": 135284225.0,
      "step": 3887
    },
    {
      "epoch": 0.7220055710306407,
      "grad_norm": 1.440242831142785,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8619904518127441,
      "num_tokens": 135319148.0,
      "step": 3888
    },
    {
      "epoch": 0.7221912720519963,
      "grad_norm": 1.4050109474620278,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.874814510345459,
      "num_tokens": 135355874.0,
      "step": 3889
    },
    {
      "epoch": 0.7223769730733519,
      "grad_norm": 1.512097692037757,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8801405429840088,
      "num_tokens": 135385268.0,
      "step": 3890
    },
    {
      "epoch": 0.7225626740947075,
      "grad_norm": 1.4030450984083556,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8735567927360535,
      "num_tokens": 135422127.0,
      "step": 3891
    },
    {
      "epoch": 0.7227483751160632,
      "grad_norm": 1.4143794199493747,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8688582181930542,
      "num_tokens": 135456789.0,
      "step": 3892
    },
    {
      "epoch": 0.7229340761374188,
      "grad_norm": 1.5246624858590438,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8628978133201599,
      "num_tokens": 135494841.0,
      "step": 3893
    },
    {
      "epoch": 0.7231197771587744,
      "grad_norm": 1.4952891112011326,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.864665150642395,
      "num_tokens": 135532898.0,
      "step": 3894
    },
    {
      "epoch": 0.72330547818013,
      "grad_norm": 1.4995708982409173,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8732860088348389,
      "num_tokens": 135568766.0,
      "step": 3895
    },
    {
      "epoch": 0.7234911792014856,
      "grad_norm": 1.7222091591426425,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.861717700958252,
      "num_tokens": 135605869.0,
      "step": 3896
    },
    {
      "epoch": 0.7236768802228413,
      "grad_norm": 1.3711709561928411,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.864652156829834,
      "num_tokens": 135649237.0,
      "step": 3897
    },
    {
      "epoch": 0.7238625812441969,
      "grad_norm": 1.5775230672004121,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8666436076164246,
      "num_tokens": 135681777.0,
      "step": 3898
    },
    {
      "epoch": 0.7240482822655525,
      "grad_norm": 1.6830566977149926,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8557977676391602,
      "num_tokens": 135711356.0,
      "step": 3899
    },
    {
      "epoch": 0.724233983286908,
      "grad_norm": 1.521267949250611,
      "learning_rate": 1e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8650423288345337,
      "num_tokens": 135749769.0,
      "step": 3900
    },
    {
      "epoch": 0.7244196843082636,
      "grad_norm": 1.5844235078690099,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8544045686721802,
      "num_tokens": 135781776.0,
      "step": 3901
    },
    {
      "epoch": 0.7246053853296193,
      "grad_norm": 1.4281037970368604,
      "learning_rate": 1e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8522972464561462,
      "num_tokens": 135819937.0,
      "step": 3902
    },
    {
      "epoch": 0.7247910863509749,
      "grad_norm": 1.5517803925718374,
      "learning_rate": 1e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8518407940864563,
      "num_tokens": 135855412.0,
      "step": 3903
    },
    {
      "epoch": 0.7249767873723305,
      "grad_norm": 1.4111457919906292,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8606365919113159,
      "num_tokens": 135894357.0,
      "step": 3904
    },
    {
      "epoch": 0.7251624883936861,
      "grad_norm": 1.6089251010424497,
      "learning_rate": 1e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8533279895782471,
      "num_tokens": 135924646.0,
      "step": 3905
    },
    {
      "epoch": 0.7253481894150418,
      "grad_norm": 1.4341642032272812,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.868318498134613,
      "num_tokens": 135962833.0,
      "step": 3906
    },
    {
      "epoch": 0.7255338904363974,
      "grad_norm": 1.3373728177696846,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8807541728019714,
      "num_tokens": 136001043.0,
      "step": 3907
    },
    {
      "epoch": 0.725719591457753,
      "grad_norm": 1.2531968988479854,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.874618411064148,
      "num_tokens": 136045690.0,
      "step": 3908
    },
    {
      "epoch": 0.7259052924791086,
      "grad_norm": 1.418185864809947,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8687117099761963,
      "num_tokens": 136084082.0,
      "step": 3909
    },
    {
      "epoch": 0.7260909935004642,
      "grad_norm": 1.4445399927587512,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8631256818771362,
      "num_tokens": 136122395.0,
      "step": 3910
    },
    {
      "epoch": 0.7262766945218199,
      "grad_norm": 1.6698090556518528,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8754006624221802,
      "num_tokens": 136150078.0,
      "step": 3911
    },
    {
      "epoch": 0.7264623955431755,
      "grad_norm": 1.6562725408698002,
      "learning_rate": 1e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8708556294441223,
      "num_tokens": 136178165.0,
      "step": 3912
    },
    {
      "epoch": 0.7266480965645311,
      "grad_norm": 1.3839583066143666,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8627510070800781,
      "num_tokens": 136217428.0,
      "step": 3913
    },
    {
      "epoch": 0.7268337975858867,
      "grad_norm": 1.6242294010583909,
      "learning_rate": 1e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8569894433021545,
      "num_tokens": 136249259.0,
      "step": 3914
    },
    {
      "epoch": 0.7270194986072424,
      "grad_norm": 1.4084753695952845,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8778789043426514,
      "num_tokens": 136287222.0,
      "step": 3915
    },
    {
      "epoch": 0.727205199628598,
      "grad_norm": 1.4547003117301038,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.866662859916687,
      "num_tokens": 136321247.0,
      "step": 3916
    },
    {
      "epoch": 0.7273909006499536,
      "grad_norm": 1.561469184815722,
      "learning_rate": 1e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8516297340393066,
      "num_tokens": 136354155.0,
      "step": 3917
    },
    {
      "epoch": 0.7275766016713092,
      "grad_norm": 1.650213804841648,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8669933676719666,
      "num_tokens": 136384421.0,
      "step": 3918
    },
    {
      "epoch": 0.7277623026926648,
      "grad_norm": 1.4177433361298752,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8787676095962524,
      "num_tokens": 136420365.0,
      "step": 3919
    },
    {
      "epoch": 0.7279480037140205,
      "grad_norm": 1.4744639983628443,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8774152398109436,
      "num_tokens": 136452345.0,
      "step": 3920
    },
    {
      "epoch": 0.7281337047353761,
      "grad_norm": 1.3586119502198255,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8605954647064209,
      "num_tokens": 136498865.0,
      "step": 3921
    },
    {
      "epoch": 0.7283194057567317,
      "grad_norm": 1.5967780542597407,
      "learning_rate": 1e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8458048105239868,
      "num_tokens": 136532568.0,
      "step": 3922
    },
    {
      "epoch": 0.7285051067780873,
      "grad_norm": 1.370749113275402,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8790133595466614,
      "num_tokens": 136572030.0,
      "step": 3923
    },
    {
      "epoch": 0.728690807799443,
      "grad_norm": 1.4813458160954216,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8667850494384766,
      "num_tokens": 136606739.0,
      "step": 3924
    },
    {
      "epoch": 0.7288765088207985,
      "grad_norm": 1.4696118100890447,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.861029863357544,
      "num_tokens": 136644754.0,
      "step": 3925
    },
    {
      "epoch": 0.7290622098421541,
      "grad_norm": 1.4105397080647697,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8785409927368164,
      "num_tokens": 136680457.0,
      "step": 3926
    },
    {
      "epoch": 0.7292479108635097,
      "grad_norm": 1.4100274055097943,
      "learning_rate": 1e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8634162545204163,
      "num_tokens": 136720278.0,
      "step": 3927
    },
    {
      "epoch": 0.7294336118848653,
      "grad_norm": 1.3994935601892171,
      "learning_rate": 1e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8586183786392212,
      "num_tokens": 136759524.0,
      "step": 3928
    },
    {
      "epoch": 0.729619312906221,
      "grad_norm": 1.4917022000731615,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8756654262542725,
      "num_tokens": 136792576.0,
      "step": 3929
    },
    {
      "epoch": 0.7298050139275766,
      "grad_norm": 1.6777817495065301,
      "learning_rate": 1e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8663526773452759,
      "num_tokens": 136818245.0,
      "step": 3930
    },
    {
      "epoch": 0.7299907149489322,
      "grad_norm": 1.3826387748517104,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8735190033912659,
      "num_tokens": 136854897.0,
      "step": 3931
    },
    {
      "epoch": 0.7301764159702878,
      "grad_norm": 1.366929261834655,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8703203201293945,
      "num_tokens": 136892693.0,
      "step": 3932
    },
    {
      "epoch": 0.7303621169916434,
      "grad_norm": 1.3576272035827528,
      "learning_rate": 1e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8455711603164673,
      "num_tokens": 136933832.0,
      "step": 3933
    },
    {
      "epoch": 0.7305478180129991,
      "grad_norm": 1.5232743955017713,
      "learning_rate": 1e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8361731171607971,
      "num_tokens": 136969707.0,
      "step": 3934
    },
    {
      "epoch": 0.7307335190343547,
      "grad_norm": 1.391433890329875,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8705657720565796,
      "num_tokens": 137008811.0,
      "step": 3935
    },
    {
      "epoch": 0.7309192200557103,
      "grad_norm": 1.5121865742148404,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8690694570541382,
      "num_tokens": 137040037.0,
      "step": 3936
    },
    {
      "epoch": 0.7311049210770659,
      "grad_norm": 1.5750179032687373,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8726276755332947,
      "num_tokens": 137069995.0,
      "step": 3937
    },
    {
      "epoch": 0.7312906220984215,
      "grad_norm": 1.5065700317911872,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.8739407658576965,
      "num_tokens": 137100803.0,
      "step": 3938
    },
    {
      "epoch": 0.7314763231197772,
      "grad_norm": 1.5387823371349596,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8721411228179932,
      "num_tokens": 137132659.0,
      "step": 3939
    },
    {
      "epoch": 0.7316620241411328,
      "grad_norm": 1.478753101971923,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8647069334983826,
      "num_tokens": 137168898.0,
      "step": 3940
    },
    {
      "epoch": 0.7318477251624884,
      "grad_norm": 1.361708732128597,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8776241540908813,
      "num_tokens": 137207878.0,
      "step": 3941
    },
    {
      "epoch": 0.732033426183844,
      "grad_norm": 1.597890077414646,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8699461221694946,
      "num_tokens": 137239884.0,
      "step": 3942
    },
    {
      "epoch": 0.7322191272051997,
      "grad_norm": 1.4892487414997846,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8729531764984131,
      "num_tokens": 137269503.0,
      "step": 3943
    },
    {
      "epoch": 0.7324048282265553,
      "grad_norm": 1.5164696272996347,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8644169569015503,
      "num_tokens": 137300809.0,
      "step": 3944
    },
    {
      "epoch": 0.7325905292479109,
      "grad_norm": 1.4220526581304325,
      "learning_rate": 1e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8529880046844482,
      "num_tokens": 137340651.0,
      "step": 3945
    },
    {
      "epoch": 0.7327762302692665,
      "grad_norm": 1.6447687941689981,
      "learning_rate": 1e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8624165654182434,
      "num_tokens": 137371025.0,
      "step": 3946
    },
    {
      "epoch": 0.7329619312906221,
      "grad_norm": 1.4082739974294483,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8676497936248779,
      "num_tokens": 137405772.0,
      "step": 3947
    },
    {
      "epoch": 0.7331476323119778,
      "grad_norm": 1.3561198423888303,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8697828054428101,
      "num_tokens": 137449221.0,
      "step": 3948
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 1.3721097360282812,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8723242282867432,
      "num_tokens": 137485725.0,
      "step": 3949
    },
    {
      "epoch": 0.7335190343546889,
      "grad_norm": 1.4758874004806108,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.860642671585083,
      "num_tokens": 137522860.0,
      "step": 3950
    },
    {
      "epoch": 0.7337047353760445,
      "grad_norm": 1.5920650894936104,
      "learning_rate": 1e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8508935570716858,
      "num_tokens": 137557311.0,
      "step": 3951
    },
    {
      "epoch": 0.7338904363974001,
      "grad_norm": 1.5764927170250063,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8774237036705017,
      "num_tokens": 137591169.0,
      "step": 3952
    },
    {
      "epoch": 0.7340761374187558,
      "grad_norm": 1.3953142971696328,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8648349046707153,
      "num_tokens": 137629245.0,
      "step": 3953
    },
    {
      "epoch": 0.7342618384401114,
      "grad_norm": 1.5335124132725602,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.870512843132019,
      "num_tokens": 137660281.0,
      "step": 3954
    },
    {
      "epoch": 0.734447539461467,
      "grad_norm": 1.498271072936686,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8623061180114746,
      "num_tokens": 137696744.0,
      "step": 3955
    },
    {
      "epoch": 0.7346332404828226,
      "grad_norm": 1.406033612909265,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.870086669921875,
      "num_tokens": 137734246.0,
      "step": 3956
    },
    {
      "epoch": 0.7348189415041783,
      "grad_norm": 1.375065601385396,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8669489622116089,
      "num_tokens": 137774681.0,
      "step": 3957
    },
    {
      "epoch": 0.7350046425255339,
      "grad_norm": 1.5345369339169703,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8571901917457581,
      "num_tokens": 137812360.0,
      "step": 3958
    },
    {
      "epoch": 0.7351903435468895,
      "grad_norm": 1.4799874792789702,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8742977380752563,
      "num_tokens": 137845388.0,
      "step": 3959
    },
    {
      "epoch": 0.7353760445682451,
      "grad_norm": 1.617074434436691,
      "learning_rate": 1e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8552179336547852,
      "num_tokens": 137873715.0,
      "step": 3960
    },
    {
      "epoch": 0.7355617455896007,
      "grad_norm": 1.4962456187816617,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8726593255996704,
      "num_tokens": 137908577.0,
      "step": 3961
    },
    {
      "epoch": 0.7357474466109564,
      "grad_norm": 1.4157239635952295,
      "learning_rate": 1e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8472223877906799,
      "num_tokens": 137953545.0,
      "step": 3962
    },
    {
      "epoch": 0.735933147632312,
      "grad_norm": 1.5018625247620612,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8567837476730347,
      "num_tokens": 137989882.0,
      "step": 3963
    },
    {
      "epoch": 0.7361188486536676,
      "grad_norm": 1.4681251513868003,
      "learning_rate": 1e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8631154298782349,
      "num_tokens": 138025092.0,
      "step": 3964
    },
    {
      "epoch": 0.7363045496750232,
      "grad_norm": 1.4806678437027254,
      "learning_rate": 1e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8513223528862,
      "num_tokens": 138061718.0,
      "step": 3965
    },
    {
      "epoch": 0.7364902506963789,
      "grad_norm": 1.5021964453791132,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8706275820732117,
      "num_tokens": 138093895.0,
      "step": 3966
    },
    {
      "epoch": 0.7366759517177345,
      "grad_norm": 1.5768795468934582,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8609748482704163,
      "num_tokens": 138123253.0,
      "step": 3967
    },
    {
      "epoch": 0.7368616527390901,
      "grad_norm": 1.479478195939347,
      "learning_rate": 1e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8591985702514648,
      "num_tokens": 138155522.0,
      "step": 3968
    },
    {
      "epoch": 0.7370473537604457,
      "grad_norm": 1.5229338689131808,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8615772128105164,
      "num_tokens": 138192906.0,
      "step": 3969
    },
    {
      "epoch": 0.7372330547818013,
      "grad_norm": 1.6152357513917368,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8637886047363281,
      "num_tokens": 138223794.0,
      "step": 3970
    },
    {
      "epoch": 0.737418755803157,
      "grad_norm": 1.6108080619246583,
      "learning_rate": 1e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8482425212860107,
      "num_tokens": 138253679.0,
      "step": 3971
    },
    {
      "epoch": 0.7376044568245126,
      "grad_norm": 1.3083559018430384,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8724278807640076,
      "num_tokens": 138294776.0,
      "step": 3972
    },
    {
      "epoch": 0.7377901578458681,
      "grad_norm": 1.6406295982160808,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8755778074264526,
      "num_tokens": 138323175.0,
      "step": 3973
    },
    {
      "epoch": 0.7379758588672237,
      "grad_norm": 1.605512560980914,
      "learning_rate": 1e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8531675338745117,
      "num_tokens": 138353295.0,
      "step": 3974
    },
    {
      "epoch": 0.7381615598885793,
      "grad_norm": 1.4854834337218037,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8811204433441162,
      "num_tokens": 138386576.0,
      "step": 3975
    },
    {
      "epoch": 0.738347260909935,
      "grad_norm": 1.4153968029253479,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8695725798606873,
      "num_tokens": 138422736.0,
      "step": 3976
    },
    {
      "epoch": 0.7385329619312906,
      "grad_norm": 1.4475495481270004,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8686618208885193,
      "num_tokens": 138456667.0,
      "step": 3977
    },
    {
      "epoch": 0.7387186629526462,
      "grad_norm": 1.4436871870381707,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8843108415603638,
      "num_tokens": 138493950.0,
      "step": 3978
    },
    {
      "epoch": 0.7389043639740018,
      "grad_norm": 1.588585080316781,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8622395992279053,
      "num_tokens": 138525094.0,
      "step": 3979
    },
    {
      "epoch": 0.7390900649953575,
      "grad_norm": 1.3901706002184917,
      "learning_rate": 1e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8647345900535583,
      "num_tokens": 138561996.0,
      "step": 3980
    },
    {
      "epoch": 0.7392757660167131,
      "grad_norm": 1.3300109588647107,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8725037574768066,
      "num_tokens": 138602854.0,
      "step": 3981
    },
    {
      "epoch": 0.7394614670380687,
      "grad_norm": 1.3294251282405967,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8669266700744629,
      "num_tokens": 138642567.0,
      "step": 3982
    },
    {
      "epoch": 0.7396471680594243,
      "grad_norm": 1.3859030292343322,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8636200428009033,
      "num_tokens": 138682036.0,
      "step": 3983
    },
    {
      "epoch": 0.7398328690807799,
      "grad_norm": 1.4368501767871351,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8666481375694275,
      "num_tokens": 138714349.0,
      "step": 3984
    },
    {
      "epoch": 0.7400185701021356,
      "grad_norm": 1.5150909042876226,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8724576234817505,
      "num_tokens": 138744686.0,
      "step": 3985
    },
    {
      "epoch": 0.7402042711234912,
      "grad_norm": 1.4842486947940823,
      "learning_rate": 1e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.859673261642456,
      "num_tokens": 138781201.0,
      "step": 3986
    },
    {
      "epoch": 0.7403899721448468,
      "grad_norm": 1.475845320238406,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.865957498550415,
      "num_tokens": 138816764.0,
      "step": 3987
    },
    {
      "epoch": 0.7405756731662024,
      "grad_norm": 1.5508008098221588,
      "learning_rate": 1e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8510128259658813,
      "num_tokens": 138853093.0,
      "step": 3988
    },
    {
      "epoch": 0.740761374187558,
      "grad_norm": 1.3781883233138488,
      "learning_rate": 1e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8575947284698486,
      "num_tokens": 138893716.0,
      "step": 3989
    },
    {
      "epoch": 0.7409470752089137,
      "grad_norm": 1.4123791190874175,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8605661392211914,
      "num_tokens": 138934559.0,
      "step": 3990
    },
    {
      "epoch": 0.7411327762302693,
      "grad_norm": 1.5580205664997637,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8644466996192932,
      "num_tokens": 138965908.0,
      "step": 3991
    },
    {
      "epoch": 0.7413184772516249,
      "grad_norm": 1.34488835033438,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8673242330551147,
      "num_tokens": 139006417.0,
      "step": 3992
    },
    {
      "epoch": 0.7415041782729805,
      "grad_norm": 1.4679959372341889,
      "learning_rate": 1e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8517321944236755,
      "num_tokens": 139046883.0,
      "step": 3993
    },
    {
      "epoch": 0.7416898792943362,
      "grad_norm": 1.4753126735144848,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8723162412643433,
      "num_tokens": 139079631.0,
      "step": 3994
    },
    {
      "epoch": 0.7418755803156918,
      "grad_norm": 1.6326987932336787,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.872899055480957,
      "num_tokens": 139108523.0,
      "step": 3995
    },
    {
      "epoch": 0.7420612813370474,
      "grad_norm": 1.68619376330547,
      "learning_rate": 1e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8513821959495544,
      "num_tokens": 139136938.0,
      "step": 3996
    },
    {
      "epoch": 0.7422469823584029,
      "grad_norm": 1.3418313814756035,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8685814142227173,
      "num_tokens": 139176471.0,
      "step": 3997
    },
    {
      "epoch": 0.7424326833797585,
      "grad_norm": 1.257698794951951,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8865083456039429,
      "num_tokens": 139216304.0,
      "step": 3998
    },
    {
      "epoch": 0.7426183844011142,
      "grad_norm": 1.576963804344246,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8711151480674744,
      "num_tokens": 139253115.0,
      "step": 3999
    },
    {
      "epoch": 0.7428040854224698,
      "grad_norm": 1.402738867500003,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8716152906417847,
      "num_tokens": 139296502.0,
      "step": 4000
    },
    {
      "epoch": 0.7429897864438254,
      "grad_norm": 1.469005039900255,
      "learning_rate": 1e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8512970209121704,
      "num_tokens": 139333539.0,
      "step": 4001
    },
    {
      "epoch": 0.743175487465181,
      "grad_norm": 1.4357859632931993,
      "learning_rate": 1e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8524887561798096,
      "num_tokens": 139371314.0,
      "step": 4002
    },
    {
      "epoch": 0.7433611884865367,
      "grad_norm": 1.453090484923566,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8688968420028687,
      "num_tokens": 139405498.0,
      "step": 4003
    },
    {
      "epoch": 0.7435468895078923,
      "grad_norm": 1.46948734716465,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8589686751365662,
      "num_tokens": 139439292.0,
      "step": 4004
    },
    {
      "epoch": 0.7437325905292479,
      "grad_norm": 1.4523342224499598,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8711024522781372,
      "num_tokens": 139474566.0,
      "step": 4005
    },
    {
      "epoch": 0.7439182915506035,
      "grad_norm": 1.5267939633612764,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8633270263671875,
      "num_tokens": 139511354.0,
      "step": 4006
    },
    {
      "epoch": 0.7441039925719591,
      "grad_norm": 1.4723800750625153,
      "learning_rate": 1e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8546844720840454,
      "num_tokens": 139550927.0,
      "step": 4007
    },
    {
      "epoch": 0.7442896935933148,
      "grad_norm": 1.68377584538203,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8784040808677673,
      "num_tokens": 139580444.0,
      "step": 4008
    },
    {
      "epoch": 0.7444753946146704,
      "grad_norm": 1.3510545449009066,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8710833787918091,
      "num_tokens": 139618409.0,
      "step": 4009
    },
    {
      "epoch": 0.744661095636026,
      "grad_norm": 1.5673103203023202,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8719743490219116,
      "num_tokens": 139646047.0,
      "step": 4010
    },
    {
      "epoch": 0.7448467966573816,
      "grad_norm": 1.4758737624695617,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8847122192382812,
      "num_tokens": 139677516.0,
      "step": 4011
    },
    {
      "epoch": 0.7450324976787372,
      "grad_norm": 1.5686928441258432,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8693402409553528,
      "num_tokens": 139710650.0,
      "step": 4012
    },
    {
      "epoch": 0.7452181987000929,
      "grad_norm": 1.3578149550720457,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8656061887741089,
      "num_tokens": 139746912.0,
      "step": 4013
    },
    {
      "epoch": 0.7454038997214485,
      "grad_norm": 1.4210249050358335,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8639469146728516,
      "num_tokens": 139785726.0,
      "step": 4014
    },
    {
      "epoch": 0.7455896007428041,
      "grad_norm": 1.453913233572018,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8896844387054443,
      "num_tokens": 139816267.0,
      "step": 4015
    },
    {
      "epoch": 0.7457753017641597,
      "grad_norm": 1.452768254593601,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8718069791793823,
      "num_tokens": 139850990.0,
      "step": 4016
    },
    {
      "epoch": 0.7459610027855154,
      "grad_norm": 1.5071539415883033,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8571240901947021,
      "num_tokens": 139889769.0,
      "step": 4017
    },
    {
      "epoch": 0.746146703806871,
      "grad_norm": 1.5928729279940042,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.852916955947876,
      "num_tokens": 139921946.0,
      "step": 4018
    },
    {
      "epoch": 0.7463324048282266,
      "grad_norm": 1.5633165135926608,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8591845631599426,
      "num_tokens": 139956573.0,
      "step": 4019
    },
    {
      "epoch": 0.7465181058495822,
      "grad_norm": 1.4112866339423997,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8682619333267212,
      "num_tokens": 139990920.0,
      "step": 4020
    },
    {
      "epoch": 0.7467038068709377,
      "grad_norm": 1.384492753460644,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8704853057861328,
      "num_tokens": 140027881.0,
      "step": 4021
    },
    {
      "epoch": 0.7468895078922934,
      "grad_norm": 1.4339006189370618,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8672828674316406,
      "num_tokens": 140065691.0,
      "step": 4022
    },
    {
      "epoch": 0.747075208913649,
      "grad_norm": 1.3619056140254076,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8734415173530579,
      "num_tokens": 140104142.0,
      "step": 4023
    },
    {
      "epoch": 0.7472609099350046,
      "grad_norm": 1.4089808745145702,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.890091061592102,
      "num_tokens": 140142902.0,
      "step": 4024
    },
    {
      "epoch": 0.7474466109563602,
      "grad_norm": 1.6379734704795648,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8668051362037659,
      "num_tokens": 140172053.0,
      "step": 4025
    },
    {
      "epoch": 0.7476323119777158,
      "grad_norm": 1.536092068031671,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8758378028869629,
      "num_tokens": 140205053.0,
      "step": 4026
    },
    {
      "epoch": 0.7478180129990715,
      "grad_norm": 1.360995261802133,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8764220476150513,
      "num_tokens": 140240695.0,
      "step": 4027
    },
    {
      "epoch": 0.7480037140204271,
      "grad_norm": 1.3519547413148079,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8836368322372437,
      "num_tokens": 140278555.0,
      "step": 4028
    },
    {
      "epoch": 0.7481894150417827,
      "grad_norm": 1.5122395876745114,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8815180063247681,
      "num_tokens": 140311309.0,
      "step": 4029
    },
    {
      "epoch": 0.7483751160631383,
      "grad_norm": 1.5281236619388936,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8768709897994995,
      "num_tokens": 140341741.0,
      "step": 4030
    },
    {
      "epoch": 0.748560817084494,
      "grad_norm": 1.4387333746318562,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8707684874534607,
      "num_tokens": 140379570.0,
      "step": 4031
    },
    {
      "epoch": 0.7487465181058496,
      "grad_norm": 1.3965182391554618,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8800944089889526,
      "num_tokens": 140416898.0,
      "step": 4032
    },
    {
      "epoch": 0.7489322191272052,
      "grad_norm": 1.5033475819695954,
      "learning_rate": 1e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8647909164428711,
      "num_tokens": 140452884.0,
      "step": 4033
    },
    {
      "epoch": 0.7491179201485608,
      "grad_norm": 1.3743697900261789,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8725636005401611,
      "num_tokens": 140490243.0,
      "step": 4034
    },
    {
      "epoch": 0.7493036211699164,
      "grad_norm": 1.492268700996671,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8602404594421387,
      "num_tokens": 140526602.0,
      "step": 4035
    },
    {
      "epoch": 0.7494893221912721,
      "grad_norm": 1.326142860431272,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8770098686218262,
      "num_tokens": 140564025.0,
      "step": 4036
    },
    {
      "epoch": 0.7496750232126277,
      "grad_norm": 1.573352341490179,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8607170581817627,
      "num_tokens": 140597651.0,
      "step": 4037
    },
    {
      "epoch": 0.7498607242339833,
      "grad_norm": 1.3940225769194885,
      "learning_rate": 1e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8551391959190369,
      "num_tokens": 140636587.0,
      "step": 4038
    },
    {
      "epoch": 0.7500464252553389,
      "grad_norm": 1.5089123626590841,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8649402260780334,
      "num_tokens": 140670480.0,
      "step": 4039
    },
    {
      "epoch": 0.7502321262766946,
      "grad_norm": 1.458755032080328,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8752310872077942,
      "num_tokens": 140706571.0,
      "step": 4040
    },
    {
      "epoch": 0.7504178272980502,
      "grad_norm": 1.3576900917052448,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8773483633995056,
      "num_tokens": 140744150.0,
      "step": 4041
    },
    {
      "epoch": 0.7506035283194058,
      "grad_norm": 1.452614492163226,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8693598508834839,
      "num_tokens": 140781837.0,
      "step": 4042
    },
    {
      "epoch": 0.7507892293407614,
      "grad_norm": 1.5554561867796743,
      "learning_rate": 1e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8545527458190918,
      "num_tokens": 140819305.0,
      "step": 4043
    },
    {
      "epoch": 0.750974930362117,
      "grad_norm": 1.5413673164550852,
      "learning_rate": 1e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8532580733299255,
      "num_tokens": 140852400.0,
      "step": 4044
    },
    {
      "epoch": 0.7511606313834726,
      "grad_norm": 1.4988277626784907,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8576116561889648,
      "num_tokens": 140890831.0,
      "step": 4045
    },
    {
      "epoch": 0.7513463324048282,
      "grad_norm": 1.4838496148615339,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8754629492759705,
      "num_tokens": 140924530.0,
      "step": 4046
    },
    {
      "epoch": 0.7515320334261838,
      "grad_norm": 1.5238846799349057,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8658404350280762,
      "num_tokens": 140958733.0,
      "step": 4047
    },
    {
      "epoch": 0.7517177344475394,
      "grad_norm": 1.5276804551852505,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8741503357887268,
      "num_tokens": 140996695.0,
      "step": 4048
    },
    {
      "epoch": 0.751903435468895,
      "grad_norm": 1.4035329045404343,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8565558195114136,
      "num_tokens": 141035997.0,
      "step": 4049
    },
    {
      "epoch": 0.7520891364902507,
      "grad_norm": 1.4987222520227108,
      "learning_rate": 1e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8654916882514954,
      "num_tokens": 141071111.0,
      "step": 4050
    },
    {
      "epoch": 0.7522748375116063,
      "grad_norm": 1.609988298132211,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.859000027179718,
      "num_tokens": 141099612.0,
      "step": 4051
    },
    {
      "epoch": 0.7524605385329619,
      "grad_norm": 1.673029516860071,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8747144937515259,
      "num_tokens": 141128407.0,
      "step": 4052
    },
    {
      "epoch": 0.7526462395543175,
      "grad_norm": 1.433342646731296,
      "learning_rate": 1e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8522239327430725,
      "num_tokens": 141165048.0,
      "step": 4053
    },
    {
      "epoch": 0.7528319405756732,
      "grad_norm": 1.4422865068618438,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8718352317810059,
      "num_tokens": 141201767.0,
      "step": 4054
    },
    {
      "epoch": 0.7530176415970288,
      "grad_norm": 1.4958141685652857,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8676345348358154,
      "num_tokens": 141240873.0,
      "step": 4055
    },
    {
      "epoch": 0.7532033426183844,
      "grad_norm": 1.4464723425652484,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8629746437072754,
      "num_tokens": 141276896.0,
      "step": 4056
    },
    {
      "epoch": 0.75338904363974,
      "grad_norm": 1.5235248449985657,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8594171404838562,
      "num_tokens": 141312245.0,
      "step": 4057
    },
    {
      "epoch": 0.7535747446610956,
      "grad_norm": 1.438246343243519,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8593846559524536,
      "num_tokens": 141350599.0,
      "step": 4058
    },
    {
      "epoch": 0.7537604456824513,
      "grad_norm": 1.5370710760766484,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8601135015487671,
      "num_tokens": 141386834.0,
      "step": 4059
    },
    {
      "epoch": 0.7539461467038069,
      "grad_norm": 1.5535978681989728,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8684334754943848,
      "num_tokens": 141422527.0,
      "step": 4060
    },
    {
      "epoch": 0.7541318477251625,
      "grad_norm": 1.4424574086430906,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8767228126525879,
      "num_tokens": 141458048.0,
      "step": 4061
    },
    {
      "epoch": 0.7543175487465181,
      "grad_norm": 1.6552097918527506,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8717690706253052,
      "num_tokens": 141487472.0,
      "step": 4062
    },
    {
      "epoch": 0.7545032497678738,
      "grad_norm": 1.3820646301017618,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8709836602210999,
      "num_tokens": 141523803.0,
      "step": 4063
    },
    {
      "epoch": 0.7546889507892294,
      "grad_norm": 1.545236383905775,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8658066391944885,
      "num_tokens": 141558390.0,
      "step": 4064
    },
    {
      "epoch": 0.754874651810585,
      "grad_norm": 1.5517279152694197,
      "learning_rate": 1e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8626892566680908,
      "num_tokens": 141594528.0,
      "step": 4065
    },
    {
      "epoch": 0.7550603528319406,
      "grad_norm": 1.6880918147528359,
      "learning_rate": 1e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8634769916534424,
      "num_tokens": 141622608.0,
      "step": 4066
    },
    {
      "epoch": 0.7552460538532962,
      "grad_norm": 1.6445223016830264,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8693752884864807,
      "num_tokens": 141650944.0,
      "step": 4067
    },
    {
      "epoch": 0.7554317548746519,
      "grad_norm": 1.4962793635763407,
      "learning_rate": 1e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.858057975769043,
      "num_tokens": 141684842.0,
      "step": 4068
    },
    {
      "epoch": 0.7556174558960074,
      "grad_norm": 1.7015110218417109,
      "learning_rate": 1e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8443354368209839,
      "num_tokens": 141716135.0,
      "step": 4069
    },
    {
      "epoch": 0.755803156917363,
      "grad_norm": 1.4885209238285533,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8756859302520752,
      "num_tokens": 141750726.0,
      "step": 4070
    },
    {
      "epoch": 0.7559888579387186,
      "grad_norm": 1.3916870508708488,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8833726644515991,
      "num_tokens": 141785946.0,
      "step": 4071
    },
    {
      "epoch": 0.7561745589600742,
      "grad_norm": 1.549331189803562,
      "learning_rate": 1e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.854824960231781,
      "num_tokens": 141821080.0,
      "step": 4072
    },
    {
      "epoch": 0.7563602599814299,
      "grad_norm": 1.5460112013166323,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8616178035736084,
      "num_tokens": 141855451.0,
      "step": 4073
    },
    {
      "epoch": 0.7565459610027855,
      "grad_norm": 1.5229927683949822,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8699911832809448,
      "num_tokens": 141887752.0,
      "step": 4074
    },
    {
      "epoch": 0.7567316620241411,
      "grad_norm": 1.5641252011585656,
      "learning_rate": 1e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8587126731872559,
      "num_tokens": 141919880.0,
      "step": 4075
    },
    {
      "epoch": 0.7569173630454967,
      "grad_norm": 1.5030085271344762,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8843182921409607,
      "num_tokens": 141948881.0,
      "step": 4076
    },
    {
      "epoch": 0.7571030640668523,
      "grad_norm": 1.3415940075828607,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8740947246551514,
      "num_tokens": 141987824.0,
      "step": 4077
    },
    {
      "epoch": 0.757288765088208,
      "grad_norm": 1.4028752628373602,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8705008029937744,
      "num_tokens": 142023566.0,
      "step": 4078
    },
    {
      "epoch": 0.7574744661095636,
      "grad_norm": 1.534536499085568,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8576129674911499,
      "num_tokens": 142056276.0,
      "step": 4079
    },
    {
      "epoch": 0.7576601671309192,
      "grad_norm": 1.621941369211538,
      "learning_rate": 1e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8449928760528564,
      "num_tokens": 142087745.0,
      "step": 4080
    },
    {
      "epoch": 0.7578458681522748,
      "grad_norm": 1.4777940127090987,
      "learning_rate": 1e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8629266023635864,
      "num_tokens": 142125175.0,
      "step": 4081
    },
    {
      "epoch": 0.7580315691736305,
      "grad_norm": 1.4478642429323791,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8705846071243286,
      "num_tokens": 142158223.0,
      "step": 4082
    },
    {
      "epoch": 0.7582172701949861,
      "grad_norm": 1.5220716843402085,
      "learning_rate": 1e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8622673749923706,
      "num_tokens": 142191276.0,
      "step": 4083
    },
    {
      "epoch": 0.7584029712163417,
      "grad_norm": 1.5062840638414654,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8683135509490967,
      "num_tokens": 142227342.0,
      "step": 4084
    },
    {
      "epoch": 0.7585886722376973,
      "grad_norm": 1.5384015562918296,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8715461492538452,
      "num_tokens": 142261722.0,
      "step": 4085
    },
    {
      "epoch": 0.758774373259053,
      "grad_norm": 1.4872138987287895,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8735904693603516,
      "num_tokens": 142298432.0,
      "step": 4086
    },
    {
      "epoch": 0.7589600742804086,
      "grad_norm": 1.4407992594284382,
      "learning_rate": 1e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8601422309875488,
      "num_tokens": 142336065.0,
      "step": 4087
    },
    {
      "epoch": 0.7591457753017642,
      "grad_norm": 1.5038065506924811,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8752208948135376,
      "num_tokens": 142365425.0,
      "step": 4088
    },
    {
      "epoch": 0.7593314763231198,
      "grad_norm": 1.4673132475552961,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8632628321647644,
      "num_tokens": 142399752.0,
      "step": 4089
    },
    {
      "epoch": 0.7595171773444754,
      "grad_norm": 1.5024450042567397,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8697206974029541,
      "num_tokens": 142435108.0,
      "step": 4090
    },
    {
      "epoch": 0.7597028783658311,
      "grad_norm": 1.454206624529475,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8613455295562744,
      "num_tokens": 142473600.0,
      "step": 4091
    },
    {
      "epoch": 0.7598885793871867,
      "grad_norm": 1.6403673827991425,
      "learning_rate": 1e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8513649702072144,
      "num_tokens": 142505759.0,
      "step": 4092
    },
    {
      "epoch": 0.7600742804085423,
      "grad_norm": 1.479079719039177,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8720583915710449,
      "num_tokens": 142542199.0,
      "step": 4093
    },
    {
      "epoch": 0.7602599814298978,
      "grad_norm": 1.3430261224349618,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8818411827087402,
      "num_tokens": 142577131.0,
      "step": 4094
    },
    {
      "epoch": 0.7604456824512534,
      "grad_norm": 1.4777726681603003,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8665651082992554,
      "num_tokens": 142613832.0,
      "step": 4095
    },
    {
      "epoch": 0.7606313834726091,
      "grad_norm": 1.4203401991438436,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8729573488235474,
      "num_tokens": 142647218.0,
      "step": 4096
    },
    {
      "epoch": 0.7608170844939647,
      "grad_norm": 1.4227007828260843,
      "learning_rate": 1e-06,
      "loss": 0.3826,
      "mean_token_accuracy": 0.8719004392623901,
      "num_tokens": 142684436.0,
      "step": 4097
    },
    {
      "epoch": 0.7610027855153203,
      "grad_norm": 1.453427218179501,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8783186674118042,
      "num_tokens": 142721660.0,
      "step": 4098
    },
    {
      "epoch": 0.7611884865366759,
      "grad_norm": 1.5545231640176644,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8746304512023926,
      "num_tokens": 142750624.0,
      "step": 4099
    },
    {
      "epoch": 0.7613741875580315,
      "grad_norm": 1.412889259160609,
      "learning_rate": 1e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8600546717643738,
      "num_tokens": 142788466.0,
      "step": 4100
    },
    {
      "epoch": 0.7615598885793872,
      "grad_norm": 1.6118293114789495,
      "learning_rate": 1e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8470785617828369,
      "num_tokens": 142821554.0,
      "step": 4101
    },
    {
      "epoch": 0.7617455896007428,
      "grad_norm": 1.4037381640079991,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8575145602226257,
      "num_tokens": 142861502.0,
      "step": 4102
    },
    {
      "epoch": 0.7619312906220984,
      "grad_norm": 1.373883251445467,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8681684136390686,
      "num_tokens": 142899541.0,
      "step": 4103
    },
    {
      "epoch": 0.762116991643454,
      "grad_norm": 1.5735373664225858,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8764323592185974,
      "num_tokens": 142927081.0,
      "step": 4104
    },
    {
      "epoch": 0.7623026926648097,
      "grad_norm": 1.502231404078274,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8672637939453125,
      "num_tokens": 142960348.0,
      "step": 4105
    },
    {
      "epoch": 0.7624883936861653,
      "grad_norm": 1.6386040262487331,
      "learning_rate": 1e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8429296016693115,
      "num_tokens": 142990660.0,
      "step": 4106
    },
    {
      "epoch": 0.7626740947075209,
      "grad_norm": 1.4495336811594406,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8747411370277405,
      "num_tokens": 143025682.0,
      "step": 4107
    },
    {
      "epoch": 0.7628597957288765,
      "grad_norm": 1.5472986638149393,
      "learning_rate": 1e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8526811599731445,
      "num_tokens": 143058505.0,
      "step": 4108
    },
    {
      "epoch": 0.7630454967502321,
      "grad_norm": 1.4538962321159876,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8617233633995056,
      "num_tokens": 143095501.0,
      "step": 4109
    },
    {
      "epoch": 0.7632311977715878,
      "grad_norm": 1.4261076397728523,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8684297204017639,
      "num_tokens": 143130213.0,
      "step": 4110
    },
    {
      "epoch": 0.7634168987929434,
      "grad_norm": 1.4344401055566842,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8634055256843567,
      "num_tokens": 143168690.0,
      "step": 4111
    },
    {
      "epoch": 0.763602599814299,
      "grad_norm": 1.4458422437965386,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.864023745059967,
      "num_tokens": 143203257.0,
      "step": 4112
    },
    {
      "epoch": 0.7637883008356546,
      "grad_norm": 1.5036271952450688,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8719222545623779,
      "num_tokens": 143235639.0,
      "step": 4113
    },
    {
      "epoch": 0.7639740018570103,
      "grad_norm": 1.4358131680724133,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8759452104568481,
      "num_tokens": 143268453.0,
      "step": 4114
    },
    {
      "epoch": 0.7641597028783659,
      "grad_norm": 1.435826411711114,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8789699077606201,
      "num_tokens": 143300558.0,
      "step": 4115
    },
    {
      "epoch": 0.7643454038997215,
      "grad_norm": 1.5397734622743344,
      "learning_rate": 1e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.858292818069458,
      "num_tokens": 143333101.0,
      "step": 4116
    },
    {
      "epoch": 0.7645311049210771,
      "grad_norm": 1.3961359283002286,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8547870516777039,
      "num_tokens": 143375034.0,
      "step": 4117
    },
    {
      "epoch": 0.7647168059424326,
      "grad_norm": 1.4545679592614689,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8763869404792786,
      "num_tokens": 143407859.0,
      "step": 4118
    },
    {
      "epoch": 0.7649025069637883,
      "grad_norm": 1.552675251280725,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8706800937652588,
      "num_tokens": 143440917.0,
      "step": 4119
    },
    {
      "epoch": 0.7650882079851439,
      "grad_norm": 1.6514311699903437,
      "learning_rate": 1e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8523381948471069,
      "num_tokens": 143471143.0,
      "step": 4120
    },
    {
      "epoch": 0.7652739090064995,
      "grad_norm": 1.4622406841700422,
      "learning_rate": 1e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8644428253173828,
      "num_tokens": 143509797.0,
      "step": 4121
    },
    {
      "epoch": 0.7654596100278551,
      "grad_norm": 1.4766357762455105,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8647783398628235,
      "num_tokens": 143544748.0,
      "step": 4122
    },
    {
      "epoch": 0.7656453110492107,
      "grad_norm": 1.501258225227022,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8764413595199585,
      "num_tokens": 143576775.0,
      "step": 4123
    },
    {
      "epoch": 0.7658310120705664,
      "grad_norm": 1.580082241701523,
      "learning_rate": 1e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8605770468711853,
      "num_tokens": 143612816.0,
      "step": 4124
    },
    {
      "epoch": 0.766016713091922,
      "grad_norm": 1.5907076009959498,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8740624189376831,
      "num_tokens": 143642732.0,
      "step": 4125
    },
    {
      "epoch": 0.7662024141132776,
      "grad_norm": 1.5106569109948556,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8689640760421753,
      "num_tokens": 143673907.0,
      "step": 4126
    },
    {
      "epoch": 0.7663881151346332,
      "grad_norm": 1.4733223180642028,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8691573143005371,
      "num_tokens": 143705407.0,
      "step": 4127
    },
    {
      "epoch": 0.7665738161559889,
      "grad_norm": 1.4266023088965523,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8812695145606995,
      "num_tokens": 143740295.0,
      "step": 4128
    },
    {
      "epoch": 0.7667595171773445,
      "grad_norm": 1.3596109188016512,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8806896209716797,
      "num_tokens": 143778359.0,
      "step": 4129
    },
    {
      "epoch": 0.7669452181987001,
      "grad_norm": 1.5336219442158614,
      "learning_rate": 1e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.857616126537323,
      "num_tokens": 143813329.0,
      "step": 4130
    },
    {
      "epoch": 0.7671309192200557,
      "grad_norm": 1.7568241922684378,
      "learning_rate": 1e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.862541675567627,
      "num_tokens": 143849116.0,
      "step": 4131
    },
    {
      "epoch": 0.7673166202414113,
      "grad_norm": 1.52546177949058,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8704326748847961,
      "num_tokens": 143881696.0,
      "step": 4132
    },
    {
      "epoch": 0.767502321262767,
      "grad_norm": 1.6899046948794558,
      "learning_rate": 1e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.8560584783554077,
      "num_tokens": 143910131.0,
      "step": 4133
    },
    {
      "epoch": 0.7676880222841226,
      "grad_norm": 1.452382888645947,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8803728818893433,
      "num_tokens": 143940473.0,
      "step": 4134
    },
    {
      "epoch": 0.7678737233054782,
      "grad_norm": 1.550482057326346,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8720036149024963,
      "num_tokens": 143971990.0,
      "step": 4135
    },
    {
      "epoch": 0.7680594243268338,
      "grad_norm": 1.5669166990729906,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8720107078552246,
      "num_tokens": 144002347.0,
      "step": 4136
    },
    {
      "epoch": 0.7682451253481895,
      "grad_norm": 1.5197557297977422,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8772425055503845,
      "num_tokens": 144035506.0,
      "step": 4137
    },
    {
      "epoch": 0.7684308263695451,
      "grad_norm": 1.6392720195383206,
      "learning_rate": 1e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8569158911705017,
      "num_tokens": 144067358.0,
      "step": 4138
    },
    {
      "epoch": 0.7686165273909007,
      "grad_norm": 1.5086638331404594,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8647277355194092,
      "num_tokens": 144104772.0,
      "step": 4139
    },
    {
      "epoch": 0.7688022284122563,
      "grad_norm": 1.5419748620186304,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8647487163543701,
      "num_tokens": 144138588.0,
      "step": 4140
    },
    {
      "epoch": 0.7689879294336119,
      "grad_norm": 1.4511378194115743,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8696142435073853,
      "num_tokens": 144174237.0,
      "step": 4141
    },
    {
      "epoch": 0.7691736304549674,
      "grad_norm": 1.4791867639185898,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8748580813407898,
      "num_tokens": 144208893.0,
      "step": 4142
    },
    {
      "epoch": 0.7693593314763231,
      "grad_norm": 1.5874068928574447,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8688290119171143,
      "num_tokens": 144238441.0,
      "step": 4143
    },
    {
      "epoch": 0.7695450324976787,
      "grad_norm": 1.5822347745126182,
      "learning_rate": 1e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8567651510238647,
      "num_tokens": 144273007.0,
      "step": 4144
    },
    {
      "epoch": 0.7697307335190343,
      "grad_norm": 1.5320909698290393,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8752697110176086,
      "num_tokens": 144303601.0,
      "step": 4145
    },
    {
      "epoch": 0.7699164345403899,
      "grad_norm": 1.5280287000942276,
      "learning_rate": 1e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8570089936256409,
      "num_tokens": 144332964.0,
      "step": 4146
    },
    {
      "epoch": 0.7701021355617456,
      "grad_norm": 1.5104307475369907,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8579257726669312,
      "num_tokens": 144366496.0,
      "step": 4147
    },
    {
      "epoch": 0.7702878365831012,
      "grad_norm": 1.547427860857208,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8703099489212036,
      "num_tokens": 144398592.0,
      "step": 4148
    },
    {
      "epoch": 0.7704735376044568,
      "grad_norm": 1.4955143912270636,
      "learning_rate": 1e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8587766885757446,
      "num_tokens": 144435739.0,
      "step": 4149
    },
    {
      "epoch": 0.7706592386258124,
      "grad_norm": 1.4076716696661147,
      "learning_rate": 1e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8575091361999512,
      "num_tokens": 144478764.0,
      "step": 4150
    },
    {
      "epoch": 0.770844939647168,
      "grad_norm": 1.5485563486230003,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.865780770778656,
      "num_tokens": 144508045.0,
      "step": 4151
    },
    {
      "epoch": 0.7710306406685237,
      "grad_norm": 1.5046687866722228,
      "learning_rate": 1e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8601930141448975,
      "num_tokens": 144540610.0,
      "step": 4152
    },
    {
      "epoch": 0.7712163416898793,
      "grad_norm": 1.5686257272466673,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8588102459907532,
      "num_tokens": 144570543.0,
      "step": 4153
    },
    {
      "epoch": 0.7714020427112349,
      "grad_norm": 1.4355797588391195,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8659958839416504,
      "num_tokens": 144606947.0,
      "step": 4154
    },
    {
      "epoch": 0.7715877437325905,
      "grad_norm": 1.4046512602201353,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8565055727958679,
      "num_tokens": 144649320.0,
      "step": 4155
    },
    {
      "epoch": 0.7717734447539462,
      "grad_norm": 1.5556780644740529,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8649952411651611,
      "num_tokens": 144681464.0,
      "step": 4156
    },
    {
      "epoch": 0.7719591457753018,
      "grad_norm": 1.4712747556575654,
      "learning_rate": 1e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.859664261341095,
      "num_tokens": 144716494.0,
      "step": 4157
    },
    {
      "epoch": 0.7721448467966574,
      "grad_norm": 1.4307487945936916,
      "learning_rate": 1e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8668172955513,
      "num_tokens": 144759375.0,
      "step": 4158
    },
    {
      "epoch": 0.772330547818013,
      "grad_norm": 1.4860438501053688,
      "learning_rate": 1e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8631976842880249,
      "num_tokens": 144794048.0,
      "step": 4159
    },
    {
      "epoch": 0.7725162488393686,
      "grad_norm": 1.4952206672353743,
      "learning_rate": 1e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8513146638870239,
      "num_tokens": 144829135.0,
      "step": 4160
    },
    {
      "epoch": 0.7727019498607243,
      "grad_norm": 1.3219623973607957,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8616814613342285,
      "num_tokens": 144874400.0,
      "step": 4161
    },
    {
      "epoch": 0.7728876508820799,
      "grad_norm": 1.5996951675941837,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8739190101623535,
      "num_tokens": 144912548.0,
      "step": 4162
    },
    {
      "epoch": 0.7730733519034355,
      "grad_norm": 1.530618995452552,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8672482967376709,
      "num_tokens": 144943562.0,
      "step": 4163
    },
    {
      "epoch": 0.7732590529247911,
      "grad_norm": 1.5878450920676583,
      "learning_rate": 1e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8544988036155701,
      "num_tokens": 144976997.0,
      "step": 4164
    },
    {
      "epoch": 0.7734447539461468,
      "grad_norm": 1.4232285090884709,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8672707080841064,
      "num_tokens": 145012855.0,
      "step": 4165
    },
    {
      "epoch": 0.7736304549675023,
      "grad_norm": 1.4776858477980914,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8624570965766907,
      "num_tokens": 145047424.0,
      "step": 4166
    },
    {
      "epoch": 0.7738161559888579,
      "grad_norm": 1.5541916719020632,
      "learning_rate": 1e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8556684255599976,
      "num_tokens": 145078704.0,
      "step": 4167
    },
    {
      "epoch": 0.7740018570102135,
      "grad_norm": 1.6351154220587099,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8684460520744324,
      "num_tokens": 145105436.0,
      "step": 4168
    },
    {
      "epoch": 0.7741875580315691,
      "grad_norm": 1.5906903669126868,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8762261271476746,
      "num_tokens": 145132220.0,
      "step": 4169
    },
    {
      "epoch": 0.7743732590529248,
      "grad_norm": 1.5707507544179635,
      "learning_rate": 1e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.863806962966919,
      "num_tokens": 145163073.0,
      "step": 4170
    },
    {
      "epoch": 0.7745589600742804,
      "grad_norm": 1.4803634528792962,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8721117377281189,
      "num_tokens": 145198412.0,
      "step": 4171
    },
    {
      "epoch": 0.774744661095636,
      "grad_norm": 1.7384924672866073,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8705407381057739,
      "num_tokens": 145225877.0,
      "step": 4172
    },
    {
      "epoch": 0.7749303621169916,
      "grad_norm": 1.521233292447511,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8674056529998779,
      "num_tokens": 145258229.0,
      "step": 4173
    },
    {
      "epoch": 0.7751160631383472,
      "grad_norm": 1.6035413899528266,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8537744879722595,
      "num_tokens": 145290497.0,
      "step": 4174
    },
    {
      "epoch": 0.7753017641597029,
      "grad_norm": 1.4853644001217106,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8704571723937988,
      "num_tokens": 145331640.0,
      "step": 4175
    },
    {
      "epoch": 0.7754874651810585,
      "grad_norm": 1.467108986174747,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8665604591369629,
      "num_tokens": 145366586.0,
      "step": 4176
    },
    {
      "epoch": 0.7756731662024141,
      "grad_norm": 1.5620658921238915,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8632861375808716,
      "num_tokens": 145401193.0,
      "step": 4177
    },
    {
      "epoch": 0.7758588672237697,
      "grad_norm": 1.5775057368657532,
      "learning_rate": 1e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8660688996315002,
      "num_tokens": 145430507.0,
      "step": 4178
    },
    {
      "epoch": 0.7760445682451254,
      "grad_norm": 1.4842457404372131,
      "learning_rate": 1e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8598310947418213,
      "num_tokens": 145471378.0,
      "step": 4179
    },
    {
      "epoch": 0.776230269266481,
      "grad_norm": 1.4502205641295416,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8755759000778198,
      "num_tokens": 145507446.0,
      "step": 4180
    },
    {
      "epoch": 0.7764159702878366,
      "grad_norm": 1.4693052885415832,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8693596124649048,
      "num_tokens": 145541311.0,
      "step": 4181
    },
    {
      "epoch": 0.7766016713091922,
      "grad_norm": 1.6084158590458113,
      "learning_rate": 1e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8604757785797119,
      "num_tokens": 145573213.0,
      "step": 4182
    },
    {
      "epoch": 0.7767873723305478,
      "grad_norm": 1.5808481948676574,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8707906603813171,
      "num_tokens": 145607764.0,
      "step": 4183
    },
    {
      "epoch": 0.7769730733519035,
      "grad_norm": 1.5529865381365882,
      "learning_rate": 1e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8543024063110352,
      "num_tokens": 145646242.0,
      "step": 4184
    },
    {
      "epoch": 0.7771587743732591,
      "grad_norm": 1.5042285262022752,
      "learning_rate": 1e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8412780165672302,
      "num_tokens": 145683304.0,
      "step": 4185
    },
    {
      "epoch": 0.7773444753946147,
      "grad_norm": 1.4948005588978808,
      "learning_rate": 1e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8525083065032959,
      "num_tokens": 145720282.0,
      "step": 4186
    },
    {
      "epoch": 0.7775301764159703,
      "grad_norm": 1.4974648516707358,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8533475399017334,
      "num_tokens": 145759777.0,
      "step": 4187
    },
    {
      "epoch": 0.777715877437326,
      "grad_norm": 1.458481850256568,
      "learning_rate": 1e-06,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8678804636001587,
      "num_tokens": 145798359.0,
      "step": 4188
    },
    {
      "epoch": 0.7779015784586816,
      "grad_norm": 1.6433612717910413,
      "learning_rate": 1e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8512299060821533,
      "num_tokens": 145832693.0,
      "step": 4189
    },
    {
      "epoch": 0.7780872794800371,
      "grad_norm": 1.4277582557143509,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.864128589630127,
      "num_tokens": 145871979.0,
      "step": 4190
    },
    {
      "epoch": 0.7782729805013927,
      "grad_norm": 1.709857637352912,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.858666181564331,
      "num_tokens": 145898247.0,
      "step": 4191
    },
    {
      "epoch": 0.7784586815227483,
      "grad_norm": 1.5747540322645277,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8771746754646301,
      "num_tokens": 145927691.0,
      "step": 4192
    },
    {
      "epoch": 0.778644382544104,
      "grad_norm": 1.6580248907026096,
      "learning_rate": 1e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8488380312919617,
      "num_tokens": 145961989.0,
      "step": 4193
    },
    {
      "epoch": 0.7788300835654596,
      "grad_norm": 1.4111297212783598,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8756136894226074,
      "num_tokens": 145998193.0,
      "step": 4194
    },
    {
      "epoch": 0.7790157845868152,
      "grad_norm": 1.393099511051301,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8608471155166626,
      "num_tokens": 146036064.0,
      "step": 4195
    },
    {
      "epoch": 0.7792014856081708,
      "grad_norm": 1.5299745417654427,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8651292324066162,
      "num_tokens": 146069052.0,
      "step": 4196
    },
    {
      "epoch": 0.7793871866295264,
      "grad_norm": 1.5751304249899456,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8850993514060974,
      "num_tokens": 146094667.0,
      "step": 4197
    },
    {
      "epoch": 0.7795728876508821,
      "grad_norm": 1.5393514994597621,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8512195944786072,
      "num_tokens": 146133983.0,
      "step": 4198
    },
    {
      "epoch": 0.7797585886722377,
      "grad_norm": 1.5117596695657098,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8603665828704834,
      "num_tokens": 146167454.0,
      "step": 4199
    },
    {
      "epoch": 0.7799442896935933,
      "grad_norm": 1.583884520603899,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8708133697509766,
      "num_tokens": 146199037.0,
      "step": 4200
    },
    {
      "epoch": 0.7801299907149489,
      "grad_norm": 1.6459253903744753,
      "learning_rate": 1e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8572738170623779,
      "num_tokens": 146228623.0,
      "step": 4201
    },
    {
      "epoch": 0.7803156917363046,
      "grad_norm": 1.5520807585858196,
      "learning_rate": 1e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8600320816040039,
      "num_tokens": 146263348.0,
      "step": 4202
    },
    {
      "epoch": 0.7805013927576602,
      "grad_norm": 1.5290362503333432,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8722648620605469,
      "num_tokens": 146294506.0,
      "step": 4203
    },
    {
      "epoch": 0.7806870937790158,
      "grad_norm": 1.530425491086965,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8763221502304077,
      "num_tokens": 146326340.0,
      "step": 4204
    },
    {
      "epoch": 0.7808727948003714,
      "grad_norm": 1.606826484409298,
      "learning_rate": 1e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8553941249847412,
      "num_tokens": 146360118.0,
      "step": 4205
    },
    {
      "epoch": 0.781058495821727,
      "grad_norm": 1.447206861217036,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8676369190216064,
      "num_tokens": 146395486.0,
      "step": 4206
    },
    {
      "epoch": 0.7812441968430827,
      "grad_norm": 1.554659865228068,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8609944581985474,
      "num_tokens": 146427655.0,
      "step": 4207
    },
    {
      "epoch": 0.7814298978644383,
      "grad_norm": 1.434860074853938,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8649697303771973,
      "num_tokens": 146464584.0,
      "step": 4208
    },
    {
      "epoch": 0.7816155988857939,
      "grad_norm": 1.4897322497201235,
      "learning_rate": 1e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8684226274490356,
      "num_tokens": 146501344.0,
      "step": 4209
    },
    {
      "epoch": 0.7818012999071495,
      "grad_norm": 1.563680096464976,
      "learning_rate": 1e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8463267087936401,
      "num_tokens": 146541277.0,
      "step": 4210
    },
    {
      "epoch": 0.7819870009285051,
      "grad_norm": 1.5686180386911515,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8625601530075073,
      "num_tokens": 146571765.0,
      "step": 4211
    },
    {
      "epoch": 0.7821727019498608,
      "grad_norm": 1.4162140826401655,
      "learning_rate": 1e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.8608170747756958,
      "num_tokens": 146611675.0,
      "step": 4212
    },
    {
      "epoch": 0.7823584029712164,
      "grad_norm": 1.3539335628368874,
      "learning_rate": 1e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8541263341903687,
      "num_tokens": 146652978.0,
      "step": 4213
    },
    {
      "epoch": 0.7825441039925719,
      "grad_norm": 1.4753108923737275,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8744933605194092,
      "num_tokens": 146688020.0,
      "step": 4214
    },
    {
      "epoch": 0.7827298050139275,
      "grad_norm": 1.4974990874577094,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8648039102554321,
      "num_tokens": 146720612.0,
      "step": 4215
    },
    {
      "epoch": 0.7829155060352831,
      "grad_norm": 1.4520720710413408,
      "learning_rate": 1e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8601466417312622,
      "num_tokens": 146755782.0,
      "step": 4216
    },
    {
      "epoch": 0.7831012070566388,
      "grad_norm": 1.4286819484752897,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8706626892089844,
      "num_tokens": 146792924.0,
      "step": 4217
    },
    {
      "epoch": 0.7832869080779944,
      "grad_norm": 1.5755760213985435,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.858535885810852,
      "num_tokens": 146823868.0,
      "step": 4218
    },
    {
      "epoch": 0.78347260909935,
      "grad_norm": 1.551982542809076,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8666520118713379,
      "num_tokens": 146856847.0,
      "step": 4219
    },
    {
      "epoch": 0.7836583101207056,
      "grad_norm": 1.3676439950920467,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8770861625671387,
      "num_tokens": 146891721.0,
      "step": 4220
    },
    {
      "epoch": 0.7838440111420613,
      "grad_norm": 1.3774522664938862,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8796138763427734,
      "num_tokens": 146927976.0,
      "step": 4221
    },
    {
      "epoch": 0.7840297121634169,
      "grad_norm": 1.4436281753911266,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8630134463310242,
      "num_tokens": 146964774.0,
      "step": 4222
    },
    {
      "epoch": 0.7842154131847725,
      "grad_norm": 1.6141743690734403,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8720021843910217,
      "num_tokens": 146995907.0,
      "step": 4223
    },
    {
      "epoch": 0.7844011142061281,
      "grad_norm": 1.4394615326456197,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.865572988986969,
      "num_tokens": 147034610.0,
      "step": 4224
    },
    {
      "epoch": 0.7845868152274837,
      "grad_norm": 1.421387902320641,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.877008318901062,
      "num_tokens": 147067157.0,
      "step": 4225
    },
    {
      "epoch": 0.7847725162488394,
      "grad_norm": 1.4253904420197934,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8648813962936401,
      "num_tokens": 147103985.0,
      "step": 4226
    },
    {
      "epoch": 0.784958217270195,
      "grad_norm": 1.5126130346930715,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8833555579185486,
      "num_tokens": 147137496.0,
      "step": 4227
    },
    {
      "epoch": 0.7851439182915506,
      "grad_norm": 1.410343745947405,
      "learning_rate": 1e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.866229772567749,
      "num_tokens": 147173983.0,
      "step": 4228
    },
    {
      "epoch": 0.7853296193129062,
      "grad_norm": 1.482525316228513,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8743544816970825,
      "num_tokens": 147207204.0,
      "step": 4229
    },
    {
      "epoch": 0.7855153203342619,
      "grad_norm": 1.3235512698663487,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8741793632507324,
      "num_tokens": 147247315.0,
      "step": 4230
    },
    {
      "epoch": 0.7857010213556175,
      "grad_norm": 1.4695760294712894,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8717707991600037,
      "num_tokens": 147280636.0,
      "step": 4231
    },
    {
      "epoch": 0.7858867223769731,
      "grad_norm": 1.4304857214600724,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8697639107704163,
      "num_tokens": 147317987.0,
      "step": 4232
    },
    {
      "epoch": 0.7860724233983287,
      "grad_norm": 1.4156546397914596,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8765571117401123,
      "num_tokens": 147355572.0,
      "step": 4233
    },
    {
      "epoch": 0.7862581244196843,
      "grad_norm": 1.4978698712435734,
      "learning_rate": 1e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8583099842071533,
      "num_tokens": 147390732.0,
      "step": 4234
    },
    {
      "epoch": 0.78644382544104,
      "grad_norm": 1.4595650352567466,
      "learning_rate": 1e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8560453653335571,
      "num_tokens": 147428476.0,
      "step": 4235
    },
    {
      "epoch": 0.7866295264623956,
      "grad_norm": 1.5330634122528977,
      "learning_rate": 1e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8569965362548828,
      "num_tokens": 147463597.0,
      "step": 4236
    },
    {
      "epoch": 0.7868152274837512,
      "grad_norm": 1.7230086708147303,
      "learning_rate": 1e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8544949889183044,
      "num_tokens": 147492696.0,
      "step": 4237
    },
    {
      "epoch": 0.7870009285051067,
      "grad_norm": 1.6348730596358958,
      "learning_rate": 1e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8564624786376953,
      "num_tokens": 147523291.0,
      "step": 4238
    },
    {
      "epoch": 0.7871866295264623,
      "grad_norm": 1.443815688833087,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8670451641082764,
      "num_tokens": 147560175.0,
      "step": 4239
    },
    {
      "epoch": 0.787372330547818,
      "grad_norm": 1.5605621235544513,
      "learning_rate": 1e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8425942659378052,
      "num_tokens": 147594322.0,
      "step": 4240
    },
    {
      "epoch": 0.7875580315691736,
      "grad_norm": 1.3739090999729866,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8748534917831421,
      "num_tokens": 147632103.0,
      "step": 4241
    },
    {
      "epoch": 0.7877437325905292,
      "grad_norm": 1.4592293845732234,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8745472431182861,
      "num_tokens": 147665875.0,
      "step": 4242
    },
    {
      "epoch": 0.7879294336118848,
      "grad_norm": 1.4904672542239932,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8716947436332703,
      "num_tokens": 147699406.0,
      "step": 4243
    },
    {
      "epoch": 0.7881151346332405,
      "grad_norm": 1.3573151991974113,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.859347939491272,
      "num_tokens": 147739793.0,
      "step": 4244
    },
    {
      "epoch": 0.7883008356545961,
      "grad_norm": 1.4171340983320397,
      "learning_rate": 1e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8652029037475586,
      "num_tokens": 147775184.0,
      "step": 4245
    },
    {
      "epoch": 0.7884865366759517,
      "grad_norm": 1.311321036940728,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8705180287361145,
      "num_tokens": 147818024.0,
      "step": 4246
    },
    {
      "epoch": 0.7886722376973073,
      "grad_norm": 1.524032399784139,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8716415166854858,
      "num_tokens": 147851003.0,
      "step": 4247
    },
    {
      "epoch": 0.7888579387186629,
      "grad_norm": 1.5612838381401415,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8640478253364563,
      "num_tokens": 147883652.0,
      "step": 4248
    },
    {
      "epoch": 0.7890436397400186,
      "grad_norm": 1.5556655818825713,
      "learning_rate": 1e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8524382710456848,
      "num_tokens": 147917716.0,
      "step": 4249
    },
    {
      "epoch": 0.7892293407613742,
      "grad_norm": 1.4371486014860038,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8743878602981567,
      "num_tokens": 147952219.0,
      "step": 4250
    },
    {
      "epoch": 0.7894150417827298,
      "grad_norm": 1.9065037074668036,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8825851678848267,
      "num_tokens": 147983046.0,
      "step": 4251
    },
    {
      "epoch": 0.7896007428040854,
      "grad_norm": 1.464385490762679,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8673442602157593,
      "num_tokens": 148017070.0,
      "step": 4252
    },
    {
      "epoch": 0.789786443825441,
      "grad_norm": 1.5523380207495205,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8631672263145447,
      "num_tokens": 148048475.0,
      "step": 4253
    },
    {
      "epoch": 0.7899721448467967,
      "grad_norm": 1.5191284073263969,
      "learning_rate": 1e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8573737144470215,
      "num_tokens": 148084126.0,
      "step": 4254
    },
    {
      "epoch": 0.7901578458681523,
      "grad_norm": 1.4603099565989515,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8674473762512207,
      "num_tokens": 148118718.0,
      "step": 4255
    },
    {
      "epoch": 0.7903435468895079,
      "grad_norm": 1.4639704225713248,
      "learning_rate": 1e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8506146669387817,
      "num_tokens": 148154577.0,
      "step": 4256
    },
    {
      "epoch": 0.7905292479108635,
      "grad_norm": 1.368217588050421,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8741777539253235,
      "num_tokens": 148195012.0,
      "step": 4257
    },
    {
      "epoch": 0.7907149489322192,
      "grad_norm": 1.4518270983443506,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8664796352386475,
      "num_tokens": 148228942.0,
      "step": 4258
    },
    {
      "epoch": 0.7909006499535748,
      "grad_norm": 1.5462043317542495,
      "learning_rate": 1e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8565140962600708,
      "num_tokens": 148263147.0,
      "step": 4259
    },
    {
      "epoch": 0.7910863509749304,
      "grad_norm": 1.360842567449257,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8658041954040527,
      "num_tokens": 148304494.0,
      "step": 4260
    },
    {
      "epoch": 0.791272051996286,
      "grad_norm": 1.444244033831251,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8570458292961121,
      "num_tokens": 148339733.0,
      "step": 4261
    },
    {
      "epoch": 0.7914577530176417,
      "grad_norm": 1.4759060539649116,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8689256906509399,
      "num_tokens": 148372131.0,
      "step": 4262
    },
    {
      "epoch": 0.7916434540389972,
      "grad_norm": 1.414738366737303,
      "learning_rate": 1e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8592601418495178,
      "num_tokens": 148411989.0,
      "step": 4263
    },
    {
      "epoch": 0.7918291550603528,
      "grad_norm": 1.485717962315931,
      "learning_rate": 1e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8619430065155029,
      "num_tokens": 148451443.0,
      "step": 4264
    },
    {
      "epoch": 0.7920148560817084,
      "grad_norm": 1.5243459889012119,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8846217393875122,
      "num_tokens": 148480406.0,
      "step": 4265
    },
    {
      "epoch": 0.792200557103064,
      "grad_norm": 1.48204859295902,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8655494451522827,
      "num_tokens": 148512644.0,
      "step": 4266
    },
    {
      "epoch": 0.7923862581244197,
      "grad_norm": 1.5588650001935733,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8634363412857056,
      "num_tokens": 148547114.0,
      "step": 4267
    },
    {
      "epoch": 0.7925719591457753,
      "grad_norm": 1.5088733860185952,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8819250464439392,
      "num_tokens": 148579702.0,
      "step": 4268
    },
    {
      "epoch": 0.7927576601671309,
      "grad_norm": 1.385690485924935,
      "learning_rate": 1e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.867306113243103,
      "num_tokens": 148616746.0,
      "step": 4269
    },
    {
      "epoch": 0.7929433611884865,
      "grad_norm": 1.4419255429600764,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8787939548492432,
      "num_tokens": 148652413.0,
      "step": 4270
    },
    {
      "epoch": 0.7931290622098421,
      "grad_norm": 1.5153633886885498,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8617557883262634,
      "num_tokens": 148689104.0,
      "step": 4271
    },
    {
      "epoch": 0.7933147632311978,
      "grad_norm": 1.3566774916978366,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8810269236564636,
      "num_tokens": 148725433.0,
      "step": 4272
    },
    {
      "epoch": 0.7935004642525534,
      "grad_norm": 1.5308018875065945,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.871123194694519,
      "num_tokens": 148757681.0,
      "step": 4273
    },
    {
      "epoch": 0.793686165273909,
      "grad_norm": 1.6169377763238844,
      "learning_rate": 1e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8623876571655273,
      "num_tokens": 148787351.0,
      "step": 4274
    },
    {
      "epoch": 0.7938718662952646,
      "grad_norm": 1.4992546589968205,
      "learning_rate": 1e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.855505108833313,
      "num_tokens": 148821518.0,
      "step": 4275
    },
    {
      "epoch": 0.7940575673166202,
      "grad_norm": 1.4009365506774858,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8748301863670349,
      "num_tokens": 148858657.0,
      "step": 4276
    },
    {
      "epoch": 0.7942432683379759,
      "grad_norm": 1.412257685759186,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.868047833442688,
      "num_tokens": 148894453.0,
      "step": 4277
    },
    {
      "epoch": 0.7944289693593315,
      "grad_norm": 1.3910403579976622,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8704497814178467,
      "num_tokens": 148928487.0,
      "step": 4278
    },
    {
      "epoch": 0.7946146703806871,
      "grad_norm": 1.5906943164222964,
      "learning_rate": 1e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8471970558166504,
      "num_tokens": 148960465.0,
      "step": 4279
    },
    {
      "epoch": 0.7948003714020427,
      "grad_norm": 1.429699052596285,
      "learning_rate": 1e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8569245338439941,
      "num_tokens": 148999323.0,
      "step": 4280
    },
    {
      "epoch": 0.7949860724233984,
      "grad_norm": 1.4906389877601254,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8695018291473389,
      "num_tokens": 149035020.0,
      "step": 4281
    },
    {
      "epoch": 0.795171773444754,
      "grad_norm": 1.4247957049761515,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8741849660873413,
      "num_tokens": 149071304.0,
      "step": 4282
    },
    {
      "epoch": 0.7953574744661096,
      "grad_norm": 1.402373899509798,
      "learning_rate": 1e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8610489964485168,
      "num_tokens": 149109534.0,
      "step": 4283
    },
    {
      "epoch": 0.7955431754874652,
      "grad_norm": 1.4764032337895912,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8725372552871704,
      "num_tokens": 149137267.0,
      "step": 4284
    },
    {
      "epoch": 0.7957288765088208,
      "grad_norm": 1.5674136628735875,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8592026829719543,
      "num_tokens": 149170568.0,
      "step": 4285
    },
    {
      "epoch": 0.7959145775301765,
      "grad_norm": 1.5499442266346668,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8604118824005127,
      "num_tokens": 149202870.0,
      "step": 4286
    },
    {
      "epoch": 0.796100278551532,
      "grad_norm": 1.3667240013492838,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.870003342628479,
      "num_tokens": 149242075.0,
      "step": 4287
    },
    {
      "epoch": 0.7962859795728876,
      "grad_norm": 1.4856209047397482,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.876001238822937,
      "num_tokens": 149275587.0,
      "step": 4288
    },
    {
      "epoch": 0.7964716805942432,
      "grad_norm": 1.4327381766317009,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8785268068313599,
      "num_tokens": 149307978.0,
      "step": 4289
    },
    {
      "epoch": 0.7966573816155988,
      "grad_norm": 1.5907679837816562,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8791106343269348,
      "num_tokens": 149341075.0,
      "step": 4290
    },
    {
      "epoch": 0.7968430826369545,
      "grad_norm": 1.4606148925464726,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8658069372177124,
      "num_tokens": 149378538.0,
      "step": 4291
    },
    {
      "epoch": 0.7970287836583101,
      "grad_norm": 1.5275199407763225,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8772276639938354,
      "num_tokens": 149408883.0,
      "step": 4292
    },
    {
      "epoch": 0.7972144846796657,
      "grad_norm": 1.5711810591070485,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8706650733947754,
      "num_tokens": 149442600.0,
      "step": 4293
    },
    {
      "epoch": 0.7974001857010213,
      "grad_norm": 1.304092020383367,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8864175081253052,
      "num_tokens": 149479355.0,
      "step": 4294
    },
    {
      "epoch": 0.797585886722377,
      "grad_norm": 1.4042378450796595,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8653228282928467,
      "num_tokens": 149515830.0,
      "step": 4295
    },
    {
      "epoch": 0.7977715877437326,
      "grad_norm": 1.5963254610966986,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8756626844406128,
      "num_tokens": 149543758.0,
      "step": 4296
    },
    {
      "epoch": 0.7979572887650882,
      "grad_norm": 1.4832252940800577,
      "learning_rate": 1e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8559428453445435,
      "num_tokens": 149577005.0,
      "step": 4297
    },
    {
      "epoch": 0.7981429897864438,
      "grad_norm": 1.4525260440312746,
      "learning_rate": 1e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8534846305847168,
      "num_tokens": 149614160.0,
      "step": 4298
    },
    {
      "epoch": 0.7983286908077994,
      "grad_norm": 1.470952754274168,
      "learning_rate": 1e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8547714352607727,
      "num_tokens": 149652166.0,
      "step": 4299
    },
    {
      "epoch": 0.7985143918291551,
      "grad_norm": 1.5000823466472661,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8650545477867126,
      "num_tokens": 149688956.0,
      "step": 4300
    },
    {
      "epoch": 0.7987000928505107,
      "grad_norm": 1.5781554214882372,
      "learning_rate": 1e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8612292408943176,
      "num_tokens": 149722521.0,
      "step": 4301
    },
    {
      "epoch": 0.7988857938718663,
      "grad_norm": 1.4310150915126967,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.860774040222168,
      "num_tokens": 149764800.0,
      "step": 4302
    },
    {
      "epoch": 0.7990714948932219,
      "grad_norm": 1.662903695522589,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8646500706672668,
      "num_tokens": 149793008.0,
      "step": 4303
    },
    {
      "epoch": 0.7992571959145776,
      "grad_norm": 1.401786869609608,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8722765445709229,
      "num_tokens": 149830519.0,
      "step": 4304
    },
    {
      "epoch": 0.7994428969359332,
      "grad_norm": 1.4736905966441167,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8571661710739136,
      "num_tokens": 149865790.0,
      "step": 4305
    },
    {
      "epoch": 0.7996285979572888,
      "grad_norm": 1.5124951861076792,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8712905645370483,
      "num_tokens": 149896306.0,
      "step": 4306
    },
    {
      "epoch": 0.7998142989786444,
      "grad_norm": 1.4425389632082708,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8668955564498901,
      "num_tokens": 149934975.0,
      "step": 4307
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4952792262434103,
      "learning_rate": 1e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8461228013038635,
      "num_tokens": 149973308.0,
      "step": 4308
    },
    {
      "epoch": 0.8001857010213557,
      "grad_norm": 1.3899438043855636,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8620162010192871,
      "num_tokens": 150011980.0,
      "step": 4309
    },
    {
      "epoch": 0.8003714020427113,
      "grad_norm": 1.6203993997827673,
      "learning_rate": 1e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8515843749046326,
      "num_tokens": 150043630.0,
      "step": 4310
    },
    {
      "epoch": 0.8005571030640668,
      "grad_norm": 1.4467831973993346,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8630563616752625,
      "num_tokens": 150082143.0,
      "step": 4311
    },
    {
      "epoch": 0.8007428040854224,
      "grad_norm": 1.3558379888137575,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8710831999778748,
      "num_tokens": 150119849.0,
      "step": 4312
    },
    {
      "epoch": 0.800928505106778,
      "grad_norm": 1.432549606764197,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8645063638687134,
      "num_tokens": 150154550.0,
      "step": 4313
    },
    {
      "epoch": 0.8011142061281337,
      "grad_norm": 1.4486918621702531,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8844729661941528,
      "num_tokens": 150188103.0,
      "step": 4314
    },
    {
      "epoch": 0.8012999071494893,
      "grad_norm": 1.4711570035909871,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8901209235191345,
      "num_tokens": 150217596.0,
      "step": 4315
    },
    {
      "epoch": 0.8014856081708449,
      "grad_norm": 1.507937743428348,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8738958239555359,
      "num_tokens": 150248230.0,
      "step": 4316
    },
    {
      "epoch": 0.8016713091922005,
      "grad_norm": 1.4637322862641,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8779586553573608,
      "num_tokens": 150282941.0,
      "step": 4317
    },
    {
      "epoch": 0.8018570102135562,
      "grad_norm": 1.52881138303281,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8745638728141785,
      "num_tokens": 150315377.0,
      "step": 4318
    },
    {
      "epoch": 0.8020427112349118,
      "grad_norm": 1.46278008543093,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8751879334449768,
      "num_tokens": 150348063.0,
      "step": 4319
    },
    {
      "epoch": 0.8022284122562674,
      "grad_norm": 1.6058320385890645,
      "learning_rate": 1e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8584473133087158,
      "num_tokens": 150380306.0,
      "step": 4320
    },
    {
      "epoch": 0.802414113277623,
      "grad_norm": 1.4734799250099142,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8760879039764404,
      "num_tokens": 150414579.0,
      "step": 4321
    },
    {
      "epoch": 0.8025998142989786,
      "grad_norm": 1.4741051789752482,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8760223388671875,
      "num_tokens": 150448092.0,
      "step": 4322
    },
    {
      "epoch": 0.8027855153203343,
      "grad_norm": 1.4276161544675252,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.879245400428772,
      "num_tokens": 150487668.0,
      "step": 4323
    },
    {
      "epoch": 0.8029712163416899,
      "grad_norm": 1.3825638782765282,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8619855046272278,
      "num_tokens": 150527530.0,
      "step": 4324
    },
    {
      "epoch": 0.8031569173630455,
      "grad_norm": 1.6015993930926877,
      "learning_rate": 1e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8599323034286499,
      "num_tokens": 150558702.0,
      "step": 4325
    },
    {
      "epoch": 0.8033426183844011,
      "grad_norm": 1.3960919359385033,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8745564818382263,
      "num_tokens": 150592768.0,
      "step": 4326
    },
    {
      "epoch": 0.8035283194057568,
      "grad_norm": 1.358037203151585,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8627766370773315,
      "num_tokens": 150630683.0,
      "step": 4327
    },
    {
      "epoch": 0.8037140204271124,
      "grad_norm": 1.7174048764491394,
      "learning_rate": 1e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8557705879211426,
      "num_tokens": 150657722.0,
      "step": 4328
    },
    {
      "epoch": 0.803899721448468,
      "grad_norm": 1.598101680080136,
      "learning_rate": 1e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8543696999549866,
      "num_tokens": 150687504.0,
      "step": 4329
    },
    {
      "epoch": 0.8040854224698236,
      "grad_norm": 1.4939860760896795,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8813724517822266,
      "num_tokens": 150713732.0,
      "step": 4330
    },
    {
      "epoch": 0.8042711234911792,
      "grad_norm": 1.619829720231655,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8669574856758118,
      "num_tokens": 150739394.0,
      "step": 4331
    },
    {
      "epoch": 0.8044568245125349,
      "grad_norm": 1.6096915937604956,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8749552965164185,
      "num_tokens": 150767728.0,
      "step": 4332
    },
    {
      "epoch": 0.8046425255338905,
      "grad_norm": 1.3985281490067702,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8714114427566528,
      "num_tokens": 150805880.0,
      "step": 4333
    },
    {
      "epoch": 0.8048282265552461,
      "grad_norm": 1.5060401962224865,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.862668514251709,
      "num_tokens": 150839233.0,
      "step": 4334
    },
    {
      "epoch": 0.8050139275766016,
      "grad_norm": 1.4792248827380374,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.875366747379303,
      "num_tokens": 150871461.0,
      "step": 4335
    },
    {
      "epoch": 0.8051996285979572,
      "grad_norm": 1.42032795775082,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8678667545318604,
      "num_tokens": 150906053.0,
      "step": 4336
    },
    {
      "epoch": 0.8053853296193129,
      "grad_norm": 1.3861363231679649,
      "learning_rate": 1e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8725826740264893,
      "num_tokens": 150945393.0,
      "step": 4337
    },
    {
      "epoch": 0.8055710306406685,
      "grad_norm": 1.3613686920253918,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8748456835746765,
      "num_tokens": 150978810.0,
      "step": 4338
    },
    {
      "epoch": 0.8057567316620241,
      "grad_norm": 1.429582114923675,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8714274168014526,
      "num_tokens": 151012757.0,
      "step": 4339
    },
    {
      "epoch": 0.8059424326833797,
      "grad_norm": 1.377558379035823,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8723278045654297,
      "num_tokens": 151051850.0,
      "step": 4340
    },
    {
      "epoch": 0.8061281337047354,
      "grad_norm": 1.3447889690468866,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.868342936038971,
      "num_tokens": 151091714.0,
      "step": 4341
    },
    {
      "epoch": 0.806313834726091,
      "grad_norm": 1.3992285600556085,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8788251876831055,
      "num_tokens": 151129583.0,
      "step": 4342
    },
    {
      "epoch": 0.8064995357474466,
      "grad_norm": 1.612702626223245,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.8701827526092529,
      "num_tokens": 151156279.0,
      "step": 4343
    },
    {
      "epoch": 0.8066852367688022,
      "grad_norm": 1.378311032834907,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8718043565750122,
      "num_tokens": 151195504.0,
      "step": 4344
    },
    {
      "epoch": 0.8068709377901578,
      "grad_norm": 1.3395217179824879,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.876457929611206,
      "num_tokens": 151230496.0,
      "step": 4345
    },
    {
      "epoch": 0.8070566388115135,
      "grad_norm": 1.3605919144406196,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8780627250671387,
      "num_tokens": 151271980.0,
      "step": 4346
    },
    {
      "epoch": 0.8072423398328691,
      "grad_norm": 1.453532818483904,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8829379081726074,
      "num_tokens": 151305756.0,
      "step": 4347
    },
    {
      "epoch": 0.8074280408542247,
      "grad_norm": 1.4882824917400261,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8745418190956116,
      "num_tokens": 151337546.0,
      "step": 4348
    },
    {
      "epoch": 0.8076137418755803,
      "grad_norm": 1.699718280426996,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8715523481369019,
      "num_tokens": 151364537.0,
      "step": 4349
    },
    {
      "epoch": 0.807799442896936,
      "grad_norm": 1.534631947761036,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8639057278633118,
      "num_tokens": 151395542.0,
      "step": 4350
    },
    {
      "epoch": 0.8079851439182916,
      "grad_norm": 1.626656195914202,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8610246181488037,
      "num_tokens": 151426647.0,
      "step": 4351
    },
    {
      "epoch": 0.8081708449396472,
      "grad_norm": 1.3760776930469205,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8783925771713257,
      "num_tokens": 151461705.0,
      "step": 4352
    },
    {
      "epoch": 0.8083565459610028,
      "grad_norm": 1.3986280750309539,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8728374242782593,
      "num_tokens": 151499351.0,
      "step": 4353
    },
    {
      "epoch": 0.8085422469823584,
      "grad_norm": 1.3282573218526417,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8734467625617981,
      "num_tokens": 151541426.0,
      "step": 4354
    },
    {
      "epoch": 0.8087279480037141,
      "grad_norm": 1.381188392993607,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8612844944000244,
      "num_tokens": 151578167.0,
      "step": 4355
    },
    {
      "epoch": 0.8089136490250697,
      "grad_norm": 1.547905200560261,
      "learning_rate": 1e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8509986400604248,
      "num_tokens": 151607801.0,
      "step": 4356
    },
    {
      "epoch": 0.8090993500464253,
      "grad_norm": 1.403685244032094,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8735089302062988,
      "num_tokens": 151644282.0,
      "step": 4357
    },
    {
      "epoch": 0.8092850510677809,
      "grad_norm": 1.545457512995733,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8723891973495483,
      "num_tokens": 151675686.0,
      "step": 4358
    },
    {
      "epoch": 0.8094707520891364,
      "grad_norm": 1.6531677665656666,
      "learning_rate": 1e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8579384088516235,
      "num_tokens": 151704821.0,
      "step": 4359
    },
    {
      "epoch": 0.8096564531104921,
      "grad_norm": 1.6162190928786713,
      "learning_rate": 1e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8610129356384277,
      "num_tokens": 151737810.0,
      "step": 4360
    },
    {
      "epoch": 0.8098421541318477,
      "grad_norm": 1.6288942550002035,
      "learning_rate": 1e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8521891832351685,
      "num_tokens": 151769481.0,
      "step": 4361
    },
    {
      "epoch": 0.8100278551532033,
      "grad_norm": 1.4787059183395557,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8712245225906372,
      "num_tokens": 151803174.0,
      "step": 4362
    },
    {
      "epoch": 0.8102135561745589,
      "grad_norm": 1.412430548859068,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8867602348327637,
      "num_tokens": 151839570.0,
      "step": 4363
    },
    {
      "epoch": 0.8103992571959145,
      "grad_norm": 1.3638637372932654,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.863042950630188,
      "num_tokens": 151880675.0,
      "step": 4364
    },
    {
      "epoch": 0.8105849582172702,
      "grad_norm": 1.5893602119986874,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8584282398223877,
      "num_tokens": 151910907.0,
      "step": 4365
    },
    {
      "epoch": 0.8107706592386258,
      "grad_norm": 1.3958033880534444,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8721839785575867,
      "num_tokens": 151947466.0,
      "step": 4366
    },
    {
      "epoch": 0.8109563602599814,
      "grad_norm": 1.3807991522655119,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8688406944274902,
      "num_tokens": 151984714.0,
      "step": 4367
    },
    {
      "epoch": 0.811142061281337,
      "grad_norm": 1.4633344895826919,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8808043003082275,
      "num_tokens": 152019980.0,
      "step": 4368
    },
    {
      "epoch": 0.8113277623026927,
      "grad_norm": 1.333966805347666,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8668236136436462,
      "num_tokens": 152059601.0,
      "step": 4369
    },
    {
      "epoch": 0.8115134633240483,
      "grad_norm": 1.651299704007678,
      "learning_rate": 1e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.862615704536438,
      "num_tokens": 152092626.0,
      "step": 4370
    },
    {
      "epoch": 0.8116991643454039,
      "grad_norm": 1.33101776092851,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8884046077728271,
      "num_tokens": 152130821.0,
      "step": 4371
    },
    {
      "epoch": 0.8118848653667595,
      "grad_norm": 1.5578984004552923,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8744201064109802,
      "num_tokens": 152162622.0,
      "step": 4372
    },
    {
      "epoch": 0.8120705663881151,
      "grad_norm": 1.5012300312520312,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8628213405609131,
      "num_tokens": 152198990.0,
      "step": 4373
    },
    {
      "epoch": 0.8122562674094708,
      "grad_norm": 1.5381857679383215,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8750361800193787,
      "num_tokens": 152229487.0,
      "step": 4374
    },
    {
      "epoch": 0.8124419684308264,
      "grad_norm": 1.6262194154701197,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8703407645225525,
      "num_tokens": 152257937.0,
      "step": 4375
    },
    {
      "epoch": 0.812627669452182,
      "grad_norm": 1.5850289098579708,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8651334047317505,
      "num_tokens": 152288026.0,
      "step": 4376
    },
    {
      "epoch": 0.8128133704735376,
      "grad_norm": 1.4972550913713332,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8682245016098022,
      "num_tokens": 152324377.0,
      "step": 4377
    },
    {
      "epoch": 0.8129990714948933,
      "grad_norm": 1.4628998228123802,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8729051947593689,
      "num_tokens": 152357795.0,
      "step": 4378
    },
    {
      "epoch": 0.8131847725162489,
      "grad_norm": 1.2913390185782156,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.866807222366333,
      "num_tokens": 152400086.0,
      "step": 4379
    },
    {
      "epoch": 0.8133704735376045,
      "grad_norm": 1.4953629927969498,
      "learning_rate": 1e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.851625382900238,
      "num_tokens": 152436060.0,
      "step": 4380
    },
    {
      "epoch": 0.8135561745589601,
      "grad_norm": 1.3806212706872636,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8682976961135864,
      "num_tokens": 152476329.0,
      "step": 4381
    },
    {
      "epoch": 0.8137418755803157,
      "grad_norm": 1.4400739626192178,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8670446276664734,
      "num_tokens": 152512282.0,
      "step": 4382
    },
    {
      "epoch": 0.8139275766016713,
      "grad_norm": 1.2857986326785615,
      "learning_rate": 1e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8666896820068359,
      "num_tokens": 152552402.0,
      "step": 4383
    },
    {
      "epoch": 0.8141132776230269,
      "grad_norm": 1.4599576441520479,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8654046058654785,
      "num_tokens": 152588661.0,
      "step": 4384
    },
    {
      "epoch": 0.8142989786443825,
      "grad_norm": 1.4633057926208586,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8718273639678955,
      "num_tokens": 152627046.0,
      "step": 4385
    },
    {
      "epoch": 0.8144846796657381,
      "grad_norm": 1.5514933160286686,
      "learning_rate": 1e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.841369092464447,
      "num_tokens": 152665097.0,
      "step": 4386
    },
    {
      "epoch": 0.8146703806870937,
      "grad_norm": 1.402770993088317,
      "learning_rate": 1e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8573397397994995,
      "num_tokens": 152704917.0,
      "step": 4387
    },
    {
      "epoch": 0.8148560817084494,
      "grad_norm": 1.4384902208800587,
      "learning_rate": 1e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8662353754043579,
      "num_tokens": 152739293.0,
      "step": 4388
    },
    {
      "epoch": 0.815041782729805,
      "grad_norm": 1.5125412899347597,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8735185861587524,
      "num_tokens": 152772877.0,
      "step": 4389
    },
    {
      "epoch": 0.8152274837511606,
      "grad_norm": 1.4946101658371367,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8582381010055542,
      "num_tokens": 152809216.0,
      "step": 4390
    },
    {
      "epoch": 0.8154131847725162,
      "grad_norm": 1.4396659318258538,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8599020838737488,
      "num_tokens": 152848927.0,
      "step": 4391
    },
    {
      "epoch": 0.8155988857938719,
      "grad_norm": 1.4890391810294192,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8677902221679688,
      "num_tokens": 152881745.0,
      "step": 4392
    },
    {
      "epoch": 0.8157845868152275,
      "grad_norm": 1.3535746950947871,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8697372674942017,
      "num_tokens": 152921478.0,
      "step": 4393
    },
    {
      "epoch": 0.8159702878365831,
      "grad_norm": 1.511911912566589,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8655416369438171,
      "num_tokens": 152953935.0,
      "step": 4394
    },
    {
      "epoch": 0.8161559888579387,
      "grad_norm": 1.6285288783403409,
      "learning_rate": 1e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8449624180793762,
      "num_tokens": 152985792.0,
      "step": 4395
    },
    {
      "epoch": 0.8163416898792943,
      "grad_norm": 1.4959213542888243,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8691904544830322,
      "num_tokens": 153016307.0,
      "step": 4396
    },
    {
      "epoch": 0.81652739090065,
      "grad_norm": 1.4768135920050423,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8723759651184082,
      "num_tokens": 153048270.0,
      "step": 4397
    },
    {
      "epoch": 0.8167130919220056,
      "grad_norm": 1.5164373469460692,
      "learning_rate": 1e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8580014705657959,
      "num_tokens": 153084147.0,
      "step": 4398
    },
    {
      "epoch": 0.8168987929433612,
      "grad_norm": 1.5542789122196374,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8629328012466431,
      "num_tokens": 153118962.0,
      "step": 4399
    },
    {
      "epoch": 0.8170844939647168,
      "grad_norm": 1.541387242507276,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.874785840511322,
      "num_tokens": 153146367.0,
      "step": 4400
    },
    {
      "epoch": 0.8172701949860725,
      "grad_norm": 1.4543938724964924,
      "learning_rate": 1e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8464445471763611,
      "num_tokens": 153186615.0,
      "step": 4401
    },
    {
      "epoch": 0.8174558960074281,
      "grad_norm": 1.4296575741956967,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8779587149620056,
      "num_tokens": 153219437.0,
      "step": 4402
    },
    {
      "epoch": 0.8176415970287837,
      "grad_norm": 1.4705615214039784,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8825085163116455,
      "num_tokens": 153251707.0,
      "step": 4403
    },
    {
      "epoch": 0.8178272980501393,
      "grad_norm": 1.6461085571276295,
      "learning_rate": 1e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8582502603530884,
      "num_tokens": 153280783.0,
      "step": 4404
    },
    {
      "epoch": 0.8180129990714949,
      "grad_norm": 1.431313877278771,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8648108243942261,
      "num_tokens": 153318392.0,
      "step": 4405
    },
    {
      "epoch": 0.8181987000928506,
      "grad_norm": 1.5329700713321173,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8698763847351074,
      "num_tokens": 153348514.0,
      "step": 4406
    },
    {
      "epoch": 0.8183844011142062,
      "grad_norm": 1.3842995327938656,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8705093860626221,
      "num_tokens": 153384724.0,
      "step": 4407
    },
    {
      "epoch": 0.8185701021355617,
      "grad_norm": 1.5005170969112953,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8633791208267212,
      "num_tokens": 153418845.0,
      "step": 4408
    },
    {
      "epoch": 0.8187558031569173,
      "grad_norm": 1.5321816639312371,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8613961935043335,
      "num_tokens": 153452463.0,
      "step": 4409
    },
    {
      "epoch": 0.8189415041782729,
      "grad_norm": 1.5006502906395527,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8615850210189819,
      "num_tokens": 153488241.0,
      "step": 4410
    },
    {
      "epoch": 0.8191272051996286,
      "grad_norm": 1.488021501263596,
      "learning_rate": 1e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8558382987976074,
      "num_tokens": 153525573.0,
      "step": 4411
    },
    {
      "epoch": 0.8193129062209842,
      "grad_norm": 1.4818100780140444,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8636569976806641,
      "num_tokens": 153559230.0,
      "step": 4412
    },
    {
      "epoch": 0.8194986072423398,
      "grad_norm": 1.348278450878133,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8634030818939209,
      "num_tokens": 153604167.0,
      "step": 4413
    },
    {
      "epoch": 0.8196843082636954,
      "grad_norm": 1.6385636076997017,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8598377704620361,
      "num_tokens": 153637271.0,
      "step": 4414
    },
    {
      "epoch": 0.819870009285051,
      "grad_norm": 1.5400686981298242,
      "learning_rate": 1e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8574961423873901,
      "num_tokens": 153670125.0,
      "step": 4415
    },
    {
      "epoch": 0.8200557103064067,
      "grad_norm": 1.3969717263971704,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8711297512054443,
      "num_tokens": 153705940.0,
      "step": 4416
    },
    {
      "epoch": 0.8202414113277623,
      "grad_norm": 1.555123988155609,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8571114540100098,
      "num_tokens": 153740731.0,
      "step": 4417
    },
    {
      "epoch": 0.8204271123491179,
      "grad_norm": 1.4720916662859058,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.867178201675415,
      "num_tokens": 153776232.0,
      "step": 4418
    },
    {
      "epoch": 0.8206128133704735,
      "grad_norm": 1.6683381366983687,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8720129728317261,
      "num_tokens": 153807874.0,
      "step": 4419
    },
    {
      "epoch": 0.8207985143918292,
      "grad_norm": 1.575910515417612,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8677090406417847,
      "num_tokens": 153834662.0,
      "step": 4420
    },
    {
      "epoch": 0.8209842154131848,
      "grad_norm": 1.415292754815036,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8696364164352417,
      "num_tokens": 153874950.0,
      "step": 4421
    },
    {
      "epoch": 0.8211699164345404,
      "grad_norm": 1.4565703917352373,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8741564750671387,
      "num_tokens": 153910424.0,
      "step": 4422
    },
    {
      "epoch": 0.821355617455896,
      "grad_norm": 1.4909564466180518,
      "learning_rate": 1e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8544407486915588,
      "num_tokens": 153945980.0,
      "step": 4423
    },
    {
      "epoch": 0.8215413184772516,
      "grad_norm": 1.387870576284454,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8775486350059509,
      "num_tokens": 153980430.0,
      "step": 4424
    },
    {
      "epoch": 0.8217270194986073,
      "grad_norm": 1.7082409611029363,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8688474893569946,
      "num_tokens": 154006774.0,
      "step": 4425
    },
    {
      "epoch": 0.8219127205199629,
      "grad_norm": 1.5438798236251536,
      "learning_rate": 1e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8561835885047913,
      "num_tokens": 154041354.0,
      "step": 4426
    },
    {
      "epoch": 0.8220984215413185,
      "grad_norm": 1.506097438893023,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8635621070861816,
      "num_tokens": 154078498.0,
      "step": 4427
    },
    {
      "epoch": 0.8222841225626741,
      "grad_norm": 1.5395303809860166,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8599153161048889,
      "num_tokens": 154111859.0,
      "step": 4428
    },
    {
      "epoch": 0.8224698235840298,
      "grad_norm": 1.5318529963670975,
      "learning_rate": 1e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.8683887124061584,
      "num_tokens": 154143791.0,
      "step": 4429
    },
    {
      "epoch": 0.8226555246053854,
      "grad_norm": 1.5120474314078707,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8845480680465698,
      "num_tokens": 154174870.0,
      "step": 4430
    },
    {
      "epoch": 0.822841225626741,
      "grad_norm": 1.4704313879120319,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8755375146865845,
      "num_tokens": 154210978.0,
      "step": 4431
    },
    {
      "epoch": 0.8230269266480965,
      "grad_norm": 1.6946278534246026,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8552782535552979,
      "num_tokens": 154241858.0,
      "step": 4432
    },
    {
      "epoch": 0.8232126276694521,
      "grad_norm": 1.7045462897443273,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8577611446380615,
      "num_tokens": 154268887.0,
      "step": 4433
    },
    {
      "epoch": 0.8233983286908078,
      "grad_norm": 1.4154491305596515,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8722890019416809,
      "num_tokens": 154301662.0,
      "step": 4434
    },
    {
      "epoch": 0.8235840297121634,
      "grad_norm": 1.4985368205150242,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.86521315574646,
      "num_tokens": 154334721.0,
      "step": 4435
    },
    {
      "epoch": 0.823769730733519,
      "grad_norm": 1.5834089530760953,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8883705139160156,
      "num_tokens": 154364241.0,
      "step": 4436
    },
    {
      "epoch": 0.8239554317548746,
      "grad_norm": 1.378028853132957,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8754702806472778,
      "num_tokens": 154403939.0,
      "step": 4437
    },
    {
      "epoch": 0.8241411327762302,
      "grad_norm": 1.364538162328377,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8725532293319702,
      "num_tokens": 154441200.0,
      "step": 4438
    },
    {
      "epoch": 0.8243268337975859,
      "grad_norm": 1.4459636006660246,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8870773315429688,
      "num_tokens": 154472194.0,
      "step": 4439
    },
    {
      "epoch": 0.8245125348189415,
      "grad_norm": 1.5000843616217956,
      "learning_rate": 1e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.853451132774353,
      "num_tokens": 154509287.0,
      "step": 4440
    },
    {
      "epoch": 0.8246982358402971,
      "grad_norm": 1.544899068153838,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8703305721282959,
      "num_tokens": 154536416.0,
      "step": 4441
    },
    {
      "epoch": 0.8248839368616527,
      "grad_norm": 1.5812652197844412,
      "learning_rate": 1e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8595988750457764,
      "num_tokens": 154566772.0,
      "step": 4442
    },
    {
      "epoch": 0.8250696378830084,
      "grad_norm": 1.5606978639461526,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8652404546737671,
      "num_tokens": 154599528.0,
      "step": 4443
    },
    {
      "epoch": 0.825255338904364,
      "grad_norm": 1.4198043731355328,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8692963123321533,
      "num_tokens": 154637177.0,
      "step": 4444
    },
    {
      "epoch": 0.8254410399257196,
      "grad_norm": 1.46052687214874,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8678162097930908,
      "num_tokens": 154671470.0,
      "step": 4445
    },
    {
      "epoch": 0.8256267409470752,
      "grad_norm": 1.4370538015484335,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8659311532974243,
      "num_tokens": 154705908.0,
      "step": 4446
    },
    {
      "epoch": 0.8258124419684308,
      "grad_norm": 1.714710624262084,
      "learning_rate": 1e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8529390096664429,
      "num_tokens": 154734917.0,
      "step": 4447
    },
    {
      "epoch": 0.8259981429897865,
      "grad_norm": 1.4615910083248413,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8761979937553406,
      "num_tokens": 154770697.0,
      "step": 4448
    },
    {
      "epoch": 0.8261838440111421,
      "grad_norm": 1.4393726747043443,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8725343942642212,
      "num_tokens": 154806689.0,
      "step": 4449
    },
    {
      "epoch": 0.8263695450324977,
      "grad_norm": 1.4666363756195377,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8757083415985107,
      "num_tokens": 154841730.0,
      "step": 4450
    },
    {
      "epoch": 0.8265552460538533,
      "grad_norm": 1.5039588900640848,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8698663711547852,
      "num_tokens": 154873074.0,
      "step": 4451
    },
    {
      "epoch": 0.826740947075209,
      "grad_norm": 1.3103580328437758,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8815388679504395,
      "num_tokens": 154913443.0,
      "step": 4452
    },
    {
      "epoch": 0.8269266480965646,
      "grad_norm": 1.416641636446052,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8752288818359375,
      "num_tokens": 154952150.0,
      "step": 4453
    },
    {
      "epoch": 0.8271123491179202,
      "grad_norm": 1.4136849599080108,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8612365126609802,
      "num_tokens": 154989535.0,
      "step": 4454
    },
    {
      "epoch": 0.8272980501392758,
      "grad_norm": 1.4406479049983987,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8731781244277954,
      "num_tokens": 155025872.0,
      "step": 4455
    },
    {
      "epoch": 0.8274837511606313,
      "grad_norm": 1.5527269526728626,
      "learning_rate": 1e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8693854808807373,
      "num_tokens": 155056333.0,
      "step": 4456
    },
    {
      "epoch": 0.827669452181987,
      "grad_norm": 1.3983761268211192,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8815708160400391,
      "num_tokens": 155093067.0,
      "step": 4457
    },
    {
      "epoch": 0.8278551532033426,
      "grad_norm": 1.3710582805507927,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8676143884658813,
      "num_tokens": 155130467.0,
      "step": 4458
    },
    {
      "epoch": 0.8280408542246982,
      "grad_norm": 1.5186869237034695,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8661022186279297,
      "num_tokens": 155165241.0,
      "step": 4459
    },
    {
      "epoch": 0.8282265552460538,
      "grad_norm": 1.5554618617040399,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8776834607124329,
      "num_tokens": 155197574.0,
      "step": 4460
    },
    {
      "epoch": 0.8284122562674094,
      "grad_norm": 1.496605745821884,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8710179328918457,
      "num_tokens": 155232102.0,
      "step": 4461
    },
    {
      "epoch": 0.8285979572887651,
      "grad_norm": 1.3897672114452644,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8759453296661377,
      "num_tokens": 155269670.0,
      "step": 4462
    },
    {
      "epoch": 0.8287836583101207,
      "grad_norm": 1.5547073080063352,
      "learning_rate": 1e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8512975573539734,
      "num_tokens": 155306335.0,
      "step": 4463
    },
    {
      "epoch": 0.8289693593314763,
      "grad_norm": 1.3905159054737815,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.88214111328125,
      "num_tokens": 155344087.0,
      "step": 4464
    },
    {
      "epoch": 0.8291550603528319,
      "grad_norm": 1.6778565366242544,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.861822783946991,
      "num_tokens": 155372503.0,
      "step": 4465
    },
    {
      "epoch": 0.8293407613741876,
      "grad_norm": 1.5549107426942066,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8676261901855469,
      "num_tokens": 155403466.0,
      "step": 4466
    },
    {
      "epoch": 0.8295264623955432,
      "grad_norm": 1.6213993424013529,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8650134801864624,
      "num_tokens": 155433645.0,
      "step": 4467
    },
    {
      "epoch": 0.8297121634168988,
      "grad_norm": 1.4366719943105497,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8693168759346008,
      "num_tokens": 155468244.0,
      "step": 4468
    },
    {
      "epoch": 0.8298978644382544,
      "grad_norm": 1.417913047249457,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8654873371124268,
      "num_tokens": 155506537.0,
      "step": 4469
    },
    {
      "epoch": 0.83008356545961,
      "grad_norm": 1.5365250708867875,
      "learning_rate": 1e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8509717583656311,
      "num_tokens": 155542334.0,
      "step": 4470
    },
    {
      "epoch": 0.8302692664809657,
      "grad_norm": 1.528758598630075,
      "learning_rate": 1e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8566211462020874,
      "num_tokens": 155576409.0,
      "step": 4471
    },
    {
      "epoch": 0.8304549675023213,
      "grad_norm": 1.5148924161161204,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8683046102523804,
      "num_tokens": 155611404.0,
      "step": 4472
    },
    {
      "epoch": 0.8306406685236769,
      "grad_norm": 1.3793328599585237,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8812109231948853,
      "num_tokens": 155644803.0,
      "step": 4473
    },
    {
      "epoch": 0.8308263695450325,
      "grad_norm": 1.3663387741228328,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8784427046775818,
      "num_tokens": 155681835.0,
      "step": 4474
    },
    {
      "epoch": 0.8310120705663882,
      "grad_norm": 1.4773647898357951,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8650100231170654,
      "num_tokens": 155717901.0,
      "step": 4475
    },
    {
      "epoch": 0.8311977715877438,
      "grad_norm": 1.5405023654405048,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8738018274307251,
      "num_tokens": 155751183.0,
      "step": 4476
    },
    {
      "epoch": 0.8313834726090994,
      "grad_norm": 1.429885817926118,
      "learning_rate": 1e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8573700785636902,
      "num_tokens": 155792264.0,
      "step": 4477
    },
    {
      "epoch": 0.831569173630455,
      "grad_norm": 1.5380375544984604,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8712077736854553,
      "num_tokens": 155824840.0,
      "step": 4478
    },
    {
      "epoch": 0.8317548746518106,
      "grad_norm": 1.650354331720813,
      "learning_rate": 1e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8594367504119873,
      "num_tokens": 155851685.0,
      "step": 4479
    },
    {
      "epoch": 0.8319405756731661,
      "grad_norm": 1.4831660232493957,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.878775417804718,
      "num_tokens": 155884709.0,
      "step": 4480
    },
    {
      "epoch": 0.8321262766945218,
      "grad_norm": 1.4583446295836164,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8714796304702759,
      "num_tokens": 155919890.0,
      "step": 4481
    },
    {
      "epoch": 0.8323119777158774,
      "grad_norm": 1.3555856226640668,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8723325133323669,
      "num_tokens": 155959743.0,
      "step": 4482
    },
    {
      "epoch": 0.832497678737233,
      "grad_norm": 1.4203583721271833,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8759092688560486,
      "num_tokens": 155992536.0,
      "step": 4483
    },
    {
      "epoch": 0.8326833797585886,
      "grad_norm": 1.365202440751337,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8663836717605591,
      "num_tokens": 156030342.0,
      "step": 4484
    },
    {
      "epoch": 0.8328690807799443,
      "grad_norm": 1.5703521712707407,
      "learning_rate": 1e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8627959489822388,
      "num_tokens": 156061411.0,
      "step": 4485
    },
    {
      "epoch": 0.8330547818012999,
      "grad_norm": 1.4750612510484453,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.878954291343689,
      "num_tokens": 156092233.0,
      "step": 4486
    },
    {
      "epoch": 0.8332404828226555,
      "grad_norm": 1.526210072224402,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8670533895492554,
      "num_tokens": 156121421.0,
      "step": 4487
    },
    {
      "epoch": 0.8334261838440111,
      "grad_norm": 1.4155903305466815,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8792708516120911,
      "num_tokens": 156159885.0,
      "step": 4488
    },
    {
      "epoch": 0.8336118848653667,
      "grad_norm": 1.4246443876909642,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8701835870742798,
      "num_tokens": 156195924.0,
      "step": 4489
    },
    {
      "epoch": 0.8337975858867224,
      "grad_norm": 1.3686524957272628,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8716930747032166,
      "num_tokens": 156232423.0,
      "step": 4490
    },
    {
      "epoch": 0.833983286908078,
      "grad_norm": 1.6521740082361052,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8670861721038818,
      "num_tokens": 156262508.0,
      "step": 4491
    },
    {
      "epoch": 0.8341689879294336,
      "grad_norm": 1.516688658553334,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8621039986610413,
      "num_tokens": 156298493.0,
      "step": 4492
    },
    {
      "epoch": 0.8343546889507892,
      "grad_norm": 1.486431176664562,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8744466304779053,
      "num_tokens": 156332136.0,
      "step": 4493
    },
    {
      "epoch": 0.8345403899721449,
      "grad_norm": 1.4343566849088942,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8714206218719482,
      "num_tokens": 156363250.0,
      "step": 4494
    },
    {
      "epoch": 0.8347260909935005,
      "grad_norm": 1.45272101474014,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8663318157196045,
      "num_tokens": 156398878.0,
      "step": 4495
    },
    {
      "epoch": 0.8349117920148561,
      "grad_norm": 1.459517339234331,
      "learning_rate": 1e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8494997620582581,
      "num_tokens": 156437215.0,
      "step": 4496
    },
    {
      "epoch": 0.8350974930362117,
      "grad_norm": 1.4645302493519836,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8743368983268738,
      "num_tokens": 156471091.0,
      "step": 4497
    },
    {
      "epoch": 0.8352831940575673,
      "grad_norm": 1.4097414233335772,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.877482533454895,
      "num_tokens": 156506568.0,
      "step": 4498
    },
    {
      "epoch": 0.835468895078923,
      "grad_norm": 1.3558764288831466,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8693023920059204,
      "num_tokens": 156545372.0,
      "step": 4499
    },
    {
      "epoch": 0.8356545961002786,
      "grad_norm": 1.3759903207552378,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8633049726486206,
      "num_tokens": 156582523.0,
      "step": 4500
    },
    {
      "epoch": 0.8358402971216342,
      "grad_norm": 1.5553772444413414,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8656983375549316,
      "num_tokens": 156612025.0,
      "step": 4501
    },
    {
      "epoch": 0.8360259981429898,
      "grad_norm": 1.631862054796136,
      "learning_rate": 1e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8535346388816833,
      "num_tokens": 156642061.0,
      "step": 4502
    },
    {
      "epoch": 0.8362116991643455,
      "grad_norm": 1.4969328400913982,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8680099248886108,
      "num_tokens": 156675370.0,
      "step": 4503
    },
    {
      "epoch": 0.836397400185701,
      "grad_norm": 1.4832978108862824,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.878920316696167,
      "num_tokens": 156707590.0,
      "step": 4504
    },
    {
      "epoch": 0.8365831012070566,
      "grad_norm": 1.4456474601827363,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8822485208511353,
      "num_tokens": 156737780.0,
      "step": 4505
    },
    {
      "epoch": 0.8367688022284122,
      "grad_norm": 1.4787338844506146,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8677605986595154,
      "num_tokens": 156771545.0,
      "step": 4506
    },
    {
      "epoch": 0.8369545032497678,
      "grad_norm": 1.3682949538055977,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8808988332748413,
      "num_tokens": 156808589.0,
      "step": 4507
    },
    {
      "epoch": 0.8371402042711235,
      "grad_norm": 1.5542003000122577,
      "learning_rate": 1e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8585883975028992,
      "num_tokens": 156842019.0,
      "step": 4508
    },
    {
      "epoch": 0.8373259052924791,
      "grad_norm": 1.4814214719883583,
      "learning_rate": 1e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8547787070274353,
      "num_tokens": 156877068.0,
      "step": 4509
    },
    {
      "epoch": 0.8375116063138347,
      "grad_norm": 1.5356642519836776,
      "learning_rate": 1e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8698617219924927,
      "num_tokens": 156906853.0,
      "step": 4510
    },
    {
      "epoch": 0.8376973073351903,
      "grad_norm": 1.451099267752201,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8879694938659668,
      "num_tokens": 156935633.0,
      "step": 4511
    },
    {
      "epoch": 0.8378830083565459,
      "grad_norm": 1.5626732002662533,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8672515749931335,
      "num_tokens": 156966433.0,
      "step": 4512
    },
    {
      "epoch": 0.8380687093779016,
      "grad_norm": 1.573585148558969,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8804742693901062,
      "num_tokens": 156994718.0,
      "step": 4513
    },
    {
      "epoch": 0.8382544103992572,
      "grad_norm": 1.5491022907134873,
      "learning_rate": 1e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.850217342376709,
      "num_tokens": 157026101.0,
      "step": 4514
    },
    {
      "epoch": 0.8384401114206128,
      "grad_norm": 1.4933242412413716,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8644477128982544,
      "num_tokens": 157063497.0,
      "step": 4515
    },
    {
      "epoch": 0.8386258124419684,
      "grad_norm": 1.4718928445753983,
      "learning_rate": 1e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8544332981109619,
      "num_tokens": 157099929.0,
      "step": 4516
    },
    {
      "epoch": 0.838811513463324,
      "grad_norm": 1.4494697664329126,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8645321130752563,
      "num_tokens": 157135095.0,
      "step": 4517
    },
    {
      "epoch": 0.8389972144846797,
      "grad_norm": 1.632868206241178,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8635252714157104,
      "num_tokens": 157164617.0,
      "step": 4518
    },
    {
      "epoch": 0.8391829155060353,
      "grad_norm": 1.3787556581157003,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8663484454154968,
      "num_tokens": 157205745.0,
      "step": 4519
    },
    {
      "epoch": 0.8393686165273909,
      "grad_norm": 1.5056918491185196,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8786274194717407,
      "num_tokens": 157237431.0,
      "step": 4520
    },
    {
      "epoch": 0.8395543175487465,
      "grad_norm": 1.4469401123420063,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8745393753051758,
      "num_tokens": 157267372.0,
      "step": 4521
    },
    {
      "epoch": 0.8397400185701022,
      "grad_norm": 1.7228538893451804,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8743954300880432,
      "num_tokens": 157290772.0,
      "step": 4522
    },
    {
      "epoch": 0.8399257195914578,
      "grad_norm": 1.48988024048556,
      "learning_rate": 1e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8550915718078613,
      "num_tokens": 157326586.0,
      "step": 4523
    },
    {
      "epoch": 0.8401114206128134,
      "grad_norm": 1.4210472139530026,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8622369766235352,
      "num_tokens": 157363474.0,
      "step": 4524
    },
    {
      "epoch": 0.840297121634169,
      "grad_norm": 1.4501567007794078,
      "learning_rate": 1e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8636301755905151,
      "num_tokens": 157405933.0,
      "step": 4525
    },
    {
      "epoch": 0.8404828226555247,
      "grad_norm": 1.4244350918718136,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8714132905006409,
      "num_tokens": 157442457.0,
      "step": 4526
    },
    {
      "epoch": 0.8406685236768803,
      "grad_norm": 1.4175552818305315,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8852390050888062,
      "num_tokens": 157477512.0,
      "step": 4527
    },
    {
      "epoch": 0.8408542246982358,
      "grad_norm": 1.4701531499823566,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.876103937625885,
      "num_tokens": 157509167.0,
      "step": 4528
    },
    {
      "epoch": 0.8410399257195914,
      "grad_norm": 1.4331163559629596,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8676283359527588,
      "num_tokens": 157547562.0,
      "step": 4529
    },
    {
      "epoch": 0.841225626740947,
      "grad_norm": 1.581014250334979,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8704670071601868,
      "num_tokens": 157577358.0,
      "step": 4530
    },
    {
      "epoch": 0.8414113277623027,
      "grad_norm": 1.6685209469747784,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8767939805984497,
      "num_tokens": 157600102.0,
      "step": 4531
    },
    {
      "epoch": 0.8415970287836583,
      "grad_norm": 1.41623510459982,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8735848665237427,
      "num_tokens": 157640664.0,
      "step": 4532
    },
    {
      "epoch": 0.8417827298050139,
      "grad_norm": 1.6077162702798926,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8749817609786987,
      "num_tokens": 157669063.0,
      "step": 4533
    },
    {
      "epoch": 0.8419684308263695,
      "grad_norm": 1.321014053952994,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8748936653137207,
      "num_tokens": 157711219.0,
      "step": 4534
    },
    {
      "epoch": 0.8421541318477251,
      "grad_norm": 1.4822657811435191,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.871468186378479,
      "num_tokens": 157741318.0,
      "step": 4535
    },
    {
      "epoch": 0.8423398328690808,
      "grad_norm": 1.536412866501071,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8703155517578125,
      "num_tokens": 157775683.0,
      "step": 4536
    },
    {
      "epoch": 0.8425255338904364,
      "grad_norm": 1.3996661900170762,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8692646026611328,
      "num_tokens": 157815144.0,
      "step": 4537
    },
    {
      "epoch": 0.842711234911792,
      "grad_norm": 1.5197402261556114,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8691951632499695,
      "num_tokens": 157848267.0,
      "step": 4538
    },
    {
      "epoch": 0.8428969359331476,
      "grad_norm": 1.6031422123952013,
      "learning_rate": 1e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8537043333053589,
      "num_tokens": 157883693.0,
      "step": 4539
    },
    {
      "epoch": 0.8430826369545033,
      "grad_norm": 1.3501439481918986,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8797938227653503,
      "num_tokens": 157920984.0,
      "step": 4540
    },
    {
      "epoch": 0.8432683379758589,
      "grad_norm": 1.560505902905751,
      "learning_rate": 1e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8518227338790894,
      "num_tokens": 157957030.0,
      "step": 4541
    },
    {
      "epoch": 0.8434540389972145,
      "grad_norm": 1.5726350091288894,
      "learning_rate": 1e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8440150022506714,
      "num_tokens": 157990973.0,
      "step": 4542
    },
    {
      "epoch": 0.8436397400185701,
      "grad_norm": 1.508542112316207,
      "learning_rate": 1e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8584045767784119,
      "num_tokens": 158028489.0,
      "step": 4543
    },
    {
      "epoch": 0.8438254410399257,
      "grad_norm": 1.474174628510145,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8790701627731323,
      "num_tokens": 158062157.0,
      "step": 4544
    },
    {
      "epoch": 0.8440111420612814,
      "grad_norm": 1.6867462642466773,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8755666017532349,
      "num_tokens": 158093155.0,
      "step": 4545
    },
    {
      "epoch": 0.844196843082637,
      "grad_norm": 1.555185057080216,
      "learning_rate": 1e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8394168019294739,
      "num_tokens": 158133424.0,
      "step": 4546
    },
    {
      "epoch": 0.8443825441039926,
      "grad_norm": 1.3691024438555837,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8664903044700623,
      "num_tokens": 158173040.0,
      "step": 4547
    },
    {
      "epoch": 0.8445682451253482,
      "grad_norm": 1.5562965603356793,
      "learning_rate": 1e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8631081581115723,
      "num_tokens": 158204805.0,
      "step": 4548
    },
    {
      "epoch": 0.8447539461467038,
      "grad_norm": 1.4714063899263785,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8819937705993652,
      "num_tokens": 158232712.0,
      "step": 4549
    },
    {
      "epoch": 0.8449396471680595,
      "grad_norm": 1.5476502676406878,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8532184362411499,
      "num_tokens": 158267852.0,
      "step": 4550
    },
    {
      "epoch": 0.8451253481894151,
      "grad_norm": 1.383515874012346,
      "learning_rate": 1e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8568017482757568,
      "num_tokens": 158307978.0,
      "step": 4551
    },
    {
      "epoch": 0.8453110492107706,
      "grad_norm": 1.397407078522352,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8730920553207397,
      "num_tokens": 158345454.0,
      "step": 4552
    },
    {
      "epoch": 0.8454967502321262,
      "grad_norm": 2.0109625121682444,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8622714281082153,
      "num_tokens": 158384015.0,
      "step": 4553
    },
    {
      "epoch": 0.8456824512534818,
      "grad_norm": 1.6262399819522335,
      "learning_rate": 1e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8594274520874023,
      "num_tokens": 158413949.0,
      "step": 4554
    },
    {
      "epoch": 0.8458681522748375,
      "grad_norm": 1.5610860171705432,
      "learning_rate": 1e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8581719994544983,
      "num_tokens": 158447093.0,
      "step": 4555
    },
    {
      "epoch": 0.8460538532961931,
      "grad_norm": 1.4704841767080339,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8648008704185486,
      "num_tokens": 158482996.0,
      "step": 4556
    },
    {
      "epoch": 0.8462395543175487,
      "grad_norm": 1.5175000515838444,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8748401403427124,
      "num_tokens": 158513156.0,
      "step": 4557
    },
    {
      "epoch": 0.8464252553389043,
      "grad_norm": 1.5071250324798773,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.883332371711731,
      "num_tokens": 158540579.0,
      "step": 4558
    },
    {
      "epoch": 0.84661095636026,
      "grad_norm": 1.5218649879224266,
      "learning_rate": 1e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8588265180587769,
      "num_tokens": 158578585.0,
      "step": 4559
    },
    {
      "epoch": 0.8467966573816156,
      "grad_norm": 1.4820713077641972,
      "learning_rate": 1e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8501696586608887,
      "num_tokens": 158617868.0,
      "step": 4560
    },
    {
      "epoch": 0.8469823584029712,
      "grad_norm": 1.592033478493862,
      "learning_rate": 1e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8427143096923828,
      "num_tokens": 158650865.0,
      "step": 4561
    },
    {
      "epoch": 0.8471680594243268,
      "grad_norm": 1.4236250590957567,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8682743310928345,
      "num_tokens": 158689720.0,
      "step": 4562
    },
    {
      "epoch": 0.8473537604456824,
      "grad_norm": 1.3659631449565781,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8808103203773499,
      "num_tokens": 158726433.0,
      "step": 4563
    },
    {
      "epoch": 0.8475394614670381,
      "grad_norm": 1.3543767370394169,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.869250476360321,
      "num_tokens": 158767899.0,
      "step": 4564
    },
    {
      "epoch": 0.8477251624883937,
      "grad_norm": 1.6186113351199116,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.86375892162323,
      "num_tokens": 158794756.0,
      "step": 4565
    },
    {
      "epoch": 0.8479108635097493,
      "grad_norm": 1.4724027073308403,
      "learning_rate": 1e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8651676774024963,
      "num_tokens": 158828637.0,
      "step": 4566
    },
    {
      "epoch": 0.8480965645311049,
      "grad_norm": 1.5385884411340478,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.859169065952301,
      "num_tokens": 158864790.0,
      "step": 4567
    },
    {
      "epoch": 0.8482822655524606,
      "grad_norm": 1.5350378236397224,
      "learning_rate": 1e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8433499336242676,
      "num_tokens": 158901583.0,
      "step": 4568
    },
    {
      "epoch": 0.8484679665738162,
      "grad_norm": 1.4979559585799305,
      "learning_rate": 1e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8574771881103516,
      "num_tokens": 158935440.0,
      "step": 4569
    },
    {
      "epoch": 0.8486536675951718,
      "grad_norm": 1.4839605801478002,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8823292255401611,
      "num_tokens": 158966518.0,
      "step": 4570
    },
    {
      "epoch": 0.8488393686165274,
      "grad_norm": 1.4636318204703223,
      "learning_rate": 1e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8541263937950134,
      "num_tokens": 159005543.0,
      "step": 4571
    },
    {
      "epoch": 0.849025069637883,
      "grad_norm": 1.3796561319247582,
      "learning_rate": 1e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8661641478538513,
      "num_tokens": 159044474.0,
      "step": 4572
    },
    {
      "epoch": 0.8492107706592387,
      "grad_norm": 1.3909819835789314,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8786337971687317,
      "num_tokens": 159078257.0,
      "step": 4573
    },
    {
      "epoch": 0.8493964716805943,
      "grad_norm": 1.4621504691005245,
      "learning_rate": 1e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8619710206985474,
      "num_tokens": 159114258.0,
      "step": 4574
    },
    {
      "epoch": 0.8495821727019499,
      "grad_norm": 1.5674424152169808,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8510444164276123,
      "num_tokens": 159146942.0,
      "step": 4575
    },
    {
      "epoch": 0.8497678737233055,
      "grad_norm": 1.3149563565578286,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8827128410339355,
      "num_tokens": 159183403.0,
      "step": 4576
    },
    {
      "epoch": 0.849953574744661,
      "grad_norm": 1.6301869664896673,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8667072653770447,
      "num_tokens": 159215488.0,
      "step": 4577
    },
    {
      "epoch": 0.8501392757660167,
      "grad_norm": 1.3255534970689646,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8874961137771606,
      "num_tokens": 159252710.0,
      "step": 4578
    },
    {
      "epoch": 0.8503249767873723,
      "grad_norm": 1.5703548931555313,
      "learning_rate": 1e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8493908047676086,
      "num_tokens": 159283757.0,
      "step": 4579
    },
    {
      "epoch": 0.8505106778087279,
      "grad_norm": 1.4097744578166158,
      "learning_rate": 1e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8565964698791504,
      "num_tokens": 159323085.0,
      "step": 4580
    },
    {
      "epoch": 0.8506963788300835,
      "grad_norm": 1.410581489290821,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8663882613182068,
      "num_tokens": 159360318.0,
      "step": 4581
    },
    {
      "epoch": 0.8508820798514392,
      "grad_norm": 1.5375972341263398,
      "learning_rate": 1e-06,
      "loss": 0.3871,
      "mean_token_accuracy": 0.8677140474319458,
      "num_tokens": 159393397.0,
      "step": 4582
    },
    {
      "epoch": 0.8510677808727948,
      "grad_norm": 1.7792992052545633,
      "learning_rate": 1e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8638667464256287,
      "num_tokens": 159416146.0,
      "step": 4583
    },
    {
      "epoch": 0.8512534818941504,
      "grad_norm": 1.4792007667404237,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8664636611938477,
      "num_tokens": 159454075.0,
      "step": 4584
    },
    {
      "epoch": 0.851439182915506,
      "grad_norm": 1.3810161037268038,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8804727792739868,
      "num_tokens": 159488885.0,
      "step": 4585
    },
    {
      "epoch": 0.8516248839368616,
      "grad_norm": 1.4056003215066304,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8614881038665771,
      "num_tokens": 159527862.0,
      "step": 4586
    },
    {
      "epoch": 0.8518105849582173,
      "grad_norm": 1.3718355367417872,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8605756759643555,
      "num_tokens": 159568737.0,
      "step": 4587
    },
    {
      "epoch": 0.8519962859795729,
      "grad_norm": 1.3823148568836379,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8769311904907227,
      "num_tokens": 159603755.0,
      "step": 4588
    },
    {
      "epoch": 0.8521819870009285,
      "grad_norm": 1.3612046013417793,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8724106550216675,
      "num_tokens": 159640445.0,
      "step": 4589
    },
    {
      "epoch": 0.8523676880222841,
      "grad_norm": 1.6088955641847973,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8795499801635742,
      "num_tokens": 159665398.0,
      "step": 4590
    },
    {
      "epoch": 0.8525533890436398,
      "grad_norm": 1.3934905765321137,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.868010938167572,
      "num_tokens": 159705354.0,
      "step": 4591
    },
    {
      "epoch": 0.8527390900649954,
      "grad_norm": 1.5263108848359184,
      "learning_rate": 1e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8589897155761719,
      "num_tokens": 159739274.0,
      "step": 4592
    },
    {
      "epoch": 0.852924791086351,
      "grad_norm": 1.4940234937662813,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8742237091064453,
      "num_tokens": 159772079.0,
      "step": 4593
    },
    {
      "epoch": 0.8531104921077066,
      "grad_norm": 1.508369395367955,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8715686798095703,
      "num_tokens": 159804554.0,
      "step": 4594
    },
    {
      "epoch": 0.8532961931290622,
      "grad_norm": 1.5232301128102244,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8664559125900269,
      "num_tokens": 159835511.0,
      "step": 4595
    },
    {
      "epoch": 0.8534818941504179,
      "grad_norm": 1.3533865591524614,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.877328634262085,
      "num_tokens": 159872444.0,
      "step": 4596
    },
    {
      "epoch": 0.8536675951717735,
      "grad_norm": 1.5680366627824351,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8710017800331116,
      "num_tokens": 159902061.0,
      "step": 4597
    },
    {
      "epoch": 0.8538532961931291,
      "grad_norm": 1.3155775373400533,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.872578501701355,
      "num_tokens": 159939791.0,
      "step": 4598
    },
    {
      "epoch": 0.8540389972144847,
      "grad_norm": 1.520330952624818,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8746185302734375,
      "num_tokens": 159969560.0,
      "step": 4599
    },
    {
      "epoch": 0.8542246982358404,
      "grad_norm": 1.4551357430470038,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8756457567214966,
      "num_tokens": 160003550.0,
      "step": 4600
    },
    {
      "epoch": 0.8544103992571959,
      "grad_norm": 1.5253007680110064,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8779250383377075,
      "num_tokens": 160034871.0,
      "step": 4601
    },
    {
      "epoch": 0.8545961002785515,
      "grad_norm": 1.6044291198570597,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8578822016716003,
      "num_tokens": 160066591.0,
      "step": 4602
    },
    {
      "epoch": 0.8547818012999071,
      "grad_norm": 1.4577688579921233,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8661007881164551,
      "num_tokens": 160098822.0,
      "step": 4603
    },
    {
      "epoch": 0.8549675023212627,
      "grad_norm": 1.4384534208579043,
      "learning_rate": 1e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8544541597366333,
      "num_tokens": 160138679.0,
      "step": 4604
    },
    {
      "epoch": 0.8551532033426184,
      "grad_norm": 1.3477663150857122,
      "learning_rate": 1e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8672723770141602,
      "num_tokens": 160178525.0,
      "step": 4605
    },
    {
      "epoch": 0.855338904363974,
      "grad_norm": 1.4563042723472102,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8662201166152954,
      "num_tokens": 160217076.0,
      "step": 4606
    },
    {
      "epoch": 0.8555246053853296,
      "grad_norm": 1.4556815657928446,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8625837564468384,
      "num_tokens": 160252572.0,
      "step": 4607
    },
    {
      "epoch": 0.8557103064066852,
      "grad_norm": 1.5248213270836,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.874668300151825,
      "num_tokens": 160283149.0,
      "step": 4608
    },
    {
      "epoch": 0.8558960074280408,
      "grad_norm": 1.5732916288491696,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8807717561721802,
      "num_tokens": 160313646.0,
      "step": 4609
    },
    {
      "epoch": 0.8560817084493965,
      "grad_norm": 1.405121821435726,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8811453580856323,
      "num_tokens": 160350147.0,
      "step": 4610
    },
    {
      "epoch": 0.8562674094707521,
      "grad_norm": 1.4037956222596633,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8666543960571289,
      "num_tokens": 160387666.0,
      "step": 4611
    },
    {
      "epoch": 0.8564531104921077,
      "grad_norm": 1.3863312934381462,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8635921478271484,
      "num_tokens": 160425443.0,
      "step": 4612
    },
    {
      "epoch": 0.8566388115134633,
      "grad_norm": 1.4562438751365672,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.881483793258667,
      "num_tokens": 160457353.0,
      "step": 4613
    },
    {
      "epoch": 0.856824512534819,
      "grad_norm": 1.6141158678862975,
      "learning_rate": 1e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.855797290802002,
      "num_tokens": 160494259.0,
      "step": 4614
    },
    {
      "epoch": 0.8570102135561746,
      "grad_norm": 1.3894341388927844,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8709231615066528,
      "num_tokens": 160533290.0,
      "step": 4615
    },
    {
      "epoch": 0.8571959145775302,
      "grad_norm": 1.5450558057029038,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8704367280006409,
      "num_tokens": 160565373.0,
      "step": 4616
    },
    {
      "epoch": 0.8573816155988858,
      "grad_norm": 1.4115174799779022,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8821842670440674,
      "num_tokens": 160601698.0,
      "step": 4617
    },
    {
      "epoch": 0.8575673166202414,
      "grad_norm": 1.4655062732410082,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.869594395160675,
      "num_tokens": 160639446.0,
      "step": 4618
    },
    {
      "epoch": 0.8577530176415971,
      "grad_norm": 1.6319229272454971,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8667657375335693,
      "num_tokens": 160673548.0,
      "step": 4619
    },
    {
      "epoch": 0.8579387186629527,
      "grad_norm": 1.4996089122646676,
      "learning_rate": 1e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8401065468788147,
      "num_tokens": 160711975.0,
      "step": 4620
    },
    {
      "epoch": 0.8581244196843083,
      "grad_norm": 1.412698568828858,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8744995594024658,
      "num_tokens": 160747728.0,
      "step": 4621
    },
    {
      "epoch": 0.8583101207056639,
      "grad_norm": 1.3833484036770962,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8714620471000671,
      "num_tokens": 160786253.0,
      "step": 4622
    },
    {
      "epoch": 0.8584958217270195,
      "grad_norm": 1.422688130681228,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8632550239562988,
      "num_tokens": 160824678.0,
      "step": 4623
    },
    {
      "epoch": 0.8586815227483752,
      "grad_norm": 1.6501650851742415,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8557363152503967,
      "num_tokens": 160854277.0,
      "step": 4624
    },
    {
      "epoch": 0.8588672237697307,
      "grad_norm": 1.4959254078366944,
      "learning_rate": 1e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8536081314086914,
      "num_tokens": 160891421.0,
      "step": 4625
    },
    {
      "epoch": 0.8590529247910863,
      "grad_norm": 1.4825545026105544,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8753868341445923,
      "num_tokens": 160926847.0,
      "step": 4626
    },
    {
      "epoch": 0.8592386258124419,
      "grad_norm": 1.4552992137375524,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8666737675666809,
      "num_tokens": 160962588.0,
      "step": 4627
    },
    {
      "epoch": 0.8594243268337975,
      "grad_norm": 1.3930008735227224,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.867615818977356,
      "num_tokens": 161001624.0,
      "step": 4628
    },
    {
      "epoch": 0.8596100278551532,
      "grad_norm": 1.61245171469405,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8622328042984009,
      "num_tokens": 161030809.0,
      "step": 4629
    },
    {
      "epoch": 0.8597957288765088,
      "grad_norm": 1.3932384965551525,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8693555593490601,
      "num_tokens": 161065650.0,
      "step": 4630
    },
    {
      "epoch": 0.8599814298978644,
      "grad_norm": 1.3918925938046334,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8743246793746948,
      "num_tokens": 161100765.0,
      "step": 4631
    },
    {
      "epoch": 0.86016713091922,
      "grad_norm": 1.5005697704623961,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8677854537963867,
      "num_tokens": 161135650.0,
      "step": 4632
    },
    {
      "epoch": 0.8603528319405757,
      "grad_norm": 1.4358641788995754,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8706544041633606,
      "num_tokens": 161171423.0,
      "step": 4633
    },
    {
      "epoch": 0.8605385329619313,
      "grad_norm": 1.38114873553705,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8849222660064697,
      "num_tokens": 161203793.0,
      "step": 4634
    },
    {
      "epoch": 0.8607242339832869,
      "grad_norm": 1.4217062031823722,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8586544990539551,
      "num_tokens": 161240540.0,
      "step": 4635
    },
    {
      "epoch": 0.8609099350046425,
      "grad_norm": 1.4796260727355457,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8630887269973755,
      "num_tokens": 161277115.0,
      "step": 4636
    },
    {
      "epoch": 0.8610956360259981,
      "grad_norm": 1.610828994434944,
      "learning_rate": 1e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8502459526062012,
      "num_tokens": 161312503.0,
      "step": 4637
    },
    {
      "epoch": 0.8612813370473538,
      "grad_norm": 1.3550945937543095,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.8712965250015259,
      "num_tokens": 161351657.0,
      "step": 4638
    },
    {
      "epoch": 0.8614670380687094,
      "grad_norm": 1.3707340674848265,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8757243752479553,
      "num_tokens": 161389952.0,
      "step": 4639
    },
    {
      "epoch": 0.861652739090065,
      "grad_norm": 1.3475781945876715,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8901397585868835,
      "num_tokens": 161425380.0,
      "step": 4640
    },
    {
      "epoch": 0.8618384401114206,
      "grad_norm": 1.3961245871835455,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8712692260742188,
      "num_tokens": 161464313.0,
      "step": 4641
    },
    {
      "epoch": 0.8620241411327763,
      "grad_norm": 1.509763993818623,
      "learning_rate": 1e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.857131838798523,
      "num_tokens": 161502501.0,
      "step": 4642
    },
    {
      "epoch": 0.8622098421541319,
      "grad_norm": 1.6662713535919078,
      "learning_rate": 1e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8554458618164062,
      "num_tokens": 161537016.0,
      "step": 4643
    },
    {
      "epoch": 0.8623955431754875,
      "grad_norm": 1.5596060992565388,
      "learning_rate": 1e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8501420021057129,
      "num_tokens": 161574597.0,
      "step": 4644
    },
    {
      "epoch": 0.8625812441968431,
      "grad_norm": 1.5516689597998914,
      "learning_rate": 1e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8547658920288086,
      "num_tokens": 161609202.0,
      "step": 4645
    },
    {
      "epoch": 0.8627669452181987,
      "grad_norm": 1.357301873763511,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8746819496154785,
      "num_tokens": 161643763.0,
      "step": 4646
    },
    {
      "epoch": 0.8629526462395544,
      "grad_norm": 1.319591673140964,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8732348084449768,
      "num_tokens": 161683436.0,
      "step": 4647
    },
    {
      "epoch": 0.86313834726091,
      "grad_norm": 1.482125096929624,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.872054934501648,
      "num_tokens": 161713818.0,
      "step": 4648
    },
    {
      "epoch": 0.8633240482822655,
      "grad_norm": 1.3999352807379652,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8570715188980103,
      "num_tokens": 161755569.0,
      "step": 4649
    },
    {
      "epoch": 0.8635097493036211,
      "grad_norm": 1.4600768701074103,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8751266002655029,
      "num_tokens": 161791761.0,
      "step": 4650
    },
    {
      "epoch": 0.8636954503249767,
      "grad_norm": 1.4601286758684566,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8691022396087646,
      "num_tokens": 161825636.0,
      "step": 4651
    },
    {
      "epoch": 0.8638811513463324,
      "grad_norm": 1.5882566116634154,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8737926483154297,
      "num_tokens": 161853321.0,
      "step": 4652
    },
    {
      "epoch": 0.864066852367688,
      "grad_norm": 1.4734331972861616,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8624658584594727,
      "num_tokens": 161889627.0,
      "step": 4653
    },
    {
      "epoch": 0.8642525533890436,
      "grad_norm": 1.5464049590306956,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8643405437469482,
      "num_tokens": 161919336.0,
      "step": 4654
    },
    {
      "epoch": 0.8644382544103992,
      "grad_norm": 1.4321596206287137,
      "learning_rate": 1e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8562834858894348,
      "num_tokens": 161962115.0,
      "step": 4655
    },
    {
      "epoch": 0.8646239554317549,
      "grad_norm": 1.5452262758991837,
      "learning_rate": 1e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8631444573402405,
      "num_tokens": 161994620.0,
      "step": 4656
    },
    {
      "epoch": 0.8648096564531105,
      "grad_norm": 1.423969528606466,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8643007874488831,
      "num_tokens": 162034041.0,
      "step": 4657
    },
    {
      "epoch": 0.8649953574744661,
      "grad_norm": 1.3925217871983346,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8826113343238831,
      "num_tokens": 162071834.0,
      "step": 4658
    },
    {
      "epoch": 0.8651810584958217,
      "grad_norm": 1.472926143936806,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.872702956199646,
      "num_tokens": 162105627.0,
      "step": 4659
    },
    {
      "epoch": 0.8653667595171773,
      "grad_norm": 1.377930142477128,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8616327047348022,
      "num_tokens": 162146144.0,
      "step": 4660
    },
    {
      "epoch": 0.865552460538533,
      "grad_norm": 1.576702619564705,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8819645643234253,
      "num_tokens": 162173700.0,
      "step": 4661
    },
    {
      "epoch": 0.8657381615598886,
      "grad_norm": 1.5938368088473591,
      "learning_rate": 1e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8573520183563232,
      "num_tokens": 162205972.0,
      "step": 4662
    },
    {
      "epoch": 0.8659238625812442,
      "grad_norm": 1.2768625102357822,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.87466961145401,
      "num_tokens": 162246122.0,
      "step": 4663
    },
    {
      "epoch": 0.8661095636025998,
      "grad_norm": 1.3917200793323292,
      "learning_rate": 1e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8631720542907715,
      "num_tokens": 162285013.0,
      "step": 4664
    },
    {
      "epoch": 0.8662952646239555,
      "grad_norm": 1.5787756606499552,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8641356229782104,
      "num_tokens": 162315810.0,
      "step": 4665
    },
    {
      "epoch": 0.8664809656453111,
      "grad_norm": 1.4375265462655833,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8651120662689209,
      "num_tokens": 162353535.0,
      "step": 4666
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 1.4334848665073472,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8843824863433838,
      "num_tokens": 162387697.0,
      "step": 4667
    },
    {
      "epoch": 0.8668523676880223,
      "grad_norm": 1.4835708047949012,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8721721768379211,
      "num_tokens": 162417694.0,
      "step": 4668
    },
    {
      "epoch": 0.8670380687093779,
      "grad_norm": 1.471922520256018,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8857123255729675,
      "num_tokens": 162447313.0,
      "step": 4669
    },
    {
      "epoch": 0.8672237697307336,
      "grad_norm": 1.4340767261974672,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8764492869377136,
      "num_tokens": 162483930.0,
      "step": 4670
    },
    {
      "epoch": 0.8674094707520892,
      "grad_norm": 1.4947011470633038,
      "learning_rate": 1e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.86357581615448,
      "num_tokens": 162520191.0,
      "step": 4671
    },
    {
      "epoch": 0.8675951717734448,
      "grad_norm": 1.5050782562147857,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8817777037620544,
      "num_tokens": 162555096.0,
      "step": 4672
    },
    {
      "epoch": 0.8677808727948003,
      "grad_norm": 1.4014051788870352,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8813287019729614,
      "num_tokens": 162592938.0,
      "step": 4673
    },
    {
      "epoch": 0.8679665738161559,
      "grad_norm": 1.3753587660403133,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8744317293167114,
      "num_tokens": 162631105.0,
      "step": 4674
    },
    {
      "epoch": 0.8681522748375116,
      "grad_norm": 1.576099096672672,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8742731213569641,
      "num_tokens": 162659325.0,
      "step": 4675
    },
    {
      "epoch": 0.8683379758588672,
      "grad_norm": 1.447230516550522,
      "learning_rate": 1e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8612360954284668,
      "num_tokens": 162696383.0,
      "step": 4676
    },
    {
      "epoch": 0.8685236768802228,
      "grad_norm": 1.5768207428746501,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8787965774536133,
      "num_tokens": 162728268.0,
      "step": 4677
    },
    {
      "epoch": 0.8687093779015784,
      "grad_norm": 1.48850190846219,
      "learning_rate": 1e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8575819730758667,
      "num_tokens": 162764992.0,
      "step": 4678
    },
    {
      "epoch": 0.868895078922934,
      "grad_norm": 1.4565313247957985,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8771039247512817,
      "num_tokens": 162798623.0,
      "step": 4679
    },
    {
      "epoch": 0.8690807799442897,
      "grad_norm": 1.398291080068569,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8635627627372742,
      "num_tokens": 162837033.0,
      "step": 4680
    },
    {
      "epoch": 0.8692664809656453,
      "grad_norm": 1.4648063995805556,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.867797315120697,
      "num_tokens": 162871869.0,
      "step": 4681
    },
    {
      "epoch": 0.8694521819870009,
      "grad_norm": 1.4387278590719126,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8631863594055176,
      "num_tokens": 162908785.0,
      "step": 4682
    },
    {
      "epoch": 0.8696378830083565,
      "grad_norm": 1.3660398969777527,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8665955066680908,
      "num_tokens": 162946985.0,
      "step": 4683
    },
    {
      "epoch": 0.8698235840297122,
      "grad_norm": 1.449866897867474,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8673285245895386,
      "num_tokens": 162984989.0,
      "step": 4684
    },
    {
      "epoch": 0.8700092850510678,
      "grad_norm": 1.5769185698095136,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8746252059936523,
      "num_tokens": 163017098.0,
      "step": 4685
    },
    {
      "epoch": 0.8701949860724234,
      "grad_norm": 1.6215393829791604,
      "learning_rate": 1e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8560206890106201,
      "num_tokens": 163047223.0,
      "step": 4686
    },
    {
      "epoch": 0.870380687093779,
      "grad_norm": 1.42856154464892,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8637955188751221,
      "num_tokens": 163085054.0,
      "step": 4687
    },
    {
      "epoch": 0.8705663881151346,
      "grad_norm": 1.550766747444798,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.861697793006897,
      "num_tokens": 163114852.0,
      "step": 4688
    },
    {
      "epoch": 0.8707520891364903,
      "grad_norm": 1.491514513544784,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8815264701843262,
      "num_tokens": 163148045.0,
      "step": 4689
    },
    {
      "epoch": 0.8709377901578459,
      "grad_norm": 1.45419304988004,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8688132166862488,
      "num_tokens": 163182922.0,
      "step": 4690
    },
    {
      "epoch": 0.8711234911792015,
      "grad_norm": 1.5780419141217035,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.856136679649353,
      "num_tokens": 163217468.0,
      "step": 4691
    },
    {
      "epoch": 0.8713091922005571,
      "grad_norm": 1.4193805091428928,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8796951770782471,
      "num_tokens": 163251904.0,
      "step": 4692
    },
    {
      "epoch": 0.8714948932219128,
      "grad_norm": 1.4231752230371366,
      "learning_rate": 1e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8568609952926636,
      "num_tokens": 163292912.0,
      "step": 4693
    },
    {
      "epoch": 0.8716805942432684,
      "grad_norm": 1.4687024116203657,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8757379055023193,
      "num_tokens": 163326066.0,
      "step": 4694
    },
    {
      "epoch": 0.871866295264624,
      "grad_norm": 1.5338941635244487,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8676187992095947,
      "num_tokens": 163359177.0,
      "step": 4695
    },
    {
      "epoch": 0.8720519962859796,
      "grad_norm": 1.6511653524362926,
      "learning_rate": 1e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8597570657730103,
      "num_tokens": 163386702.0,
      "step": 4696
    },
    {
      "epoch": 0.8722376973073351,
      "grad_norm": 1.5657983516548626,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8674437999725342,
      "num_tokens": 163422731.0,
      "step": 4697
    },
    {
      "epoch": 0.8724233983286908,
      "grad_norm": 1.7467344375865972,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8664224147796631,
      "num_tokens": 163449988.0,
      "step": 4698
    },
    {
      "epoch": 0.8726090993500464,
      "grad_norm": 1.4830791477075727,
      "learning_rate": 1e-06,
      "loss": 0.3972,
      "mean_token_accuracy": 0.8668991923332214,
      "num_tokens": 163482732.0,
      "step": 4699
    },
    {
      "epoch": 0.872794800371402,
      "grad_norm": 1.6036050192405764,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8738813400268555,
      "num_tokens": 163513182.0,
      "step": 4700
    },
    {
      "epoch": 0.8729805013927576,
      "grad_norm": 1.51055491716169,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8688861131668091,
      "num_tokens": 163546288.0,
      "step": 4701
    },
    {
      "epoch": 0.8731662024141132,
      "grad_norm": 1.4790760817876707,
      "learning_rate": 1e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8604294061660767,
      "num_tokens": 163587516.0,
      "step": 4702
    },
    {
      "epoch": 0.8733519034354689,
      "grad_norm": 1.3279153748937227,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8674495220184326,
      "num_tokens": 163630891.0,
      "step": 4703
    },
    {
      "epoch": 0.8735376044568245,
      "grad_norm": 1.4555455399558528,
      "learning_rate": 1e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8587932586669922,
      "num_tokens": 163671849.0,
      "step": 4704
    },
    {
      "epoch": 0.8737233054781801,
      "grad_norm": 1.665308089193699,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8694306015968323,
      "num_tokens": 163698954.0,
      "step": 4705
    },
    {
      "epoch": 0.8739090064995357,
      "grad_norm": 1.4541040235356586,
      "learning_rate": 1e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8559050559997559,
      "num_tokens": 163738151.0,
      "step": 4706
    },
    {
      "epoch": 0.8740947075208914,
      "grad_norm": 1.5382010789516654,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8575521111488342,
      "num_tokens": 163774262.0,
      "step": 4707
    },
    {
      "epoch": 0.874280408542247,
      "grad_norm": 1.3117528802872471,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8804301619529724,
      "num_tokens": 163813768.0,
      "step": 4708
    },
    {
      "epoch": 0.8744661095636026,
      "grad_norm": 1.4342316508495936,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8709698915481567,
      "num_tokens": 163849862.0,
      "step": 4709
    },
    {
      "epoch": 0.8746518105849582,
      "grad_norm": 1.588131975793491,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8604687452316284,
      "num_tokens": 163878027.0,
      "step": 4710
    },
    {
      "epoch": 0.8748375116063138,
      "grad_norm": 1.537575059424595,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.85465407371521,
      "num_tokens": 163913331.0,
      "step": 4711
    },
    {
      "epoch": 0.8750232126276695,
      "grad_norm": 1.3930160345481832,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8672701120376587,
      "num_tokens": 163950833.0,
      "step": 4712
    },
    {
      "epoch": 0.8752089136490251,
      "grad_norm": 1.499311677663325,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8731386661529541,
      "num_tokens": 163979803.0,
      "step": 4713
    },
    {
      "epoch": 0.8753946146703807,
      "grad_norm": 1.4257174519686358,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8731615543365479,
      "num_tokens": 164017411.0,
      "step": 4714
    },
    {
      "epoch": 0.8755803156917363,
      "grad_norm": 1.5405281695522532,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8640162348747253,
      "num_tokens": 164047245.0,
      "step": 4715
    },
    {
      "epoch": 0.875766016713092,
      "grad_norm": 1.4440759214932892,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.866413950920105,
      "num_tokens": 164084561.0,
      "step": 4716
    },
    {
      "epoch": 0.8759517177344476,
      "grad_norm": 1.2988545535972227,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8702663779258728,
      "num_tokens": 164124484.0,
      "step": 4717
    },
    {
      "epoch": 0.8761374187558032,
      "grad_norm": 1.4553370779460126,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.880440354347229,
      "num_tokens": 164156446.0,
      "step": 4718
    },
    {
      "epoch": 0.8763231197771588,
      "grad_norm": 1.4557023184328521,
      "learning_rate": 1e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8576818704605103,
      "num_tokens": 164192142.0,
      "step": 4719
    },
    {
      "epoch": 0.8765088207985144,
      "grad_norm": 1.3235302290857756,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8791012763977051,
      "num_tokens": 164231625.0,
      "step": 4720
    },
    {
      "epoch": 0.87669452181987,
      "grad_norm": 1.4516696813773722,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8678125143051147,
      "num_tokens": 164268019.0,
      "step": 4721
    },
    {
      "epoch": 0.8768802228412256,
      "grad_norm": 1.4337346395390187,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.877968430519104,
      "num_tokens": 164303597.0,
      "step": 4722
    },
    {
      "epoch": 0.8770659238625812,
      "grad_norm": 1.7512424676197922,
      "learning_rate": 1e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8480799794197083,
      "num_tokens": 164330926.0,
      "step": 4723
    },
    {
      "epoch": 0.8772516248839368,
      "grad_norm": 1.39180520502405,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8698515892028809,
      "num_tokens": 164368072.0,
      "step": 4724
    },
    {
      "epoch": 0.8774373259052924,
      "grad_norm": 1.3472835938956345,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8643869161605835,
      "num_tokens": 164406676.0,
      "step": 4725
    },
    {
      "epoch": 0.8776230269266481,
      "grad_norm": 1.4106837880860963,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8719513416290283,
      "num_tokens": 164442583.0,
      "step": 4726
    },
    {
      "epoch": 0.8778087279480037,
      "grad_norm": 1.6235286099094894,
      "learning_rate": 1e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8636622428894043,
      "num_tokens": 164474754.0,
      "step": 4727
    },
    {
      "epoch": 0.8779944289693593,
      "grad_norm": 1.228043063635264,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8755820989608765,
      "num_tokens": 164519125.0,
      "step": 4728
    },
    {
      "epoch": 0.8781801299907149,
      "grad_norm": 1.7147252717474344,
      "learning_rate": 1e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8495543003082275,
      "num_tokens": 164550598.0,
      "step": 4729
    },
    {
      "epoch": 0.8783658310120706,
      "grad_norm": 1.5012866072549853,
      "learning_rate": 1e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8592569828033447,
      "num_tokens": 164589044.0,
      "step": 4730
    },
    {
      "epoch": 0.8785515320334262,
      "grad_norm": 1.5807340999309516,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8646987676620483,
      "num_tokens": 164619800.0,
      "step": 4731
    },
    {
      "epoch": 0.8787372330547818,
      "grad_norm": 1.4883753088056333,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8660790324211121,
      "num_tokens": 164653737.0,
      "step": 4732
    },
    {
      "epoch": 0.8789229340761374,
      "grad_norm": 1.5253271679800087,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8654229640960693,
      "num_tokens": 164685434.0,
      "step": 4733
    },
    {
      "epoch": 0.879108635097493,
      "grad_norm": 1.314486208610829,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8766813278198242,
      "num_tokens": 164725271.0,
      "step": 4734
    },
    {
      "epoch": 0.8792943361188487,
      "grad_norm": 1.4706808935373172,
      "learning_rate": 1e-06,
      "loss": 0.3948,
      "mean_token_accuracy": 0.865203857421875,
      "num_tokens": 164761546.0,
      "step": 4735
    },
    {
      "epoch": 0.8794800371402043,
      "grad_norm": 1.4674948300704096,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8686268329620361,
      "num_tokens": 164794334.0,
      "step": 4736
    },
    {
      "epoch": 0.8796657381615599,
      "grad_norm": 1.6993376796761688,
      "learning_rate": 1e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8627547025680542,
      "num_tokens": 164823256.0,
      "step": 4737
    },
    {
      "epoch": 0.8798514391829155,
      "grad_norm": 1.5456181658164783,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8720887303352356,
      "num_tokens": 164853651.0,
      "step": 4738
    },
    {
      "epoch": 0.8800371402042712,
      "grad_norm": 1.499906741652043,
      "learning_rate": 1e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8586674928665161,
      "num_tokens": 164888952.0,
      "step": 4739
    },
    {
      "epoch": 0.8802228412256268,
      "grad_norm": 1.5217681628686113,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8773098587989807,
      "num_tokens": 164922089.0,
      "step": 4740
    },
    {
      "epoch": 0.8804085422469824,
      "grad_norm": 1.3762352092929377,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8723909258842468,
      "num_tokens": 164960359.0,
      "step": 4741
    },
    {
      "epoch": 0.880594243268338,
      "grad_norm": 1.5632537616095716,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8628629446029663,
      "num_tokens": 164992996.0,
      "step": 4742
    },
    {
      "epoch": 0.8807799442896936,
      "grad_norm": 1.3051306937607545,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8785167932510376,
      "num_tokens": 165032535.0,
      "step": 4743
    },
    {
      "epoch": 0.8809656453110493,
      "grad_norm": 1.4979463463476095,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.869796872138977,
      "num_tokens": 165065460.0,
      "step": 4744
    },
    {
      "epoch": 0.8811513463324049,
      "grad_norm": 1.3912986791896642,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8767550587654114,
      "num_tokens": 165100875.0,
      "step": 4745
    },
    {
      "epoch": 0.8813370473537604,
      "grad_norm": 1.664394962959484,
      "learning_rate": 1e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8512610793113708,
      "num_tokens": 165131173.0,
      "step": 4746
    },
    {
      "epoch": 0.881522748375116,
      "grad_norm": 1.3859869230086261,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8707555532455444,
      "num_tokens": 165171211.0,
      "step": 4747
    },
    {
      "epoch": 0.8817084493964716,
      "grad_norm": 1.5842169738977456,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8572365641593933,
      "num_tokens": 165201167.0,
      "step": 4748
    },
    {
      "epoch": 0.8818941504178273,
      "grad_norm": 1.2669615183133704,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8777160048484802,
      "num_tokens": 165242034.0,
      "step": 4749
    },
    {
      "epoch": 0.8820798514391829,
      "grad_norm": 1.3980313475502153,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8746246099472046,
      "num_tokens": 165277685.0,
      "step": 4750
    },
    {
      "epoch": 0.8822655524605385,
      "grad_norm": 1.4465362766215755,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8742127418518066,
      "num_tokens": 165315975.0,
      "step": 4751
    },
    {
      "epoch": 0.8824512534818941,
      "grad_norm": 1.3596073381138902,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8777514696121216,
      "num_tokens": 165352095.0,
      "step": 4752
    },
    {
      "epoch": 0.8826369545032497,
      "grad_norm": 1.3574842598932606,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8686627149581909,
      "num_tokens": 165392246.0,
      "step": 4753
    },
    {
      "epoch": 0.8828226555246054,
      "grad_norm": 1.7162469831082214,
      "learning_rate": 1e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8526464700698853,
      "num_tokens": 165421267.0,
      "step": 4754
    },
    {
      "epoch": 0.883008356545961,
      "grad_norm": 1.5013919052894569,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8784926533699036,
      "num_tokens": 165451690.0,
      "step": 4755
    },
    {
      "epoch": 0.8831940575673166,
      "grad_norm": 1.5114065816619247,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8809241652488708,
      "num_tokens": 165483081.0,
      "step": 4756
    },
    {
      "epoch": 0.8833797585886722,
      "grad_norm": 1.5724612783458702,
      "learning_rate": 1e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8570340275764465,
      "num_tokens": 165515812.0,
      "step": 4757
    },
    {
      "epoch": 0.8835654596100279,
      "grad_norm": 1.4299645986496683,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8834334015846252,
      "num_tokens": 165547268.0,
      "step": 4758
    },
    {
      "epoch": 0.8837511606313835,
      "grad_norm": 1.4490087248405237,
      "learning_rate": 1e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8609806895256042,
      "num_tokens": 165585159.0,
      "step": 4759
    },
    {
      "epoch": 0.8839368616527391,
      "grad_norm": 1.371463266524094,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8835618495941162,
      "num_tokens": 165619851.0,
      "step": 4760
    },
    {
      "epoch": 0.8841225626740947,
      "grad_norm": 1.468763435239614,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8688315749168396,
      "num_tokens": 165655289.0,
      "step": 4761
    },
    {
      "epoch": 0.8843082636954503,
      "grad_norm": 1.4847395748313217,
      "learning_rate": 1e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8621435165405273,
      "num_tokens": 165692859.0,
      "step": 4762
    },
    {
      "epoch": 0.884493964716806,
      "grad_norm": 1.538591744839335,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8699413537979126,
      "num_tokens": 165722382.0,
      "step": 4763
    },
    {
      "epoch": 0.8846796657381616,
      "grad_norm": 1.516693456273196,
      "learning_rate": 1e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8481810688972473,
      "num_tokens": 165758042.0,
      "step": 4764
    },
    {
      "epoch": 0.8848653667595172,
      "grad_norm": 1.6166896202570207,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8769530057907104,
      "num_tokens": 165788145.0,
      "step": 4765
    },
    {
      "epoch": 0.8850510677808728,
      "grad_norm": 1.5065084752201297,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8620914816856384,
      "num_tokens": 165820734.0,
      "step": 4766
    },
    {
      "epoch": 0.8852367688022285,
      "grad_norm": 1.4530461853185612,
      "learning_rate": 1e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8560546636581421,
      "num_tokens": 165860430.0,
      "step": 4767
    },
    {
      "epoch": 0.8854224698235841,
      "grad_norm": 1.3642543939125513,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8740094304084778,
      "num_tokens": 165899527.0,
      "step": 4768
    },
    {
      "epoch": 0.8856081708449397,
      "grad_norm": 1.5872732016171063,
      "learning_rate": 1e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8580663204193115,
      "num_tokens": 165933629.0,
      "step": 4769
    },
    {
      "epoch": 0.8857938718662952,
      "grad_norm": 1.3040761752618781,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8758847117424011,
      "num_tokens": 165974824.0,
      "step": 4770
    },
    {
      "epoch": 0.8859795728876508,
      "grad_norm": 1.2965430932114645,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8829318284988403,
      "num_tokens": 166012751.0,
      "step": 4771
    },
    {
      "epoch": 0.8861652739090065,
      "grad_norm": 1.3239133983061147,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.8790180087089539,
      "num_tokens": 166051027.0,
      "step": 4772
    },
    {
      "epoch": 0.8863509749303621,
      "grad_norm": 1.3132020771606243,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8684750199317932,
      "num_tokens": 166094787.0,
      "step": 4773
    },
    {
      "epoch": 0.8865366759517177,
      "grad_norm": 1.5555120691031776,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8666676878929138,
      "num_tokens": 166124568.0,
      "step": 4774
    },
    {
      "epoch": 0.8867223769730733,
      "grad_norm": 1.5616018448154771,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8689600229263306,
      "num_tokens": 166155994.0,
      "step": 4775
    },
    {
      "epoch": 0.886908077994429,
      "grad_norm": 1.4570823202190184,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8717377185821533,
      "num_tokens": 166189379.0,
      "step": 4776
    },
    {
      "epoch": 0.8870937790157846,
      "grad_norm": 1.5139879218684629,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.885050892829895,
      "num_tokens": 166218381.0,
      "step": 4777
    },
    {
      "epoch": 0.8872794800371402,
      "grad_norm": 1.4078705268635605,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8673498034477234,
      "num_tokens": 166256410.0,
      "step": 4778
    },
    {
      "epoch": 0.8874651810584958,
      "grad_norm": 1.3423977135278462,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8623117208480835,
      "num_tokens": 166300766.0,
      "step": 4779
    },
    {
      "epoch": 0.8876508820798514,
      "grad_norm": 1.50773475379056,
      "learning_rate": 1e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8601492047309875,
      "num_tokens": 166335915.0,
      "step": 4780
    },
    {
      "epoch": 0.887836583101207,
      "grad_norm": 1.519222874756176,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8762626647949219,
      "num_tokens": 166364675.0,
      "step": 4781
    },
    {
      "epoch": 0.8880222841225627,
      "grad_norm": 1.350549074378312,
      "learning_rate": 1e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8607730865478516,
      "num_tokens": 166404014.0,
      "step": 4782
    },
    {
      "epoch": 0.8882079851439183,
      "grad_norm": 1.4193747949537183,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8803039789199829,
      "num_tokens": 166440532.0,
      "step": 4783
    },
    {
      "epoch": 0.8883936861652739,
      "grad_norm": 1.5077891419625156,
      "learning_rate": 1e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8559643626213074,
      "num_tokens": 166477680.0,
      "step": 4784
    },
    {
      "epoch": 0.8885793871866295,
      "grad_norm": 1.285138897400322,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8852230310440063,
      "num_tokens": 166512719.0,
      "step": 4785
    },
    {
      "epoch": 0.8887650882079852,
      "grad_norm": 1.396812596997217,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8713729381561279,
      "num_tokens": 166552533.0,
      "step": 4786
    },
    {
      "epoch": 0.8889507892293408,
      "grad_norm": 1.4432070181384384,
      "learning_rate": 1e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8695081472396851,
      "num_tokens": 166588055.0,
      "step": 4787
    },
    {
      "epoch": 0.8891364902506964,
      "grad_norm": 1.582245992090737,
      "learning_rate": 1e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8487608432769775,
      "num_tokens": 166620984.0,
      "step": 4788
    },
    {
      "epoch": 0.889322191272052,
      "grad_norm": 1.4777818502009272,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.876072883605957,
      "num_tokens": 166653432.0,
      "step": 4789
    },
    {
      "epoch": 0.8895078922934077,
      "grad_norm": 1.4878053712826573,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8804234266281128,
      "num_tokens": 166685165.0,
      "step": 4790
    },
    {
      "epoch": 0.8896935933147633,
      "grad_norm": 1.5986663249715083,
      "learning_rate": 1e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8511137962341309,
      "num_tokens": 166719067.0,
      "step": 4791
    },
    {
      "epoch": 0.8898792943361189,
      "grad_norm": 1.458337438810813,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8624354600906372,
      "num_tokens": 166757115.0,
      "step": 4792
    },
    {
      "epoch": 0.8900649953574745,
      "grad_norm": 1.3035265541476573,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8730477094650269,
      "num_tokens": 166799994.0,
      "step": 4793
    },
    {
      "epoch": 0.89025069637883,
      "grad_norm": 1.5300999611295467,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8728984594345093,
      "num_tokens": 166834262.0,
      "step": 4794
    },
    {
      "epoch": 0.8904363974001857,
      "grad_norm": 1.3735614004554229,
      "learning_rate": 1e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8589190244674683,
      "num_tokens": 166876468.0,
      "step": 4795
    },
    {
      "epoch": 0.8906220984215413,
      "grad_norm": 1.6006300547346153,
      "learning_rate": 1e-06,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8615862727165222,
      "num_tokens": 166905856.0,
      "step": 4796
    },
    {
      "epoch": 0.8908077994428969,
      "grad_norm": 1.443432102937222,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.857559323310852,
      "num_tokens": 166943344.0,
      "step": 4797
    },
    {
      "epoch": 0.8909935004642525,
      "grad_norm": 1.5167701182233755,
      "learning_rate": 1e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8670406341552734,
      "num_tokens": 166979635.0,
      "step": 4798
    },
    {
      "epoch": 0.8911792014856081,
      "grad_norm": 1.5349661970858477,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8719801306724548,
      "num_tokens": 167009502.0,
      "step": 4799
    },
    {
      "epoch": 0.8913649025069638,
      "grad_norm": 1.4494966594706744,
      "learning_rate": 1e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8668068051338196,
      "num_tokens": 167043499.0,
      "step": 4800
    },
    {
      "epoch": 0.8915506035283194,
      "grad_norm": 1.520807162189909,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.869554877281189,
      "num_tokens": 167080525.0,
      "step": 4801
    },
    {
      "epoch": 0.891736304549675,
      "grad_norm": 1.517805749377572,
      "learning_rate": 1e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8420776724815369,
      "num_tokens": 167119458.0,
      "step": 4802
    },
    {
      "epoch": 0.8919220055710306,
      "grad_norm": 1.4345528736448152,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8827093839645386,
      "num_tokens": 167154971.0,
      "step": 4803
    },
    {
      "epoch": 0.8921077065923863,
      "grad_norm": 1.3879939685900342,
      "learning_rate": 1e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8619636297225952,
      "num_tokens": 167194228.0,
      "step": 4804
    },
    {
      "epoch": 0.8922934076137419,
      "grad_norm": 1.5110042723944554,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8671392798423767,
      "num_tokens": 167228486.0,
      "step": 4805
    },
    {
      "epoch": 0.8924791086350975,
      "grad_norm": 1.4155689512663658,
      "learning_rate": 1e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8607310056686401,
      "num_tokens": 167268647.0,
      "step": 4806
    },
    {
      "epoch": 0.8926648096564531,
      "grad_norm": 1.4660007018410133,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8721895217895508,
      "num_tokens": 167300278.0,
      "step": 4807
    },
    {
      "epoch": 0.8928505106778087,
      "grad_norm": 1.4127212208025417,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8664485216140747,
      "num_tokens": 167338341.0,
      "step": 4808
    },
    {
      "epoch": 0.8930362116991644,
      "grad_norm": 1.3953343993620069,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8721074461936951,
      "num_tokens": 167377239.0,
      "step": 4809
    },
    {
      "epoch": 0.89322191272052,
      "grad_norm": 1.349809813994802,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8760412335395813,
      "num_tokens": 167415183.0,
      "step": 4810
    },
    {
      "epoch": 0.8934076137418756,
      "grad_norm": 1.464116670513948,
      "learning_rate": 1e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.852544367313385,
      "num_tokens": 167455499.0,
      "step": 4811
    },
    {
      "epoch": 0.8935933147632312,
      "grad_norm": 1.3698214357085468,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.874172568321228,
      "num_tokens": 167491785.0,
      "step": 4812
    },
    {
      "epoch": 0.8937790157845868,
      "grad_norm": 1.4322059102174758,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.883423388004303,
      "num_tokens": 167524240.0,
      "step": 4813
    },
    {
      "epoch": 0.8939647168059425,
      "grad_norm": 1.3607164650229346,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8723319172859192,
      "num_tokens": 167560920.0,
      "step": 4814
    },
    {
      "epoch": 0.8941504178272981,
      "grad_norm": 1.4019914993217562,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8697599172592163,
      "num_tokens": 167597883.0,
      "step": 4815
    },
    {
      "epoch": 0.8943361188486537,
      "grad_norm": 1.4909793544165284,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8714213967323303,
      "num_tokens": 167627633.0,
      "step": 4816
    },
    {
      "epoch": 0.8945218198700093,
      "grad_norm": 1.3716330551277107,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8812527656555176,
      "num_tokens": 167665030.0,
      "step": 4817
    },
    {
      "epoch": 0.8947075208913648,
      "grad_norm": 1.5227407695430377,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8749628663063049,
      "num_tokens": 167697769.0,
      "step": 4818
    },
    {
      "epoch": 0.8948932219127205,
      "grad_norm": 1.7342581711670852,
      "learning_rate": 1e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8458998203277588,
      "num_tokens": 167727520.0,
      "step": 4819
    },
    {
      "epoch": 0.8950789229340761,
      "grad_norm": 1.4828899405585776,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8708348870277405,
      "num_tokens": 167757645.0,
      "step": 4820
    },
    {
      "epoch": 0.8952646239554317,
      "grad_norm": 1.4055704490009675,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8797471523284912,
      "num_tokens": 167792882.0,
      "step": 4821
    },
    {
      "epoch": 0.8954503249767873,
      "grad_norm": 1.513601117885975,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.863063395023346,
      "num_tokens": 167829944.0,
      "step": 4822
    },
    {
      "epoch": 0.895636025998143,
      "grad_norm": 1.4706716697178008,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8683212399482727,
      "num_tokens": 167863184.0,
      "step": 4823
    },
    {
      "epoch": 0.8958217270194986,
      "grad_norm": 1.541947807531352,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8787916302680969,
      "num_tokens": 167892201.0,
      "step": 4824
    },
    {
      "epoch": 0.8960074280408542,
      "grad_norm": 1.2471359615398059,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8822821974754333,
      "num_tokens": 167935222.0,
      "step": 4825
    },
    {
      "epoch": 0.8961931290622098,
      "grad_norm": 1.4272142625169104,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8667376041412354,
      "num_tokens": 167969114.0,
      "step": 4826
    },
    {
      "epoch": 0.8963788300835654,
      "grad_norm": 1.4427373629742763,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8735524415969849,
      "num_tokens": 168008450.0,
      "step": 4827
    },
    {
      "epoch": 0.8965645311049211,
      "grad_norm": 1.5865128653900318,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8695323467254639,
      "num_tokens": 168041536.0,
      "step": 4828
    },
    {
      "epoch": 0.8967502321262767,
      "grad_norm": 1.4906353482509676,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8798476457595825,
      "num_tokens": 168071003.0,
      "step": 4829
    },
    {
      "epoch": 0.8969359331476323,
      "grad_norm": 1.4416889872179837,
      "learning_rate": 1e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8660696744918823,
      "num_tokens": 168105249.0,
      "step": 4830
    },
    {
      "epoch": 0.8971216341689879,
      "grad_norm": 1.5266170073775032,
      "learning_rate": 1e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.848976731300354,
      "num_tokens": 168139499.0,
      "step": 4831
    },
    {
      "epoch": 0.8973073351903436,
      "grad_norm": 1.4942240899145567,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8677911162376404,
      "num_tokens": 168172552.0,
      "step": 4832
    },
    {
      "epoch": 0.8974930362116992,
      "grad_norm": 1.549592914780991,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8771675825119019,
      "num_tokens": 168204344.0,
      "step": 4833
    },
    {
      "epoch": 0.8976787372330548,
      "grad_norm": 1.6761838788048005,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8676050305366516,
      "num_tokens": 168235200.0,
      "step": 4834
    },
    {
      "epoch": 0.8978644382544104,
      "grad_norm": 1.5335243274879349,
      "learning_rate": 1e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8552103042602539,
      "num_tokens": 168269424.0,
      "step": 4835
    },
    {
      "epoch": 0.898050139275766,
      "grad_norm": 1.5741979977872922,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8627989888191223,
      "num_tokens": 168304913.0,
      "step": 4836
    },
    {
      "epoch": 0.8982358402971217,
      "grad_norm": 1.4171636700257972,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8667914867401123,
      "num_tokens": 168343536.0,
      "step": 4837
    },
    {
      "epoch": 0.8984215413184773,
      "grad_norm": 1.453092605541091,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8656201362609863,
      "num_tokens": 168378882.0,
      "step": 4838
    },
    {
      "epoch": 0.8986072423398329,
      "grad_norm": 1.4701163061063796,
      "learning_rate": 1e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8645184636116028,
      "num_tokens": 168416253.0,
      "step": 4839
    },
    {
      "epoch": 0.8987929433611885,
      "grad_norm": 1.501465292109413,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8690235614776611,
      "num_tokens": 168447978.0,
      "step": 4840
    },
    {
      "epoch": 0.8989786443825442,
      "grad_norm": 1.2934862562165634,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8815022706985474,
      "num_tokens": 168484668.0,
      "step": 4841
    },
    {
      "epoch": 0.8991643454038997,
      "grad_norm": 1.3443464632521112,
      "learning_rate": 1e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.850513219833374,
      "num_tokens": 168526827.0,
      "step": 4842
    },
    {
      "epoch": 0.8993500464252553,
      "grad_norm": 1.5663582028771086,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8631325960159302,
      "num_tokens": 168558139.0,
      "step": 4843
    },
    {
      "epoch": 0.8995357474466109,
      "grad_norm": 1.2984354686467414,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8676264882087708,
      "num_tokens": 168600368.0,
      "step": 4844
    },
    {
      "epoch": 0.8997214484679665,
      "grad_norm": 1.4864919803454761,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8657718896865845,
      "num_tokens": 168633571.0,
      "step": 4845
    },
    {
      "epoch": 0.8999071494893222,
      "grad_norm": 1.4269297609924727,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8725715279579163,
      "num_tokens": 168667151.0,
      "step": 4846
    },
    {
      "epoch": 0.9000928505106778,
      "grad_norm": 1.459380299131501,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.874556839466095,
      "num_tokens": 168702646.0,
      "step": 4847
    },
    {
      "epoch": 0.9002785515320334,
      "grad_norm": 1.5150795780135105,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8883349299430847,
      "num_tokens": 168732953.0,
      "step": 4848
    },
    {
      "epoch": 0.900464252553389,
      "grad_norm": 1.3617542725202678,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8765407800674438,
      "num_tokens": 168771633.0,
      "step": 4849
    },
    {
      "epoch": 0.9006499535747446,
      "grad_norm": 1.4055422761934204,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8659376502037048,
      "num_tokens": 168808528.0,
      "step": 4850
    },
    {
      "epoch": 0.9008356545961003,
      "grad_norm": 1.378568591693933,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8680233359336853,
      "num_tokens": 168850302.0,
      "step": 4851
    },
    {
      "epoch": 0.9010213556174559,
      "grad_norm": 1.412911136558092,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8711344599723816,
      "num_tokens": 168888318.0,
      "step": 4852
    },
    {
      "epoch": 0.9012070566388115,
      "grad_norm": 1.5199037603698016,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8640813827514648,
      "num_tokens": 168918827.0,
      "step": 4853
    },
    {
      "epoch": 0.9013927576601671,
      "grad_norm": 1.3535358901014536,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8783202767372131,
      "num_tokens": 168957040.0,
      "step": 4854
    },
    {
      "epoch": 0.9015784586815228,
      "grad_norm": 1.4077223914123782,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8658453822135925,
      "num_tokens": 168992872.0,
      "step": 4855
    },
    {
      "epoch": 0.9017641597028784,
      "grad_norm": 1.4486990619388564,
      "learning_rate": 1e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.857093095779419,
      "num_tokens": 169032096.0,
      "step": 4856
    },
    {
      "epoch": 0.901949860724234,
      "grad_norm": 1.4353856428817666,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8657320737838745,
      "num_tokens": 169069526.0,
      "step": 4857
    },
    {
      "epoch": 0.9021355617455896,
      "grad_norm": 1.419782155530766,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8781575560569763,
      "num_tokens": 169102670.0,
      "step": 4858
    },
    {
      "epoch": 0.9023212627669452,
      "grad_norm": 1.5085713990001415,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8789631128311157,
      "num_tokens": 169135245.0,
      "step": 4859
    },
    {
      "epoch": 0.9025069637883009,
      "grad_norm": 1.5041397603759825,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8678314685821533,
      "num_tokens": 169166686.0,
      "step": 4860
    },
    {
      "epoch": 0.9026926648096565,
      "grad_norm": 1.529963754315425,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8656023740768433,
      "num_tokens": 169200039.0,
      "step": 4861
    },
    {
      "epoch": 0.9028783658310121,
      "grad_norm": 1.2898823743755377,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8689228296279907,
      "num_tokens": 169239872.0,
      "step": 4862
    },
    {
      "epoch": 0.9030640668523677,
      "grad_norm": 1.3709662338982673,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.877120852470398,
      "num_tokens": 169279963.0,
      "step": 4863
    },
    {
      "epoch": 0.9032497678737234,
      "grad_norm": 1.3440357855557348,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8716720342636108,
      "num_tokens": 169321586.0,
      "step": 4864
    },
    {
      "epoch": 0.903435468895079,
      "grad_norm": 1.3959526183855338,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8684994578361511,
      "num_tokens": 169357394.0,
      "step": 4865
    },
    {
      "epoch": 0.9036211699164345,
      "grad_norm": 1.484587216096836,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8721737861633301,
      "num_tokens": 169389647.0,
      "step": 4866
    },
    {
      "epoch": 0.9038068709377901,
      "grad_norm": 1.4702561563432097,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8843518495559692,
      "num_tokens": 169421280.0,
      "step": 4867
    },
    {
      "epoch": 0.9039925719591457,
      "grad_norm": 1.4461782563685548,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8732719421386719,
      "num_tokens": 169462308.0,
      "step": 4868
    },
    {
      "epoch": 0.9041782729805014,
      "grad_norm": 1.5663170710745453,
      "learning_rate": 1e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8666046857833862,
      "num_tokens": 169494690.0,
      "step": 4869
    },
    {
      "epoch": 0.904363974001857,
      "grad_norm": 1.4400498481169994,
      "learning_rate": 1e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8534437417984009,
      "num_tokens": 169530711.0,
      "step": 4870
    },
    {
      "epoch": 0.9045496750232126,
      "grad_norm": 1.3795088120189802,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8655128479003906,
      "num_tokens": 169566618.0,
      "step": 4871
    },
    {
      "epoch": 0.9047353760445682,
      "grad_norm": 1.5346528277149885,
      "learning_rate": 1e-06,
      "loss": 0.3866,
      "mean_token_accuracy": 0.8671219348907471,
      "num_tokens": 169596626.0,
      "step": 4872
    },
    {
      "epoch": 0.9049210770659238,
      "grad_norm": 1.3883901706241693,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8669130802154541,
      "num_tokens": 169635830.0,
      "step": 4873
    },
    {
      "epoch": 0.9051067780872795,
      "grad_norm": 1.574123287132734,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.879535436630249,
      "num_tokens": 169665949.0,
      "step": 4874
    },
    {
      "epoch": 0.9052924791086351,
      "grad_norm": 1.483612230382151,
      "learning_rate": 1e-06,
      "loss": 0.3921,
      "mean_token_accuracy": 0.8726259469985962,
      "num_tokens": 169696571.0,
      "step": 4875
    },
    {
      "epoch": 0.9054781801299907,
      "grad_norm": 1.3328626728448438,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8757153749465942,
      "num_tokens": 169736462.0,
      "step": 4876
    },
    {
      "epoch": 0.9056638811513463,
      "grad_norm": 1.3551805919767406,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.875548779964447,
      "num_tokens": 169773997.0,
      "step": 4877
    },
    {
      "epoch": 0.905849582172702,
      "grad_norm": 1.3933947986151998,
      "learning_rate": 1e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8669499158859253,
      "num_tokens": 169811009.0,
      "step": 4878
    },
    {
      "epoch": 0.9060352831940576,
      "grad_norm": 1.4174752176017147,
      "learning_rate": 1e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8650189638137817,
      "num_tokens": 169846081.0,
      "step": 4879
    },
    {
      "epoch": 0.9062209842154132,
      "grad_norm": 1.5776183966122768,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8709261417388916,
      "num_tokens": 169876961.0,
      "step": 4880
    },
    {
      "epoch": 0.9064066852367688,
      "grad_norm": 1.5982345870190127,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8660269975662231,
      "num_tokens": 169909541.0,
      "step": 4881
    },
    {
      "epoch": 0.9065923862581244,
      "grad_norm": 1.6107715921516554,
      "learning_rate": 1e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8506540060043335,
      "num_tokens": 169942167.0,
      "step": 4882
    },
    {
      "epoch": 0.9067780872794801,
      "grad_norm": 1.4240770692764466,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8725026845932007,
      "num_tokens": 169976122.0,
      "step": 4883
    },
    {
      "epoch": 0.9069637883008357,
      "grad_norm": 1.415752889609257,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8734246492385864,
      "num_tokens": 170016028.0,
      "step": 4884
    },
    {
      "epoch": 0.9071494893221913,
      "grad_norm": 1.4877653141631206,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8622546792030334,
      "num_tokens": 170051262.0,
      "step": 4885
    },
    {
      "epoch": 0.9073351903435469,
      "grad_norm": 1.5123059758177737,
      "learning_rate": 1e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8604713678359985,
      "num_tokens": 170086494.0,
      "step": 4886
    },
    {
      "epoch": 0.9075208913649025,
      "grad_norm": 1.3832268625603426,
      "learning_rate": 1e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.856595516204834,
      "num_tokens": 170126936.0,
      "step": 4887
    },
    {
      "epoch": 0.9077065923862582,
      "grad_norm": 1.5487561671151435,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8748382925987244,
      "num_tokens": 170158128.0,
      "step": 4888
    },
    {
      "epoch": 0.9078922934076138,
      "grad_norm": 1.5442677155334388,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.865593433380127,
      "num_tokens": 170198093.0,
      "step": 4889
    },
    {
      "epoch": 0.9080779944289693,
      "grad_norm": 1.5041689385139734,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8638802766799927,
      "num_tokens": 170235875.0,
      "step": 4890
    },
    {
      "epoch": 0.9082636954503249,
      "grad_norm": 1.5501154403620974,
      "learning_rate": 1e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8611128330230713,
      "num_tokens": 170270581.0,
      "step": 4891
    },
    {
      "epoch": 0.9084493964716805,
      "grad_norm": 1.4675543343488164,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8862214088439941,
      "num_tokens": 170303698.0,
      "step": 4892
    },
    {
      "epoch": 0.9086350974930362,
      "grad_norm": 1.3088027740380452,
      "learning_rate": 1e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8562045693397522,
      "num_tokens": 170349721.0,
      "step": 4893
    },
    {
      "epoch": 0.9088207985143918,
      "grad_norm": 1.4592114001284897,
      "learning_rate": 1e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8623985052108765,
      "num_tokens": 170384507.0,
      "step": 4894
    },
    {
      "epoch": 0.9090064995357474,
      "grad_norm": 2.5592446775233206,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8601357340812683,
      "num_tokens": 170420231.0,
      "step": 4895
    },
    {
      "epoch": 0.909192200557103,
      "grad_norm": 1.7054192686008698,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8841536045074463,
      "num_tokens": 170446953.0,
      "step": 4896
    },
    {
      "epoch": 0.9093779015784587,
      "grad_norm": 1.3846772591300975,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8806978464126587,
      "num_tokens": 170485995.0,
      "step": 4897
    },
    {
      "epoch": 0.9095636025998143,
      "grad_norm": 1.539462854780979,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.879043459892273,
      "num_tokens": 170518646.0,
      "step": 4898
    },
    {
      "epoch": 0.9097493036211699,
      "grad_norm": 1.5595502355425845,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8801491260528564,
      "num_tokens": 170548499.0,
      "step": 4899
    },
    {
      "epoch": 0.9099350046425255,
      "grad_norm": 1.372722157158572,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8709268569946289,
      "num_tokens": 170586211.0,
      "step": 4900
    },
    {
      "epoch": 0.9101207056638811,
      "grad_norm": 1.3473132583677956,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8738824725151062,
      "num_tokens": 170627876.0,
      "step": 4901
    },
    {
      "epoch": 0.9103064066852368,
      "grad_norm": 1.5218400963759784,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8686788082122803,
      "num_tokens": 170664350.0,
      "step": 4902
    },
    {
      "epoch": 0.9104921077065924,
      "grad_norm": 1.3755055473010092,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.867404580116272,
      "num_tokens": 170703127.0,
      "step": 4903
    },
    {
      "epoch": 0.910677808727948,
      "grad_norm": 1.7034097466506337,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8600214719772339,
      "num_tokens": 170730855.0,
      "step": 4904
    },
    {
      "epoch": 0.9108635097493036,
      "grad_norm": 1.3389794483360402,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8801934123039246,
      "num_tokens": 170771132.0,
      "step": 4905
    },
    {
      "epoch": 0.9110492107706593,
      "grad_norm": 1.3651328952751336,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8758819103240967,
      "num_tokens": 170807927.0,
      "step": 4906
    },
    {
      "epoch": 0.9112349117920149,
      "grad_norm": 1.639469920124253,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.862977921962738,
      "num_tokens": 170837211.0,
      "step": 4907
    },
    {
      "epoch": 0.9114206128133705,
      "grad_norm": 1.6020452996090144,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8697170615196228,
      "num_tokens": 170867657.0,
      "step": 4908
    },
    {
      "epoch": 0.9116063138347261,
      "grad_norm": 1.4609040125451176,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8682922720909119,
      "num_tokens": 170904558.0,
      "step": 4909
    },
    {
      "epoch": 0.9117920148560817,
      "grad_norm": 1.3578652911612838,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8818527460098267,
      "num_tokens": 170943056.0,
      "step": 4910
    },
    {
      "epoch": 0.9119777158774374,
      "grad_norm": 1.5855040549832806,
      "learning_rate": 1e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8492494821548462,
      "num_tokens": 170974557.0,
      "step": 4911
    },
    {
      "epoch": 0.912163416898793,
      "grad_norm": 1.3953363206778429,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8820482492446899,
      "num_tokens": 171009977.0,
      "step": 4912
    },
    {
      "epoch": 0.9123491179201486,
      "grad_norm": 1.547908534685064,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.870147705078125,
      "num_tokens": 171040961.0,
      "step": 4913
    },
    {
      "epoch": 0.9125348189415042,
      "grad_norm": 1.4878113907039476,
      "learning_rate": 1e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8442492485046387,
      "num_tokens": 171079785.0,
      "step": 4914
    },
    {
      "epoch": 0.9127205199628597,
      "grad_norm": 1.6160734536869743,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8706042766571045,
      "num_tokens": 171114794.0,
      "step": 4915
    },
    {
      "epoch": 0.9129062209842154,
      "grad_norm": 1.3514071320491234,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.865717351436615,
      "num_tokens": 171155406.0,
      "step": 4916
    },
    {
      "epoch": 0.913091922005571,
      "grad_norm": 1.4861926850402853,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8714558482170105,
      "num_tokens": 171189436.0,
      "step": 4917
    },
    {
      "epoch": 0.9132776230269266,
      "grad_norm": 1.506835800181798,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.873894214630127,
      "num_tokens": 171222643.0,
      "step": 4918
    },
    {
      "epoch": 0.9134633240482822,
      "grad_norm": 1.4559955053172897,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8734506964683533,
      "num_tokens": 171255769.0,
      "step": 4919
    },
    {
      "epoch": 0.9136490250696379,
      "grad_norm": 1.5269574329061955,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8652037978172302,
      "num_tokens": 171290233.0,
      "step": 4920
    },
    {
      "epoch": 0.9138347260909935,
      "grad_norm": 1.4405062405577091,
      "learning_rate": 1e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8585009574890137,
      "num_tokens": 171330817.0,
      "step": 4921
    },
    {
      "epoch": 0.9140204271123491,
      "grad_norm": 1.7667279939961307,
      "learning_rate": 1e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8495548367500305,
      "num_tokens": 171358989.0,
      "step": 4922
    },
    {
      "epoch": 0.9142061281337047,
      "grad_norm": 1.468248759588019,
      "learning_rate": 1e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8556985855102539,
      "num_tokens": 171395356.0,
      "step": 4923
    },
    {
      "epoch": 0.9143918291550603,
      "grad_norm": 1.4150329191517719,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8723382949829102,
      "num_tokens": 171430128.0,
      "step": 4924
    },
    {
      "epoch": 0.914577530176416,
      "grad_norm": 1.431320898564392,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8701829314231873,
      "num_tokens": 171466384.0,
      "step": 4925
    },
    {
      "epoch": 0.9147632311977716,
      "grad_norm": 1.4771629812657734,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8809018135070801,
      "num_tokens": 171497977.0,
      "step": 4926
    },
    {
      "epoch": 0.9149489322191272,
      "grad_norm": 1.4459622366780367,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8649736642837524,
      "num_tokens": 171534002.0,
      "step": 4927
    },
    {
      "epoch": 0.9151346332404828,
      "grad_norm": 1.3852515619063603,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8732770681381226,
      "num_tokens": 171573553.0,
      "step": 4928
    },
    {
      "epoch": 0.9153203342618385,
      "grad_norm": 1.372528265903015,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8797006607055664,
      "num_tokens": 171607283.0,
      "step": 4929
    },
    {
      "epoch": 0.9155060352831941,
      "grad_norm": 1.5105510749140458,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8687753677368164,
      "num_tokens": 171639548.0,
      "step": 4930
    },
    {
      "epoch": 0.9156917363045497,
      "grad_norm": 1.4140372484123378,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8698961734771729,
      "num_tokens": 171673876.0,
      "step": 4931
    },
    {
      "epoch": 0.9158774373259053,
      "grad_norm": 1.3080255851945823,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8697367906570435,
      "num_tokens": 171716661.0,
      "step": 4932
    },
    {
      "epoch": 0.9160631383472609,
      "grad_norm": 1.466172416506764,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8666003942489624,
      "num_tokens": 171751314.0,
      "step": 4933
    },
    {
      "epoch": 0.9162488393686166,
      "grad_norm": 1.5697732534888809,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8753520250320435,
      "num_tokens": 171779000.0,
      "step": 4934
    },
    {
      "epoch": 0.9164345403899722,
      "grad_norm": 1.3880872123092098,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.877833366394043,
      "num_tokens": 171813548.0,
      "step": 4935
    },
    {
      "epoch": 0.9166202414113278,
      "grad_norm": 1.5424413593933488,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.875775933265686,
      "num_tokens": 171848259.0,
      "step": 4936
    },
    {
      "epoch": 0.9168059424326834,
      "grad_norm": 1.4320150389819215,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8690717220306396,
      "num_tokens": 171885131.0,
      "step": 4937
    },
    {
      "epoch": 0.916991643454039,
      "grad_norm": 1.6503787147465743,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8853272199630737,
      "num_tokens": 171912733.0,
      "step": 4938
    },
    {
      "epoch": 0.9171773444753946,
      "grad_norm": 1.3645899753311126,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.867958664894104,
      "num_tokens": 171951698.0,
      "step": 4939
    },
    {
      "epoch": 0.9173630454967502,
      "grad_norm": 1.4544440699869028,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8653013706207275,
      "num_tokens": 171988606.0,
      "step": 4940
    },
    {
      "epoch": 0.9175487465181058,
      "grad_norm": 1.6842942232094162,
      "learning_rate": 1e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8539276719093323,
      "num_tokens": 172015286.0,
      "step": 4941
    },
    {
      "epoch": 0.9177344475394614,
      "grad_norm": 1.536478934224909,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8716527223587036,
      "num_tokens": 172049722.0,
      "step": 4942
    },
    {
      "epoch": 0.917920148560817,
      "grad_norm": 1.417173867913357,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8705757856369019,
      "num_tokens": 172086151.0,
      "step": 4943
    },
    {
      "epoch": 0.9181058495821727,
      "grad_norm": 1.5522168711663094,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8692958950996399,
      "num_tokens": 172116783.0,
      "step": 4944
    },
    {
      "epoch": 0.9182915506035283,
      "grad_norm": 1.3154207089318801,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8604567050933838,
      "num_tokens": 172161031.0,
      "step": 4945
    },
    {
      "epoch": 0.9184772516248839,
      "grad_norm": 1.3956645685449798,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8859775066375732,
      "num_tokens": 172196874.0,
      "step": 4946
    },
    {
      "epoch": 0.9186629526462395,
      "grad_norm": 1.4266575811956899,
      "learning_rate": 1e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8699690699577332,
      "num_tokens": 172232345.0,
      "step": 4947
    },
    {
      "epoch": 0.9188486536675952,
      "grad_norm": 1.375415636162658,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8818330764770508,
      "num_tokens": 172265885.0,
      "step": 4948
    },
    {
      "epoch": 0.9190343546889508,
      "grad_norm": 1.32773225274978,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8766106367111206,
      "num_tokens": 172304793.0,
      "step": 4949
    },
    {
      "epoch": 0.9192200557103064,
      "grad_norm": 1.539063337393577,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.8829237818717957,
      "num_tokens": 172334661.0,
      "step": 4950
    },
    {
      "epoch": 0.919405756731662,
      "grad_norm": 1.5564702148778127,
      "learning_rate": 1e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8445292711257935,
      "num_tokens": 172368604.0,
      "step": 4951
    },
    {
      "epoch": 0.9195914577530176,
      "grad_norm": 1.536153339467277,
      "learning_rate": 1e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8564580678939819,
      "num_tokens": 172400436.0,
      "step": 4952
    },
    {
      "epoch": 0.9197771587743733,
      "grad_norm": 1.6313961009342706,
      "learning_rate": 1e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8465135097503662,
      "num_tokens": 172431905.0,
      "step": 4953
    },
    {
      "epoch": 0.9199628597957289,
      "grad_norm": 1.5047568830631817,
      "learning_rate": 1e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8528951406478882,
      "num_tokens": 172467215.0,
      "step": 4954
    },
    {
      "epoch": 0.9201485608170845,
      "grad_norm": 1.49353135867049,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8684363961219788,
      "num_tokens": 172500857.0,
      "step": 4955
    },
    {
      "epoch": 0.9203342618384401,
      "grad_norm": 1.3852738912069618,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8781759738922119,
      "num_tokens": 172536724.0,
      "step": 4956
    },
    {
      "epoch": 0.9205199628597958,
      "grad_norm": 1.5043957491998776,
      "learning_rate": 1e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8552483916282654,
      "num_tokens": 172567812.0,
      "step": 4957
    },
    {
      "epoch": 0.9207056638811514,
      "grad_norm": 1.6349467930055765,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8728424310684204,
      "num_tokens": 172592675.0,
      "step": 4958
    },
    {
      "epoch": 0.920891364902507,
      "grad_norm": 1.368365159562417,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8844067454338074,
      "num_tokens": 172627272.0,
      "step": 4959
    },
    {
      "epoch": 0.9210770659238626,
      "grad_norm": 1.412565006329723,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8731987476348877,
      "num_tokens": 172662911.0,
      "step": 4960
    },
    {
      "epoch": 0.9212627669452182,
      "grad_norm": 1.4566660171800954,
      "learning_rate": 1e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8587656021118164,
      "num_tokens": 172700373.0,
      "step": 4961
    },
    {
      "epoch": 0.9214484679665739,
      "grad_norm": 1.411290048455514,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8609323501586914,
      "num_tokens": 172740283.0,
      "step": 4962
    },
    {
      "epoch": 0.9216341689879294,
      "grad_norm": 1.5382789575607363,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8717721700668335,
      "num_tokens": 172769979.0,
      "step": 4963
    },
    {
      "epoch": 0.921819870009285,
      "grad_norm": 1.530513448603761,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8701339960098267,
      "num_tokens": 172798843.0,
      "step": 4964
    },
    {
      "epoch": 0.9220055710306406,
      "grad_norm": 1.2248749132284964,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8751945495605469,
      "num_tokens": 172845573.0,
      "step": 4965
    },
    {
      "epoch": 0.9221912720519962,
      "grad_norm": 1.5979233946034261,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8708339929580688,
      "num_tokens": 172874286.0,
      "step": 4966
    },
    {
      "epoch": 0.9223769730733519,
      "grad_norm": 1.4834769724040728,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8868395686149597,
      "num_tokens": 172906920.0,
      "step": 4967
    },
    {
      "epoch": 0.9225626740947075,
      "grad_norm": 1.4460558392079936,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8583805561065674,
      "num_tokens": 172943720.0,
      "step": 4968
    },
    {
      "epoch": 0.9227483751160631,
      "grad_norm": 1.4328843537097606,
      "learning_rate": 1e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8525938987731934,
      "num_tokens": 172979800.0,
      "step": 4969
    },
    {
      "epoch": 0.9229340761374187,
      "grad_norm": 1.4391634041219425,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8587372899055481,
      "num_tokens": 173016658.0,
      "step": 4970
    },
    {
      "epoch": 0.9231197771587744,
      "grad_norm": 1.4272536451448987,
      "learning_rate": 1e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.862418532371521,
      "num_tokens": 173056981.0,
      "step": 4971
    },
    {
      "epoch": 0.92330547818013,
      "grad_norm": 1.3169988818324705,
      "learning_rate": 1e-06,
      "loss": 0.3246,
      "mean_token_accuracy": 0.8861715793609619,
      "num_tokens": 173093110.0,
      "step": 4972
    },
    {
      "epoch": 0.9234911792014856,
      "grad_norm": 1.6494803907544322,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.8681321740150452,
      "num_tokens": 173122249.0,
      "step": 4973
    },
    {
      "epoch": 0.9236768802228412,
      "grad_norm": 1.5240523054145412,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.872643232345581,
      "num_tokens": 173150874.0,
      "step": 4974
    },
    {
      "epoch": 0.9238625812441968,
      "grad_norm": 1.4210153284663833,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8560464382171631,
      "num_tokens": 173188760.0,
      "step": 4975
    },
    {
      "epoch": 0.9240482822655525,
      "grad_norm": 1.5798093122208658,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8711804151535034,
      "num_tokens": 173220575.0,
      "step": 4976
    },
    {
      "epoch": 0.9242339832869081,
      "grad_norm": 1.516026608562874,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.86711585521698,
      "num_tokens": 173250522.0,
      "step": 4977
    },
    {
      "epoch": 0.9244196843082637,
      "grad_norm": 1.4462959468841707,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8768121004104614,
      "num_tokens": 173287471.0,
      "step": 4978
    },
    {
      "epoch": 0.9246053853296193,
      "grad_norm": 1.4339029685511833,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8776787519454956,
      "num_tokens": 173321050.0,
      "step": 4979
    },
    {
      "epoch": 0.924791086350975,
      "grad_norm": 1.4316525547993557,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.878564715385437,
      "num_tokens": 173357247.0,
      "step": 4980
    },
    {
      "epoch": 0.9249767873723306,
      "grad_norm": 1.4025728267535,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8694757223129272,
      "num_tokens": 173395454.0,
      "step": 4981
    },
    {
      "epoch": 0.9251624883936862,
      "grad_norm": 1.4140395227956142,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8811227679252625,
      "num_tokens": 173431381.0,
      "step": 4982
    },
    {
      "epoch": 0.9253481894150418,
      "grad_norm": 1.6598404864304759,
      "learning_rate": 1e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8582000732421875,
      "num_tokens": 173459686.0,
      "step": 4983
    },
    {
      "epoch": 0.9255338904363974,
      "grad_norm": 1.6534359896251347,
      "learning_rate": 1e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8376650214195251,
      "num_tokens": 173492869.0,
      "step": 4984
    },
    {
      "epoch": 0.9257195914577531,
      "grad_norm": 1.4701481542222523,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.872635006904602,
      "num_tokens": 173525294.0,
      "step": 4985
    },
    {
      "epoch": 0.9259052924791087,
      "grad_norm": 1.4601359156391285,
      "learning_rate": 1e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8586826324462891,
      "num_tokens": 173564092.0,
      "step": 4986
    },
    {
      "epoch": 0.9260909935004642,
      "grad_norm": 1.6364757970646115,
      "learning_rate": 1e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8644384145736694,
      "num_tokens": 173591836.0,
      "step": 4987
    },
    {
      "epoch": 0.9262766945218198,
      "grad_norm": 1.652509826521168,
      "learning_rate": 1e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8647710084915161,
      "num_tokens": 173622384.0,
      "step": 4988
    },
    {
      "epoch": 0.9264623955431754,
      "grad_norm": 1.5070150626162986,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8689669370651245,
      "num_tokens": 173659598.0,
      "step": 4989
    },
    {
      "epoch": 0.9266480965645311,
      "grad_norm": 1.4263674382610536,
      "learning_rate": 1e-06,
      "loss": 0.41,
      "mean_token_accuracy": 0.863542914390564,
      "num_tokens": 173700622.0,
      "step": 4990
    },
    {
      "epoch": 0.9268337975858867,
      "grad_norm": 1.4076124157614756,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.87645024061203,
      "num_tokens": 173733656.0,
      "step": 4991
    },
    {
      "epoch": 0.9270194986072423,
      "grad_norm": 1.3615815447917956,
      "learning_rate": 1e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8581176996231079,
      "num_tokens": 173774031.0,
      "step": 4992
    },
    {
      "epoch": 0.9272051996285979,
      "grad_norm": 1.4306654205111522,
      "learning_rate": 1e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8598377704620361,
      "num_tokens": 173810451.0,
      "step": 4993
    },
    {
      "epoch": 0.9273909006499536,
      "grad_norm": 1.3547531135291377,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8808963298797607,
      "num_tokens": 173846369.0,
      "step": 4994
    },
    {
      "epoch": 0.9275766016713092,
      "grad_norm": 1.4061038528354555,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8693948984146118,
      "num_tokens": 173881828.0,
      "step": 4995
    },
    {
      "epoch": 0.9277623026926648,
      "grad_norm": 1.591877335150539,
      "learning_rate": 1e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8519229292869568,
      "num_tokens": 173913228.0,
      "step": 4996
    },
    {
      "epoch": 0.9279480037140204,
      "grad_norm": 1.3867553293268329,
      "learning_rate": 1e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8619639277458191,
      "num_tokens": 173951145.0,
      "step": 4997
    },
    {
      "epoch": 0.928133704735376,
      "grad_norm": 1.3710424156368948,
      "learning_rate": 1e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8556962013244629,
      "num_tokens": 173988385.0,
      "step": 4998
    },
    {
      "epoch": 0.9283194057567317,
      "grad_norm": 1.543309220996457,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.863834023475647,
      "num_tokens": 174022365.0,
      "step": 4999
    },
    {
      "epoch": 0.9285051067780873,
      "grad_norm": 1.403056857403535,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8845267295837402,
      "num_tokens": 174054438.0,
      "step": 5000
    },
    {
      "epoch": 0.9286908077994429,
      "grad_norm": 1.4056552358121923,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8701250553131104,
      "num_tokens": 174091556.0,
      "step": 5001
    },
    {
      "epoch": 0.9288765088207985,
      "grad_norm": 1.3402976216816425,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8657171726226807,
      "num_tokens": 174133529.0,
      "step": 5002
    },
    {
      "epoch": 0.9290622098421542,
      "grad_norm": 1.5018168300678882,
      "learning_rate": 1e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8576840162277222,
      "num_tokens": 174169212.0,
      "step": 5003
    },
    {
      "epoch": 0.9292479108635098,
      "grad_norm": 1.4300305565476668,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8687567114830017,
      "num_tokens": 174207542.0,
      "step": 5004
    },
    {
      "epoch": 0.9294336118848654,
      "grad_norm": 1.421234960547653,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.866075873374939,
      "num_tokens": 174245275.0,
      "step": 5005
    },
    {
      "epoch": 0.929619312906221,
      "grad_norm": 1.4596544326615162,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8766685128211975,
      "num_tokens": 174279528.0,
      "step": 5006
    },
    {
      "epoch": 0.9298050139275766,
      "grad_norm": 1.3767575573573914,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8584878444671631,
      "num_tokens": 174320597.0,
      "step": 5007
    },
    {
      "epoch": 0.9299907149489323,
      "grad_norm": 1.299297329128673,
      "learning_rate": 1e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8664873838424683,
      "num_tokens": 174363267.0,
      "step": 5008
    },
    {
      "epoch": 0.9301764159702879,
      "grad_norm": 1.311586773548775,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8744409680366516,
      "num_tokens": 174404487.0,
      "step": 5009
    },
    {
      "epoch": 0.9303621169916435,
      "grad_norm": 1.425779855326786,
      "learning_rate": 1e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.8736790418624878,
      "num_tokens": 174439428.0,
      "step": 5010
    },
    {
      "epoch": 0.930547818012999,
      "grad_norm": 1.3287139395995873,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8698670864105225,
      "num_tokens": 174480345.0,
      "step": 5011
    },
    {
      "epoch": 0.9307335190343546,
      "grad_norm": 1.5367904110865707,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8830094933509827,
      "num_tokens": 174510778.0,
      "step": 5012
    },
    {
      "epoch": 0.9309192200557103,
      "grad_norm": 1.428209037831767,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8766259551048279,
      "num_tokens": 174547784.0,
      "step": 5013
    },
    {
      "epoch": 0.9311049210770659,
      "grad_norm": 1.770601069872366,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8566358089447021,
      "num_tokens": 174572619.0,
      "step": 5014
    },
    {
      "epoch": 0.9312906220984215,
      "grad_norm": 1.5251854781493093,
      "learning_rate": 1e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8504951000213623,
      "num_tokens": 174606297.0,
      "step": 5015
    },
    {
      "epoch": 0.9314763231197771,
      "grad_norm": 1.5462331449822193,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8714052438735962,
      "num_tokens": 174637321.0,
      "step": 5016
    },
    {
      "epoch": 0.9316620241411327,
      "grad_norm": 1.5120375300572289,
      "learning_rate": 1e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8608312606811523,
      "num_tokens": 174677591.0,
      "step": 5017
    },
    {
      "epoch": 0.9318477251624884,
      "grad_norm": 1.6247157292709777,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8674994111061096,
      "num_tokens": 174705990.0,
      "step": 5018
    },
    {
      "epoch": 0.932033426183844,
      "grad_norm": 1.573482971956071,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8638993501663208,
      "num_tokens": 174737502.0,
      "step": 5019
    },
    {
      "epoch": 0.9322191272051996,
      "grad_norm": 1.4834343181081626,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8736231923103333,
      "num_tokens": 174767499.0,
      "step": 5020
    },
    {
      "epoch": 0.9324048282265552,
      "grad_norm": 1.6014263389369106,
      "learning_rate": 1e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8512990474700928,
      "num_tokens": 174798360.0,
      "step": 5021
    },
    {
      "epoch": 0.9325905292479109,
      "grad_norm": 1.7933927149711155,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8647781610488892,
      "num_tokens": 174825583.0,
      "step": 5022
    },
    {
      "epoch": 0.9327762302692665,
      "grad_norm": 1.3719757922893607,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8729984760284424,
      "num_tokens": 174862244.0,
      "step": 5023
    },
    {
      "epoch": 0.9329619312906221,
      "grad_norm": 1.4790726609129754,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8735083341598511,
      "num_tokens": 174898501.0,
      "step": 5024
    },
    {
      "epoch": 0.9331476323119777,
      "grad_norm": 1.3693090114460347,
      "learning_rate": 1e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8554450273513794,
      "num_tokens": 174938799.0,
      "step": 5025
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 1.541401194644281,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8613439798355103,
      "num_tokens": 174974401.0,
      "step": 5026
    },
    {
      "epoch": 0.933519034354689,
      "grad_norm": 1.5026983414360802,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8752129077911377,
      "num_tokens": 175005066.0,
      "step": 5027
    },
    {
      "epoch": 0.9337047353760446,
      "grad_norm": 1.5312036701793346,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8690544962882996,
      "num_tokens": 175035923.0,
      "step": 5028
    },
    {
      "epoch": 0.9338904363974002,
      "grad_norm": 1.2893810218314703,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8772086501121521,
      "num_tokens": 175075452.0,
      "step": 5029
    },
    {
      "epoch": 0.9340761374187558,
      "grad_norm": 1.3909686005837512,
      "learning_rate": 1e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8674976825714111,
      "num_tokens": 175112259.0,
      "step": 5030
    },
    {
      "epoch": 0.9342618384401115,
      "grad_norm": 1.5791842829650484,
      "learning_rate": 1e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8609778881072998,
      "num_tokens": 175140568.0,
      "step": 5031
    },
    {
      "epoch": 0.9344475394614671,
      "grad_norm": 1.5837365798850709,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8557662963867188,
      "num_tokens": 175172359.0,
      "step": 5032
    },
    {
      "epoch": 0.9346332404828227,
      "grad_norm": 1.4079612375414166,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8775577545166016,
      "num_tokens": 175209049.0,
      "step": 5033
    },
    {
      "epoch": 0.9348189415041783,
      "grad_norm": 1.4511735035484208,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8711487054824829,
      "num_tokens": 175246503.0,
      "step": 5034
    },
    {
      "epoch": 0.9350046425255338,
      "grad_norm": 1.408685615326733,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8760203123092651,
      "num_tokens": 175282914.0,
      "step": 5035
    },
    {
      "epoch": 0.9351903435468895,
      "grad_norm": 1.4987332275848848,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8694225549697876,
      "num_tokens": 175315849.0,
      "step": 5036
    },
    {
      "epoch": 0.9353760445682451,
      "grad_norm": 1.3797234764878163,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8700898289680481,
      "num_tokens": 175352595.0,
      "step": 5037
    },
    {
      "epoch": 0.9355617455896007,
      "grad_norm": 1.3045054306054764,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8819382786750793,
      "num_tokens": 175392209.0,
      "step": 5038
    },
    {
      "epoch": 0.9357474466109563,
      "grad_norm": 1.4376956627031823,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8687387108802795,
      "num_tokens": 175426101.0,
      "step": 5039
    },
    {
      "epoch": 0.935933147632312,
      "grad_norm": 1.4932944541903794,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8633520603179932,
      "num_tokens": 175462164.0,
      "step": 5040
    },
    {
      "epoch": 0.9361188486536676,
      "grad_norm": 1.4294746301733017,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8660302758216858,
      "num_tokens": 175499357.0,
      "step": 5041
    },
    {
      "epoch": 0.9363045496750232,
      "grad_norm": 1.640969408200219,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8871669769287109,
      "num_tokens": 175526131.0,
      "step": 5042
    },
    {
      "epoch": 0.9364902506963788,
      "grad_norm": 1.366014738208076,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8626992702484131,
      "num_tokens": 175563544.0,
      "step": 5043
    },
    {
      "epoch": 0.9366759517177344,
      "grad_norm": 1.4990952203017125,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8857441544532776,
      "num_tokens": 175591480.0,
      "step": 5044
    },
    {
      "epoch": 0.9368616527390901,
      "grad_norm": 1.6621433006428858,
      "learning_rate": 1e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8539485931396484,
      "num_tokens": 175621321.0,
      "step": 5045
    },
    {
      "epoch": 0.9370473537604457,
      "grad_norm": 1.5092627669689205,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8666057586669922,
      "num_tokens": 175653552.0,
      "step": 5046
    },
    {
      "epoch": 0.9372330547818013,
      "grad_norm": 1.4790188600189684,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8682254552841187,
      "num_tokens": 175687275.0,
      "step": 5047
    },
    {
      "epoch": 0.9374187558031569,
      "grad_norm": 1.4446478047949283,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8659666180610657,
      "num_tokens": 175727158.0,
      "step": 5048
    },
    {
      "epoch": 0.9376044568245125,
      "grad_norm": 1.4304759256174397,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8681990504264832,
      "num_tokens": 175763453.0,
      "step": 5049
    },
    {
      "epoch": 0.9377901578458682,
      "grad_norm": 1.4601450776601523,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8636726140975952,
      "num_tokens": 175800182.0,
      "step": 5050
    },
    {
      "epoch": 0.9379758588672238,
      "grad_norm": 1.415204742922175,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8637974262237549,
      "num_tokens": 175838338.0,
      "step": 5051
    },
    {
      "epoch": 0.9381615598885794,
      "grad_norm": 1.561568676843093,
      "learning_rate": 1e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.861801266670227,
      "num_tokens": 175872749.0,
      "step": 5052
    },
    {
      "epoch": 0.938347260909935,
      "grad_norm": 1.4245281620397219,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8780429363250732,
      "num_tokens": 175914132.0,
      "step": 5053
    },
    {
      "epoch": 0.9385329619312907,
      "grad_norm": 1.4398723275495424,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8819316625595093,
      "num_tokens": 175946604.0,
      "step": 5054
    },
    {
      "epoch": 0.9387186629526463,
      "grad_norm": 1.4390579403387143,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.873293936252594,
      "num_tokens": 175982604.0,
      "step": 5055
    },
    {
      "epoch": 0.9389043639740019,
      "grad_norm": 1.3738009453610538,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8775174617767334,
      "num_tokens": 176018569.0,
      "step": 5056
    },
    {
      "epoch": 0.9390900649953575,
      "grad_norm": 1.787088457822525,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8787305355072021,
      "num_tokens": 176051237.0,
      "step": 5057
    },
    {
      "epoch": 0.9392757660167131,
      "grad_norm": 1.5415501885211635,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8685343861579895,
      "num_tokens": 176085498.0,
      "step": 5058
    },
    {
      "epoch": 0.9394614670380688,
      "grad_norm": 1.60011747851261,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8742125034332275,
      "num_tokens": 176114602.0,
      "step": 5059
    },
    {
      "epoch": 0.9396471680594243,
      "grad_norm": 1.4187504712384227,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8705041408538818,
      "num_tokens": 176147844.0,
      "step": 5060
    },
    {
      "epoch": 0.9398328690807799,
      "grad_norm": 1.5146113002839106,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8766757845878601,
      "num_tokens": 176177819.0,
      "step": 5061
    },
    {
      "epoch": 0.9400185701021355,
      "grad_norm": 1.504965027892784,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8693453669548035,
      "num_tokens": 176210799.0,
      "step": 5062
    },
    {
      "epoch": 0.9402042711234911,
      "grad_norm": 1.4848562243343253,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8699144124984741,
      "num_tokens": 176247844.0,
      "step": 5063
    },
    {
      "epoch": 0.9403899721448468,
      "grad_norm": 1.4466838315003414,
      "learning_rate": 1e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8569951057434082,
      "num_tokens": 176285468.0,
      "step": 5064
    },
    {
      "epoch": 0.9405756731662024,
      "grad_norm": 1.3568675912937398,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8626481294631958,
      "num_tokens": 176329129.0,
      "step": 5065
    },
    {
      "epoch": 0.940761374187558,
      "grad_norm": 1.5516499248552396,
      "learning_rate": 1e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8570525050163269,
      "num_tokens": 176364806.0,
      "step": 5066
    },
    {
      "epoch": 0.9409470752089136,
      "grad_norm": 1.5259316702429724,
      "learning_rate": 1e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.858841061592102,
      "num_tokens": 176399442.0,
      "step": 5067
    },
    {
      "epoch": 0.9411327762302693,
      "grad_norm": 1.5080796382256467,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8677471876144409,
      "num_tokens": 176430229.0,
      "step": 5068
    },
    {
      "epoch": 0.9413184772516249,
      "grad_norm": 1.4954979550747738,
      "learning_rate": 1e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8630934953689575,
      "num_tokens": 176465741.0,
      "step": 5069
    },
    {
      "epoch": 0.9415041782729805,
      "grad_norm": 1.3778879312903467,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.880681037902832,
      "num_tokens": 176499790.0,
      "step": 5070
    },
    {
      "epoch": 0.9416898792943361,
      "grad_norm": 1.4223008601788765,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8778814077377319,
      "num_tokens": 176533180.0,
      "step": 5071
    },
    {
      "epoch": 0.9418755803156917,
      "grad_norm": 1.4441462198059634,
      "learning_rate": 1e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.860901951789856,
      "num_tokens": 176570361.0,
      "step": 5072
    },
    {
      "epoch": 0.9420612813370474,
      "grad_norm": 1.5755316485324313,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8618598580360413,
      "num_tokens": 176601480.0,
      "step": 5073
    },
    {
      "epoch": 0.942246982358403,
      "grad_norm": 1.524762882091971,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8829265832901001,
      "num_tokens": 176634705.0,
      "step": 5074
    },
    {
      "epoch": 0.9424326833797586,
      "grad_norm": 1.4429222766358365,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8765726089477539,
      "num_tokens": 176668941.0,
      "step": 5075
    },
    {
      "epoch": 0.9426183844011142,
      "grad_norm": 1.3113173618380498,
      "learning_rate": 1e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8633735775947571,
      "num_tokens": 176712801.0,
      "step": 5076
    },
    {
      "epoch": 0.9428040854224699,
      "grad_norm": 1.3839053580617962,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8694404363632202,
      "num_tokens": 176755858.0,
      "step": 5077
    },
    {
      "epoch": 0.9429897864438255,
      "grad_norm": 1.3979538540731478,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.878064751625061,
      "num_tokens": 176789573.0,
      "step": 5078
    },
    {
      "epoch": 0.9431754874651811,
      "grad_norm": 1.3952211463964306,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8787643909454346,
      "num_tokens": 176828615.0,
      "step": 5079
    },
    {
      "epoch": 0.9433611884865367,
      "grad_norm": 1.677426770758008,
      "learning_rate": 1e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8713129162788391,
      "num_tokens": 176857911.0,
      "step": 5080
    },
    {
      "epoch": 0.9435468895078923,
      "grad_norm": 1.4609238642971314,
      "learning_rate": 1e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8528743982315063,
      "num_tokens": 176894650.0,
      "step": 5081
    },
    {
      "epoch": 0.943732590529248,
      "grad_norm": 1.522631635884013,
      "learning_rate": 1e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8370565176010132,
      "num_tokens": 176932516.0,
      "step": 5082
    },
    {
      "epoch": 0.9439182915506036,
      "grad_norm": 1.4563945467429875,
      "learning_rate": 1e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8511607050895691,
      "num_tokens": 176970897.0,
      "step": 5083
    },
    {
      "epoch": 0.9441039925719591,
      "grad_norm": 1.5181263621818444,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8740746974945068,
      "num_tokens": 177000157.0,
      "step": 5084
    },
    {
      "epoch": 0.9442896935933147,
      "grad_norm": 1.3756092396229034,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8864754438400269,
      "num_tokens": 177037433.0,
      "step": 5085
    },
    {
      "epoch": 0.9444753946146703,
      "grad_norm": 1.435495442966244,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8675622344017029,
      "num_tokens": 177072549.0,
      "step": 5086
    },
    {
      "epoch": 0.944661095636026,
      "grad_norm": 1.412072942321495,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8741576671600342,
      "num_tokens": 177110064.0,
      "step": 5087
    },
    {
      "epoch": 0.9448467966573816,
      "grad_norm": 1.5343547424367054,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8746870756149292,
      "num_tokens": 177137850.0,
      "step": 5088
    },
    {
      "epoch": 0.9450324976787372,
      "grad_norm": 1.4123892734193766,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8740348815917969,
      "num_tokens": 177175962.0,
      "step": 5089
    },
    {
      "epoch": 0.9452181987000928,
      "grad_norm": 1.4519597994529765,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8716527223587036,
      "num_tokens": 177205888.0,
      "step": 5090
    },
    {
      "epoch": 0.9454038997214484,
      "grad_norm": 1.6295264798681521,
      "learning_rate": 1e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8463630080223083,
      "num_tokens": 177236533.0,
      "step": 5091
    },
    {
      "epoch": 0.9455896007428041,
      "grad_norm": 1.4670752502705704,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8651889562606812,
      "num_tokens": 177270839.0,
      "step": 5092
    },
    {
      "epoch": 0.9457753017641597,
      "grad_norm": 1.6545779165632064,
      "learning_rate": 1e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8565165996551514,
      "num_tokens": 177298344.0,
      "step": 5093
    },
    {
      "epoch": 0.9459610027855153,
      "grad_norm": 1.456658258980213,
      "learning_rate": 1e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8492765426635742,
      "num_tokens": 177336892.0,
      "step": 5094
    },
    {
      "epoch": 0.9461467038068709,
      "grad_norm": 1.4275816031834674,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8814849853515625,
      "num_tokens": 177372070.0,
      "step": 5095
    },
    {
      "epoch": 0.9463324048282266,
      "grad_norm": 1.421574104041457,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8702868223190308,
      "num_tokens": 177407849.0,
      "step": 5096
    },
    {
      "epoch": 0.9465181058495822,
      "grad_norm": 1.5036010561557327,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8675556182861328,
      "num_tokens": 177438800.0,
      "step": 5097
    },
    {
      "epoch": 0.9467038068709378,
      "grad_norm": 1.3381985570847044,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.883699357509613,
      "num_tokens": 177476458.0,
      "step": 5098
    },
    {
      "epoch": 0.9468895078922934,
      "grad_norm": 1.4246334577663262,
      "learning_rate": 1e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8510443568229675,
      "num_tokens": 177515287.0,
      "step": 5099
    },
    {
      "epoch": 0.947075208913649,
      "grad_norm": 1.5566223472524618,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8809967041015625,
      "num_tokens": 177542085.0,
      "step": 5100
    },
    {
      "epoch": 0.9472609099350047,
      "grad_norm": 1.508499869993718,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8595507740974426,
      "num_tokens": 177573763.0,
      "step": 5101
    },
    {
      "epoch": 0.9474466109563603,
      "grad_norm": 1.5573370975487844,
      "learning_rate": 1e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8538594245910645,
      "num_tokens": 177605088.0,
      "step": 5102
    },
    {
      "epoch": 0.9476323119777159,
      "grad_norm": 1.381446218874283,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8692842125892639,
      "num_tokens": 177644253.0,
      "step": 5103
    },
    {
      "epoch": 0.9478180129990715,
      "grad_norm": 1.420219748518766,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8796267509460449,
      "num_tokens": 177676637.0,
      "step": 5104
    },
    {
      "epoch": 0.9480037140204272,
      "grad_norm": 3.3659734154492287,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.861507773399353,
      "num_tokens": 177708915.0,
      "step": 5105
    },
    {
      "epoch": 0.9481894150417828,
      "grad_norm": 1.4325832631857371,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.882436990737915,
      "num_tokens": 177744247.0,
      "step": 5106
    },
    {
      "epoch": 0.9483751160631384,
      "grad_norm": 1.469987784601844,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8739075064659119,
      "num_tokens": 177775274.0,
      "step": 5107
    },
    {
      "epoch": 0.9485608170844939,
      "grad_norm": 1.4414258476515396,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8764525651931763,
      "num_tokens": 177809263.0,
      "step": 5108
    },
    {
      "epoch": 0.9487465181058495,
      "grad_norm": 1.656393456740434,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8726263046264648,
      "num_tokens": 177838736.0,
      "step": 5109
    },
    {
      "epoch": 0.9489322191272052,
      "grad_norm": 1.495876738500721,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.872301459312439,
      "num_tokens": 177873532.0,
      "step": 5110
    },
    {
      "epoch": 0.9491179201485608,
      "grad_norm": 1.2733797515412524,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8796055316925049,
      "num_tokens": 177916549.0,
      "step": 5111
    },
    {
      "epoch": 0.9493036211699164,
      "grad_norm": 1.3574527665119474,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8684959411621094,
      "num_tokens": 177954394.0,
      "step": 5112
    },
    {
      "epoch": 0.949489322191272,
      "grad_norm": 1.53371248329001,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8588293194770813,
      "num_tokens": 177990542.0,
      "step": 5113
    },
    {
      "epoch": 0.9496750232126276,
      "grad_norm": 1.4912839771942565,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8762103319168091,
      "num_tokens": 178026236.0,
      "step": 5114
    },
    {
      "epoch": 0.9498607242339833,
      "grad_norm": 1.6671301349291374,
      "learning_rate": 1e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8660520315170288,
      "num_tokens": 178054860.0,
      "step": 5115
    },
    {
      "epoch": 0.9500464252553389,
      "grad_norm": 1.4074905572088539,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8818674087524414,
      "num_tokens": 178092797.0,
      "step": 5116
    },
    {
      "epoch": 0.9502321262766945,
      "grad_norm": 1.8142111121295001,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8540836572647095,
      "num_tokens": 178118105.0,
      "step": 5117
    },
    {
      "epoch": 0.9504178272980501,
      "grad_norm": 1.4475001006461405,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.873139500617981,
      "num_tokens": 178150230.0,
      "step": 5118
    },
    {
      "epoch": 0.9506035283194058,
      "grad_norm": 1.6580228658579153,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8730829954147339,
      "num_tokens": 178183534.0,
      "step": 5119
    },
    {
      "epoch": 0.9507892293407614,
      "grad_norm": 1.4204900377325824,
      "learning_rate": 1e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8557089567184448,
      "num_tokens": 178220708.0,
      "step": 5120
    },
    {
      "epoch": 0.950974930362117,
      "grad_norm": 1.6453011573505913,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8706663846969604,
      "num_tokens": 178251067.0,
      "step": 5121
    },
    {
      "epoch": 0.9511606313834726,
      "grad_norm": 1.5564243850103898,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8654390573501587,
      "num_tokens": 178279962.0,
      "step": 5122
    },
    {
      "epoch": 0.9513463324048282,
      "grad_norm": 1.3456841237868942,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8869062662124634,
      "num_tokens": 178315262.0,
      "step": 5123
    },
    {
      "epoch": 0.9515320334261839,
      "grad_norm": 1.2305167940623405,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8819125890731812,
      "num_tokens": 178354048.0,
      "step": 5124
    },
    {
      "epoch": 0.9517177344475395,
      "grad_norm": 1.3359466058793865,
      "learning_rate": 1e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8710087537765503,
      "num_tokens": 178393755.0,
      "step": 5125
    },
    {
      "epoch": 0.9519034354688951,
      "grad_norm": 1.454672534451275,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8752543926239014,
      "num_tokens": 178424916.0,
      "step": 5126
    },
    {
      "epoch": 0.9520891364902507,
      "grad_norm": 1.4709228596921557,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8663564920425415,
      "num_tokens": 178458026.0,
      "step": 5127
    },
    {
      "epoch": 0.9522748375116064,
      "grad_norm": 1.4979405385042939,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8729948401451111,
      "num_tokens": 178490435.0,
      "step": 5128
    },
    {
      "epoch": 0.952460538532962,
      "grad_norm": 1.466602792698476,
      "learning_rate": 1e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8581674098968506,
      "num_tokens": 178525141.0,
      "step": 5129
    },
    {
      "epoch": 0.9526462395543176,
      "grad_norm": 1.5977142553858976,
      "learning_rate": 1e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8518288135528564,
      "num_tokens": 178561437.0,
      "step": 5130
    },
    {
      "epoch": 0.9528319405756732,
      "grad_norm": 1.4807193599301336,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8837385177612305,
      "num_tokens": 178594262.0,
      "step": 5131
    },
    {
      "epoch": 0.9530176415970287,
      "grad_norm": 1.5788525346038016,
      "learning_rate": 1e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.846097469329834,
      "num_tokens": 178626055.0,
      "step": 5132
    },
    {
      "epoch": 0.9532033426183844,
      "grad_norm": 1.5670886392704084,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8716379404067993,
      "num_tokens": 178661691.0,
      "step": 5133
    },
    {
      "epoch": 0.95338904363974,
      "grad_norm": 1.4397683381398763,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8651816844940186,
      "num_tokens": 178697362.0,
      "step": 5134
    },
    {
      "epoch": 0.9535747446610956,
      "grad_norm": 1.3489000860487173,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8906243443489075,
      "num_tokens": 178733140.0,
      "step": 5135
    },
    {
      "epoch": 0.9537604456824512,
      "grad_norm": 1.5034685749591765,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8803553581237793,
      "num_tokens": 178766075.0,
      "step": 5136
    },
    {
      "epoch": 0.9539461467038068,
      "grad_norm": 1.511742277096472,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8748977780342102,
      "num_tokens": 178795675.0,
      "step": 5137
    },
    {
      "epoch": 0.9541318477251625,
      "grad_norm": 1.3916369841166396,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8790336847305298,
      "num_tokens": 178828211.0,
      "step": 5138
    },
    {
      "epoch": 0.9543175487465181,
      "grad_norm": 1.4910893735975868,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.867216944694519,
      "num_tokens": 178861024.0,
      "step": 5139
    },
    {
      "epoch": 0.9545032497678737,
      "grad_norm": 1.529586138206765,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.86386638879776,
      "num_tokens": 178894814.0,
      "step": 5140
    },
    {
      "epoch": 0.9546889507892293,
      "grad_norm": 1.3540611863446468,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8722600340843201,
      "num_tokens": 178935570.0,
      "step": 5141
    },
    {
      "epoch": 0.954874651810585,
      "grad_norm": 1.4225495523668439,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8608583211898804,
      "num_tokens": 178973683.0,
      "step": 5142
    },
    {
      "epoch": 0.9550603528319406,
      "grad_norm": 1.5424948669865506,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8650583028793335,
      "num_tokens": 179004957.0,
      "step": 5143
    },
    {
      "epoch": 0.9552460538532962,
      "grad_norm": 1.409113894864568,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8753273487091064,
      "num_tokens": 179039477.0,
      "step": 5144
    },
    {
      "epoch": 0.9554317548746518,
      "grad_norm": 1.5126837537379922,
      "learning_rate": 1e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.849557638168335,
      "num_tokens": 179073341.0,
      "step": 5145
    },
    {
      "epoch": 0.9556174558960074,
      "grad_norm": 1.5240561693181376,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8756667375564575,
      "num_tokens": 179105359.0,
      "step": 5146
    },
    {
      "epoch": 0.9558031569173631,
      "grad_norm": 1.3339535845306785,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8818814754486084,
      "num_tokens": 179143491.0,
      "step": 5147
    },
    {
      "epoch": 0.9559888579387187,
      "grad_norm": 1.5083585209952706,
      "learning_rate": 1e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.860765278339386,
      "num_tokens": 179179944.0,
      "step": 5148
    },
    {
      "epoch": 0.9561745589600743,
      "grad_norm": 1.4263498703712603,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8632591366767883,
      "num_tokens": 179216434.0,
      "step": 5149
    },
    {
      "epoch": 0.9563602599814299,
      "grad_norm": 1.4454435636857144,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8763426542282104,
      "num_tokens": 179250682.0,
      "step": 5150
    },
    {
      "epoch": 0.9565459610027855,
      "grad_norm": 1.5511739476975062,
      "learning_rate": 1e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8597255945205688,
      "num_tokens": 179292024.0,
      "step": 5151
    },
    {
      "epoch": 0.9567316620241412,
      "grad_norm": 1.4608133369261707,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.890544593334198,
      "num_tokens": 179319625.0,
      "step": 5152
    },
    {
      "epoch": 0.9569173630454968,
      "grad_norm": 1.3938296624163253,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8813381791114807,
      "num_tokens": 179354861.0,
      "step": 5153
    },
    {
      "epoch": 0.9571030640668524,
      "grad_norm": 1.3681846145117493,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8772112131118774,
      "num_tokens": 179391564.0,
      "step": 5154
    },
    {
      "epoch": 0.957288765088208,
      "grad_norm": 1.513147964791183,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8788485527038574,
      "num_tokens": 179419118.0,
      "step": 5155
    },
    {
      "epoch": 0.9574744661095635,
      "grad_norm": 1.3506680751673141,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8844305276870728,
      "num_tokens": 179456109.0,
      "step": 5156
    },
    {
      "epoch": 0.9576601671309192,
      "grad_norm": 1.546759983555546,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8733271956443787,
      "num_tokens": 179484772.0,
      "step": 5157
    },
    {
      "epoch": 0.9578458681522748,
      "grad_norm": 1.4446563163878159,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8704805374145508,
      "num_tokens": 179517052.0,
      "step": 5158
    },
    {
      "epoch": 0.9580315691736304,
      "grad_norm": 1.5341797457747115,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8770543336868286,
      "num_tokens": 179549659.0,
      "step": 5159
    },
    {
      "epoch": 0.958217270194986,
      "grad_norm": 1.474206049464073,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.873410165309906,
      "num_tokens": 179580317.0,
      "step": 5160
    },
    {
      "epoch": 0.9584029712163417,
      "grad_norm": 1.2445984615879915,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.883557915687561,
      "num_tokens": 179621022.0,
      "step": 5161
    },
    {
      "epoch": 0.9585886722376973,
      "grad_norm": 1.351913747485346,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8745579719543457,
      "num_tokens": 179659711.0,
      "step": 5162
    },
    {
      "epoch": 0.9587743732590529,
      "grad_norm": 1.5877895228672803,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8735640048980713,
      "num_tokens": 179694566.0,
      "step": 5163
    },
    {
      "epoch": 0.9589600742804085,
      "grad_norm": 1.4974732777909288,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8725684881210327,
      "num_tokens": 179725216.0,
      "step": 5164
    },
    {
      "epoch": 0.9591457753017641,
      "grad_norm": 1.4372780210401466,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8667224645614624,
      "num_tokens": 179759216.0,
      "step": 5165
    },
    {
      "epoch": 0.9593314763231198,
      "grad_norm": 1.328653751318199,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8760777711868286,
      "num_tokens": 179796491.0,
      "step": 5166
    },
    {
      "epoch": 0.9595171773444754,
      "grad_norm": 1.3731533395491295,
      "learning_rate": 1e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8633008003234863,
      "num_tokens": 179836925.0,
      "step": 5167
    },
    {
      "epoch": 0.959702878365831,
      "grad_norm": 1.3828079837670888,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8776506185531616,
      "num_tokens": 179872203.0,
      "step": 5168
    },
    {
      "epoch": 0.9598885793871866,
      "grad_norm": 1.4379265831149328,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8810515403747559,
      "num_tokens": 179905430.0,
      "step": 5169
    },
    {
      "epoch": 0.9600742804085423,
      "grad_norm": 1.4871537228250313,
      "learning_rate": 1e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.847562313079834,
      "num_tokens": 179940640.0,
      "step": 5170
    },
    {
      "epoch": 0.9602599814298979,
      "grad_norm": 1.462816528460238,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8853398561477661,
      "num_tokens": 179971890.0,
      "step": 5171
    },
    {
      "epoch": 0.9604456824512535,
      "grad_norm": 1.4229794390263335,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8612481951713562,
      "num_tokens": 180009254.0,
      "step": 5172
    },
    {
      "epoch": 0.9606313834726091,
      "grad_norm": 1.5519662528111278,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.881971001625061,
      "num_tokens": 180038294.0,
      "step": 5173
    },
    {
      "epoch": 0.9608170844939647,
      "grad_norm": 1.469366816718349,
      "learning_rate": 1e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8558382987976074,
      "num_tokens": 180071981.0,
      "step": 5174
    },
    {
      "epoch": 0.9610027855153204,
      "grad_norm": 1.36580496931248,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8779584169387817,
      "num_tokens": 180106676.0,
      "step": 5175
    },
    {
      "epoch": 0.961188486536676,
      "grad_norm": 1.3193915565570584,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8830596804618835,
      "num_tokens": 180148002.0,
      "step": 5176
    },
    {
      "epoch": 0.9613741875580316,
      "grad_norm": 1.4340642829135988,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8684282302856445,
      "num_tokens": 180181411.0,
      "step": 5177
    },
    {
      "epoch": 0.9615598885793872,
      "grad_norm": 1.4418017973557486,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8833438158035278,
      "num_tokens": 180211691.0,
      "step": 5178
    },
    {
      "epoch": 0.9617455896007429,
      "grad_norm": 1.4779003587477388,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8692253828048706,
      "num_tokens": 180247157.0,
      "step": 5179
    },
    {
      "epoch": 0.9619312906220984,
      "grad_norm": 1.5258100123963485,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8742675185203552,
      "num_tokens": 180275762.0,
      "step": 5180
    },
    {
      "epoch": 0.962116991643454,
      "grad_norm": 1.3517686647407388,
      "learning_rate": 1e-06,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8611773252487183,
      "num_tokens": 180318894.0,
      "step": 5181
    },
    {
      "epoch": 0.9623026926648096,
      "grad_norm": 1.360916453009593,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8808163404464722,
      "num_tokens": 180355934.0,
      "step": 5182
    },
    {
      "epoch": 0.9624883936861652,
      "grad_norm": 1.4738276490616604,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8783445358276367,
      "num_tokens": 180390648.0,
      "step": 5183
    },
    {
      "epoch": 0.9626740947075209,
      "grad_norm": 1.5820236318980312,
      "learning_rate": 1e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.854619026184082,
      "num_tokens": 180424952.0,
      "step": 5184
    },
    {
      "epoch": 0.9628597957288765,
      "grad_norm": 1.366547667472241,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8742436766624451,
      "num_tokens": 180464056.0,
      "step": 5185
    },
    {
      "epoch": 0.9630454967502321,
      "grad_norm": 1.386172655776011,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8751630187034607,
      "num_tokens": 180504910.0,
      "step": 5186
    },
    {
      "epoch": 0.9632311977715877,
      "grad_norm": 1.4707546078495521,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8669087290763855,
      "num_tokens": 180535555.0,
      "step": 5187
    },
    {
      "epoch": 0.9634168987929433,
      "grad_norm": 1.59145305678027,
      "learning_rate": 1e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8561497926712036,
      "num_tokens": 180568231.0,
      "step": 5188
    },
    {
      "epoch": 0.963602599814299,
      "grad_norm": 1.3794824552141034,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8746963739395142,
      "num_tokens": 180602836.0,
      "step": 5189
    },
    {
      "epoch": 0.9637883008356546,
      "grad_norm": 1.3485466761174638,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.886520266532898,
      "num_tokens": 180639963.0,
      "step": 5190
    },
    {
      "epoch": 0.9639740018570102,
      "grad_norm": 1.4820296052435364,
      "learning_rate": 1e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8545575737953186,
      "num_tokens": 180676879.0,
      "step": 5191
    },
    {
      "epoch": 0.9641597028783658,
      "grad_norm": 1.5892467780359985,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8764322996139526,
      "num_tokens": 180705461.0,
      "step": 5192
    },
    {
      "epoch": 0.9643454038997215,
      "grad_norm": 1.38054964022802,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8674214482307434,
      "num_tokens": 180745371.0,
      "step": 5193
    },
    {
      "epoch": 0.9645311049210771,
      "grad_norm": 1.5913622716375224,
      "learning_rate": 1e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8669010400772095,
      "num_tokens": 180775155.0,
      "step": 5194
    },
    {
      "epoch": 0.9647168059424327,
      "grad_norm": 1.4822483746491648,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8761151432991028,
      "num_tokens": 180804506.0,
      "step": 5195
    },
    {
      "epoch": 0.9649025069637883,
      "grad_norm": 1.4305005570159213,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8758771419525146,
      "num_tokens": 180841153.0,
      "step": 5196
    },
    {
      "epoch": 0.9650882079851439,
      "grad_norm": 1.4135330278827518,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8812733888626099,
      "num_tokens": 180878109.0,
      "step": 5197
    },
    {
      "epoch": 0.9652739090064996,
      "grad_norm": 1.4430142861178354,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8767480850219727,
      "num_tokens": 180911262.0,
      "step": 5198
    },
    {
      "epoch": 0.9654596100278552,
      "grad_norm": 1.599799531965126,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8608559966087341,
      "num_tokens": 180941188.0,
      "step": 5199
    },
    {
      "epoch": 0.9656453110492108,
      "grad_norm": 1.4366252478090324,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8710223436355591,
      "num_tokens": 180977370.0,
      "step": 5200
    },
    {
      "epoch": 0.9658310120705664,
      "grad_norm": 1.3496959794550212,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8754196166992188,
      "num_tokens": 181013692.0,
      "step": 5201
    },
    {
      "epoch": 0.966016713091922,
      "grad_norm": 1.3867889533416322,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8736845254898071,
      "num_tokens": 181052195.0,
      "step": 5202
    },
    {
      "epoch": 0.9662024141132777,
      "grad_norm": 1.4519066336454396,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8659778237342834,
      "num_tokens": 181094006.0,
      "step": 5203
    },
    {
      "epoch": 0.9663881151346332,
      "grad_norm": 1.5897392125314367,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8673593401908875,
      "num_tokens": 181124338.0,
      "step": 5204
    },
    {
      "epoch": 0.9665738161559888,
      "grad_norm": 1.5554984784375705,
      "learning_rate": 1e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8640275001525879,
      "num_tokens": 181154620.0,
      "step": 5205
    },
    {
      "epoch": 0.9667595171773444,
      "grad_norm": 1.3601849071141852,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8592996597290039,
      "num_tokens": 181194126.0,
      "step": 5206
    },
    {
      "epoch": 0.9669452181987,
      "grad_norm": 1.5039609603104842,
      "learning_rate": 1e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8687334060668945,
      "num_tokens": 181227789.0,
      "step": 5207
    },
    {
      "epoch": 0.9671309192200557,
      "grad_norm": 1.4530884289776482,
      "learning_rate": 1e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8588649034500122,
      "num_tokens": 181263738.0,
      "step": 5208
    },
    {
      "epoch": 0.9673166202414113,
      "grad_norm": 1.4352781858574086,
      "learning_rate": 1e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8581112623214722,
      "num_tokens": 181298719.0,
      "step": 5209
    },
    {
      "epoch": 0.9675023212627669,
      "grad_norm": 1.4704589157653025,
      "learning_rate": 1e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8520793914794922,
      "num_tokens": 181331975.0,
      "step": 5210
    },
    {
      "epoch": 0.9676880222841225,
      "grad_norm": 1.5884884757591589,
      "learning_rate": 1e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8569790124893188,
      "num_tokens": 181365213.0,
      "step": 5211
    },
    {
      "epoch": 0.9678737233054782,
      "grad_norm": 1.4206008898238824,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8838109374046326,
      "num_tokens": 181400697.0,
      "step": 5212
    },
    {
      "epoch": 0.9680594243268338,
      "grad_norm": 1.449688107744415,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8733406662940979,
      "num_tokens": 181434241.0,
      "step": 5213
    },
    {
      "epoch": 0.9682451253481894,
      "grad_norm": 1.5172696170445445,
      "learning_rate": 1e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8617384433746338,
      "num_tokens": 181471743.0,
      "step": 5214
    },
    {
      "epoch": 0.968430826369545,
      "grad_norm": 1.2929519801852234,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8646522164344788,
      "num_tokens": 181516629.0,
      "step": 5215
    },
    {
      "epoch": 0.9686165273909007,
      "grad_norm": 1.3564327625374397,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8707970976829529,
      "num_tokens": 181558072.0,
      "step": 5216
    },
    {
      "epoch": 0.9688022284122563,
      "grad_norm": 1.5565046724599796,
      "learning_rate": 1e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8646848201751709,
      "num_tokens": 181591255.0,
      "step": 5217
    },
    {
      "epoch": 0.9689879294336119,
      "grad_norm": 1.516686968676039,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8827974796295166,
      "num_tokens": 181621559.0,
      "step": 5218
    },
    {
      "epoch": 0.9691736304549675,
      "grad_norm": 1.5831716493034422,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.870781660079956,
      "num_tokens": 181653177.0,
      "step": 5219
    },
    {
      "epoch": 0.9693593314763231,
      "grad_norm": 1.5959011824975289,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8723468780517578,
      "num_tokens": 181681926.0,
      "step": 5220
    },
    {
      "epoch": 0.9695450324976788,
      "grad_norm": 1.6306771736807744,
      "learning_rate": 1e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.856330156326294,
      "num_tokens": 181710485.0,
      "step": 5221
    },
    {
      "epoch": 0.9697307335190344,
      "grad_norm": 1.633079716014953,
      "learning_rate": 1e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8524424433708191,
      "num_tokens": 181739566.0,
      "step": 5222
    },
    {
      "epoch": 0.96991643454039,
      "grad_norm": 1.5286667257359476,
      "learning_rate": 1e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8588893413543701,
      "num_tokens": 181771467.0,
      "step": 5223
    },
    {
      "epoch": 0.9701021355617456,
      "grad_norm": 1.376381367293433,
      "learning_rate": 1e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8622418642044067,
      "num_tokens": 181815814.0,
      "step": 5224
    },
    {
      "epoch": 0.9702878365831012,
      "grad_norm": 1.4962060036066536,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8669748902320862,
      "num_tokens": 181848620.0,
      "step": 5225
    },
    {
      "epoch": 0.9704735376044569,
      "grad_norm": 1.3601808768399009,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8866272568702698,
      "num_tokens": 181883120.0,
      "step": 5226
    },
    {
      "epoch": 0.9706592386258125,
      "grad_norm": 1.768356354772317,
      "learning_rate": 1e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8601511716842651,
      "num_tokens": 181911861.0,
      "step": 5227
    },
    {
      "epoch": 0.9708449396471681,
      "grad_norm": 1.337865708494078,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8755043148994446,
      "num_tokens": 181956206.0,
      "step": 5228
    },
    {
      "epoch": 0.9710306406685236,
      "grad_norm": 1.4428701004929103,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.885312557220459,
      "num_tokens": 181985685.0,
      "step": 5229
    },
    {
      "epoch": 0.9712163416898792,
      "grad_norm": 1.531715529562484,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8581230640411377,
      "num_tokens": 182016952.0,
      "step": 5230
    },
    {
      "epoch": 0.9714020427112349,
      "grad_norm": 1.4418328049302813,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.865119218826294,
      "num_tokens": 182051405.0,
      "step": 5231
    },
    {
      "epoch": 0.9715877437325905,
      "grad_norm": 1.490557791996795,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8761610388755798,
      "num_tokens": 182084370.0,
      "step": 5232
    },
    {
      "epoch": 0.9717734447539461,
      "grad_norm": 1.4538645389121228,
      "learning_rate": 1e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8542587757110596,
      "num_tokens": 182121468.0,
      "step": 5233
    },
    {
      "epoch": 0.9719591457753017,
      "grad_norm": 1.4684817969925916,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8719372749328613,
      "num_tokens": 182155797.0,
      "step": 5234
    },
    {
      "epoch": 0.9721448467966574,
      "grad_norm": 1.537378664772036,
      "learning_rate": 1e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8594046831130981,
      "num_tokens": 182188894.0,
      "step": 5235
    },
    {
      "epoch": 0.972330547818013,
      "grad_norm": 1.3914201787627245,
      "learning_rate": 1e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8634226322174072,
      "num_tokens": 182227709.0,
      "step": 5236
    },
    {
      "epoch": 0.9725162488393686,
      "grad_norm": 1.6371683181339876,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8538267016410828,
      "num_tokens": 182256108.0,
      "step": 5237
    },
    {
      "epoch": 0.9727019498607242,
      "grad_norm": 1.3742179829860945,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8758029937744141,
      "num_tokens": 182289457.0,
      "step": 5238
    },
    {
      "epoch": 0.9728876508820798,
      "grad_norm": 1.541102898877976,
      "learning_rate": 1e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8562873601913452,
      "num_tokens": 182322235.0,
      "step": 5239
    },
    {
      "epoch": 0.9730733519034355,
      "grad_norm": 1.4500741515573528,
      "learning_rate": 1e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8451130390167236,
      "num_tokens": 182359875.0,
      "step": 5240
    },
    {
      "epoch": 0.9732590529247911,
      "grad_norm": 1.513822195910607,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8790024518966675,
      "num_tokens": 182395673.0,
      "step": 5241
    },
    {
      "epoch": 0.9734447539461467,
      "grad_norm": 1.60217703089017,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8695127367973328,
      "num_tokens": 182424081.0,
      "step": 5242
    },
    {
      "epoch": 0.9736304549675023,
      "grad_norm": 1.6184896257214882,
      "learning_rate": 1e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8640105128288269,
      "num_tokens": 182455394.0,
      "step": 5243
    },
    {
      "epoch": 0.973816155988858,
      "grad_norm": 1.5864199951489346,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8755393624305725,
      "num_tokens": 182490492.0,
      "step": 5244
    },
    {
      "epoch": 0.9740018570102136,
      "grad_norm": 1.4694931498216732,
      "learning_rate": 1e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8647181391716003,
      "num_tokens": 182523709.0,
      "step": 5245
    },
    {
      "epoch": 0.9741875580315692,
      "grad_norm": 1.4905038602720697,
      "learning_rate": 1e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8626618385314941,
      "num_tokens": 182559355.0,
      "step": 5246
    },
    {
      "epoch": 0.9743732590529248,
      "grad_norm": 1.4451907791609642,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8789243698120117,
      "num_tokens": 182596687.0,
      "step": 5247
    },
    {
      "epoch": 0.9745589600742804,
      "grad_norm": 1.5404457634836215,
      "learning_rate": 1e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8554854989051819,
      "num_tokens": 182632826.0,
      "step": 5248
    },
    {
      "epoch": 0.9747446610956361,
      "grad_norm": 1.3820242208115323,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8816211223602295,
      "num_tokens": 182669052.0,
      "step": 5249
    },
    {
      "epoch": 0.9749303621169917,
      "grad_norm": 1.5073251582971343,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8721352815628052,
      "num_tokens": 182705939.0,
      "step": 5250
    },
    {
      "epoch": 0.9751160631383473,
      "grad_norm": 1.4953925279893645,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8740259408950806,
      "num_tokens": 182736204.0,
      "step": 5251
    },
    {
      "epoch": 0.9753017641597029,
      "grad_norm": 1.4959305639963605,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8701989650726318,
      "num_tokens": 182769711.0,
      "step": 5252
    },
    {
      "epoch": 0.9754874651810584,
      "grad_norm": 1.3943635087087147,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8774909973144531,
      "num_tokens": 182807481.0,
      "step": 5253
    },
    {
      "epoch": 0.9756731662024141,
      "grad_norm": 1.428670122467262,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8728964924812317,
      "num_tokens": 182843790.0,
      "step": 5254
    },
    {
      "epoch": 0.9758588672237697,
      "grad_norm": 1.562253940584194,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8802976608276367,
      "num_tokens": 182875708.0,
      "step": 5255
    },
    {
      "epoch": 0.9760445682451253,
      "grad_norm": 1.459813487618455,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8616667985916138,
      "num_tokens": 182911596.0,
      "step": 5256
    },
    {
      "epoch": 0.9762302692664809,
      "grad_norm": 1.4890620559980248,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8709220290184021,
      "num_tokens": 182949076.0,
      "step": 5257
    },
    {
      "epoch": 0.9764159702878366,
      "grad_norm": 1.332869499698735,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8837383985519409,
      "num_tokens": 182986911.0,
      "step": 5258
    },
    {
      "epoch": 0.9766016713091922,
      "grad_norm": 1.399576345947275,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8689180612564087,
      "num_tokens": 183023622.0,
      "step": 5259
    },
    {
      "epoch": 0.9767873723305478,
      "grad_norm": 1.4209306707624376,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8807193040847778,
      "num_tokens": 183060295.0,
      "step": 5260
    },
    {
      "epoch": 0.9769730733519034,
      "grad_norm": 1.262634335321216,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8804668188095093,
      "num_tokens": 183100484.0,
      "step": 5261
    },
    {
      "epoch": 0.977158774373259,
      "grad_norm": 1.5385112514268093,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8778319954872131,
      "num_tokens": 183133829.0,
      "step": 5262
    },
    {
      "epoch": 0.9773444753946147,
      "grad_norm": 1.556513514749067,
      "learning_rate": 1e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.860080361366272,
      "num_tokens": 183164917.0,
      "step": 5263
    },
    {
      "epoch": 0.9775301764159703,
      "grad_norm": 1.3966595379744884,
      "learning_rate": 1e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8628020286560059,
      "num_tokens": 183203973.0,
      "step": 5264
    },
    {
      "epoch": 0.9777158774373259,
      "grad_norm": 1.465661484435703,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8620300889015198,
      "num_tokens": 183239349.0,
      "step": 5265
    },
    {
      "epoch": 0.9779015784586815,
      "grad_norm": 1.3035010571682744,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.880011796951294,
      "num_tokens": 183281318.0,
      "step": 5266
    },
    {
      "epoch": 0.9780872794800372,
      "grad_norm": 1.3158063410263565,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8697302341461182,
      "num_tokens": 183323310.0,
      "step": 5267
    },
    {
      "epoch": 0.9782729805013928,
      "grad_norm": 1.4532053187597487,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8778699636459351,
      "num_tokens": 183356704.0,
      "step": 5268
    },
    {
      "epoch": 0.9784586815227484,
      "grad_norm": 1.4340081534273907,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8734650015830994,
      "num_tokens": 183391666.0,
      "step": 5269
    },
    {
      "epoch": 0.978644382544104,
      "grad_norm": 1.5714372978649032,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8652931451797485,
      "num_tokens": 183424850.0,
      "step": 5270
    },
    {
      "epoch": 0.9788300835654596,
      "grad_norm": 1.4016303763771805,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.880826473236084,
      "num_tokens": 183460472.0,
      "step": 5271
    },
    {
      "epoch": 0.9790157845868153,
      "grad_norm": 1.4339715096541197,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.873717188835144,
      "num_tokens": 183492521.0,
      "step": 5272
    },
    {
      "epoch": 0.9792014856081709,
      "grad_norm": 1.404575321812096,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8782706260681152,
      "num_tokens": 183527632.0,
      "step": 5273
    },
    {
      "epoch": 0.9793871866295265,
      "grad_norm": 1.4593648155911267,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8662906885147095,
      "num_tokens": 183558068.0,
      "step": 5274
    },
    {
      "epoch": 0.9795728876508821,
      "grad_norm": 1.5777294381425049,
      "learning_rate": 1e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8486168384552002,
      "num_tokens": 183590171.0,
      "step": 5275
    },
    {
      "epoch": 0.9797585886722378,
      "grad_norm": 1.4121947627728006,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8792420029640198,
      "num_tokens": 183625238.0,
      "step": 5276
    },
    {
      "epoch": 0.9799442896935933,
      "grad_norm": 1.4445799703680406,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8653867244720459,
      "num_tokens": 183660568.0,
      "step": 5277
    },
    {
      "epoch": 0.9801299907149489,
      "grad_norm": 1.455513214784746,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8661949038505554,
      "num_tokens": 183693586.0,
      "step": 5278
    },
    {
      "epoch": 0.9803156917363045,
      "grad_norm": 1.5119716079269225,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8605435490608215,
      "num_tokens": 183726557.0,
      "step": 5279
    },
    {
      "epoch": 0.9805013927576601,
      "grad_norm": 1.4896016320687027,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.870835542678833,
      "num_tokens": 183758046.0,
      "step": 5280
    },
    {
      "epoch": 0.9806870937790158,
      "grad_norm": 1.5379162815606473,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8640949130058289,
      "num_tokens": 183788353.0,
      "step": 5281
    },
    {
      "epoch": 0.9808727948003714,
      "grad_norm": 1.4847388312461813,
      "learning_rate": 1e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8603363037109375,
      "num_tokens": 183822900.0,
      "step": 5282
    },
    {
      "epoch": 0.981058495821727,
      "grad_norm": 1.455333107245228,
      "learning_rate": 1e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8598327040672302,
      "num_tokens": 183859669.0,
      "step": 5283
    },
    {
      "epoch": 0.9812441968430826,
      "grad_norm": 1.4666760353068042,
      "learning_rate": 1e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8575747013092041,
      "num_tokens": 183894770.0,
      "step": 5284
    },
    {
      "epoch": 0.9814298978644382,
      "grad_norm": 1.5895557027220713,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.87237149477005,
      "num_tokens": 183924240.0,
      "step": 5285
    },
    {
      "epoch": 0.9816155988857939,
      "grad_norm": 1.419854350325578,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8670416474342346,
      "num_tokens": 183959981.0,
      "step": 5286
    },
    {
      "epoch": 0.9818012999071495,
      "grad_norm": 1.3735789100965377,
      "learning_rate": 1e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8511697053909302,
      "num_tokens": 184001420.0,
      "step": 5287
    },
    {
      "epoch": 0.9819870009285051,
      "grad_norm": 1.4838205215985099,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8690145015716553,
      "num_tokens": 184034412.0,
      "step": 5288
    },
    {
      "epoch": 0.9821727019498607,
      "grad_norm": 1.4629613842147515,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.869193434715271,
      "num_tokens": 184072181.0,
      "step": 5289
    },
    {
      "epoch": 0.9823584029712163,
      "grad_norm": 1.513346007287111,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8619872331619263,
      "num_tokens": 184109455.0,
      "step": 5290
    },
    {
      "epoch": 0.982544103992572,
      "grad_norm": 1.3825624276844226,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.880393922328949,
      "num_tokens": 184145126.0,
      "step": 5291
    },
    {
      "epoch": 0.9827298050139276,
      "grad_norm": 1.515478265170859,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8621777892112732,
      "num_tokens": 184179019.0,
      "step": 5292
    },
    {
      "epoch": 0.9829155060352832,
      "grad_norm": 1.3909487124643536,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.876311182975769,
      "num_tokens": 184215463.0,
      "step": 5293
    },
    {
      "epoch": 0.9831012070566388,
      "grad_norm": 1.5029562498534914,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8703212141990662,
      "num_tokens": 184249028.0,
      "step": 5294
    },
    {
      "epoch": 0.9832869080779945,
      "grad_norm": 1.4294448308650958,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8689699769020081,
      "num_tokens": 184287261.0,
      "step": 5295
    },
    {
      "epoch": 0.9834726090993501,
      "grad_norm": 1.4577163940365157,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8685152530670166,
      "num_tokens": 184322426.0,
      "step": 5296
    },
    {
      "epoch": 0.9836583101207057,
      "grad_norm": 1.4965781674365086,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8766711950302124,
      "num_tokens": 184351132.0,
      "step": 5297
    },
    {
      "epoch": 0.9838440111420613,
      "grad_norm": 1.3922049329111743,
      "learning_rate": 1e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8566503524780273,
      "num_tokens": 184391647.0,
      "step": 5298
    },
    {
      "epoch": 0.984029712163417,
      "grad_norm": 1.4310704942227332,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8731899261474609,
      "num_tokens": 184428209.0,
      "step": 5299
    },
    {
      "epoch": 0.9842154131847726,
      "grad_norm": 1.4738440502628978,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8743231296539307,
      "num_tokens": 184459360.0,
      "step": 5300
    },
    {
      "epoch": 0.9844011142061281,
      "grad_norm": 1.4177741010747396,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8727572560310364,
      "num_tokens": 184494640.0,
      "step": 5301
    },
    {
      "epoch": 0.9845868152274837,
      "grad_norm": 1.328414409993927,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8770344257354736,
      "num_tokens": 184535540.0,
      "step": 5302
    },
    {
      "epoch": 0.9847725162488393,
      "grad_norm": 1.5124706765143947,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8609037399291992,
      "num_tokens": 184569429.0,
      "step": 5303
    },
    {
      "epoch": 0.984958217270195,
      "grad_norm": 1.346549025251112,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8840857148170471,
      "num_tokens": 184608185.0,
      "step": 5304
    },
    {
      "epoch": 0.9851439182915506,
      "grad_norm": 1.50742073088859,
      "learning_rate": 1e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8592291474342346,
      "num_tokens": 184645568.0,
      "step": 5305
    },
    {
      "epoch": 0.9853296193129062,
      "grad_norm": 1.5916753655731077,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.859677791595459,
      "num_tokens": 184677788.0,
      "step": 5306
    },
    {
      "epoch": 0.9855153203342618,
      "grad_norm": 1.3976259668099174,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8828411102294922,
      "num_tokens": 184711833.0,
      "step": 5307
    },
    {
      "epoch": 0.9857010213556174,
      "grad_norm": 1.5042796822439974,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8758745789527893,
      "num_tokens": 184741897.0,
      "step": 5308
    },
    {
      "epoch": 0.9858867223769731,
      "grad_norm": 1.4816851531638167,
      "learning_rate": 1e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8587111234664917,
      "num_tokens": 184776860.0,
      "step": 5309
    },
    {
      "epoch": 0.9860724233983287,
      "grad_norm": 1.5702737154201794,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8657786846160889,
      "num_tokens": 184812909.0,
      "step": 5310
    },
    {
      "epoch": 0.9862581244196843,
      "grad_norm": 1.4299941396769753,
      "learning_rate": 1e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8621951341629028,
      "num_tokens": 184851351.0,
      "step": 5311
    },
    {
      "epoch": 0.9864438254410399,
      "grad_norm": 1.3655022173738118,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8736363649368286,
      "num_tokens": 184889633.0,
      "step": 5312
    },
    {
      "epoch": 0.9866295264623955,
      "grad_norm": 1.4840347064829926,
      "learning_rate": 1e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.851802408695221,
      "num_tokens": 184924385.0,
      "step": 5313
    },
    {
      "epoch": 0.9868152274837512,
      "grad_norm": 1.447694162546122,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.871977686882019,
      "num_tokens": 184957304.0,
      "step": 5314
    },
    {
      "epoch": 0.9870009285051068,
      "grad_norm": 1.3603030929572462,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8813267946243286,
      "num_tokens": 184993030.0,
      "step": 5315
    },
    {
      "epoch": 0.9871866295264624,
      "grad_norm": 1.5109760131930379,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8689531683921814,
      "num_tokens": 185029120.0,
      "step": 5316
    },
    {
      "epoch": 0.987372330547818,
      "grad_norm": 1.3711169242319952,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8688745498657227,
      "num_tokens": 185069474.0,
      "step": 5317
    },
    {
      "epoch": 0.9875580315691737,
      "grad_norm": 1.6191595240878456,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8731541633605957,
      "num_tokens": 185097050.0,
      "step": 5318
    },
    {
      "epoch": 0.9877437325905293,
      "grad_norm": 1.5205535949660351,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.870429515838623,
      "num_tokens": 185126868.0,
      "step": 5319
    },
    {
      "epoch": 0.9879294336118849,
      "grad_norm": 1.4554255119904933,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.874604344367981,
      "num_tokens": 185160664.0,
      "step": 5320
    },
    {
      "epoch": 0.9881151346332405,
      "grad_norm": 1.6706090903780253,
      "learning_rate": 1e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8647545576095581,
      "num_tokens": 185192607.0,
      "step": 5321
    },
    {
      "epoch": 0.9883008356545961,
      "grad_norm": 1.4547467771868043,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8698275685310364,
      "num_tokens": 185228075.0,
      "step": 5322
    },
    {
      "epoch": 0.9884865366759518,
      "grad_norm": 1.6697113585666112,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8655586838722229,
      "num_tokens": 185257451.0,
      "step": 5323
    },
    {
      "epoch": 0.9886722376973074,
      "grad_norm": 1.5805823502575405,
      "learning_rate": 1e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8556158542633057,
      "num_tokens": 185285992.0,
      "step": 5324
    },
    {
      "epoch": 0.9888579387186629,
      "grad_norm": 1.496136600024957,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8628129959106445,
      "num_tokens": 185319334.0,
      "step": 5325
    },
    {
      "epoch": 0.9890436397400185,
      "grad_norm": 1.354901305070098,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8634132146835327,
      "num_tokens": 185362869.0,
      "step": 5326
    },
    {
      "epoch": 0.9892293407613741,
      "grad_norm": 1.3303309906943377,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8832886219024658,
      "num_tokens": 185402601.0,
      "step": 5327
    },
    {
      "epoch": 0.9894150417827298,
      "grad_norm": 1.4814487364113802,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8764961361885071,
      "num_tokens": 185435361.0,
      "step": 5328
    },
    {
      "epoch": 0.9896007428040854,
      "grad_norm": 1.6067153202620155,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8674225807189941,
      "num_tokens": 185467151.0,
      "step": 5329
    },
    {
      "epoch": 0.989786443825441,
      "grad_norm": 1.362116442368359,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.871812641620636,
      "num_tokens": 185504651.0,
      "step": 5330
    },
    {
      "epoch": 0.9899721448467966,
      "grad_norm": 1.5415793574250374,
      "learning_rate": 1e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8595437407493591,
      "num_tokens": 185535531.0,
      "step": 5331
    },
    {
      "epoch": 0.9901578458681523,
      "grad_norm": 1.498033704003686,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8745483160018921,
      "num_tokens": 185574843.0,
      "step": 5332
    },
    {
      "epoch": 0.9903435468895079,
      "grad_norm": 1.424781742364614,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8805649280548096,
      "num_tokens": 185607580.0,
      "step": 5333
    },
    {
      "epoch": 0.9905292479108635,
      "grad_norm": 1.3377101951748553,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8781524300575256,
      "num_tokens": 185646130.0,
      "step": 5334
    },
    {
      "epoch": 0.9907149489322191,
      "grad_norm": 1.787002941402061,
      "learning_rate": 1e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8596359491348267,
      "num_tokens": 185671303.0,
      "step": 5335
    },
    {
      "epoch": 0.9909006499535747,
      "grad_norm": 1.5553107657601541,
      "learning_rate": 1e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8551939129829407,
      "num_tokens": 185706635.0,
      "step": 5336
    },
    {
      "epoch": 0.9910863509749304,
      "grad_norm": 1.4231110015250918,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8785194158554077,
      "num_tokens": 185739730.0,
      "step": 5337
    },
    {
      "epoch": 0.991272051996286,
      "grad_norm": 1.3091324488458604,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8786131143569946,
      "num_tokens": 185776330.0,
      "step": 5338
    },
    {
      "epoch": 0.9914577530176416,
      "grad_norm": 1.4504355581637949,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8713820576667786,
      "num_tokens": 185814365.0,
      "step": 5339
    },
    {
      "epoch": 0.9916434540389972,
      "grad_norm": 1.464900568712284,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8788385391235352,
      "num_tokens": 185847915.0,
      "step": 5340
    },
    {
      "epoch": 0.9918291550603529,
      "grad_norm": 1.4596653220372982,
      "learning_rate": 1e-06,
      "loss": 0.3993,
      "mean_token_accuracy": 0.8655881881713867,
      "num_tokens": 185878710.0,
      "step": 5341
    },
    {
      "epoch": 0.9920148560817085,
      "grad_norm": 1.4567904424573899,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8690477013587952,
      "num_tokens": 185914169.0,
      "step": 5342
    },
    {
      "epoch": 0.9922005571030641,
      "grad_norm": 1.5206760074390344,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8777714967727661,
      "num_tokens": 185947943.0,
      "step": 5343
    },
    {
      "epoch": 0.9923862581244197,
      "grad_norm": 1.3045939363741232,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8819653987884521,
      "num_tokens": 185988213.0,
      "step": 5344
    },
    {
      "epoch": 0.9925719591457753,
      "grad_norm": 1.370838657715902,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.882783055305481,
      "num_tokens": 186022198.0,
      "step": 5345
    },
    {
      "epoch": 0.992757660167131,
      "grad_norm": 1.3747152803048241,
      "learning_rate": 1e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.858363151550293,
      "num_tokens": 186065303.0,
      "step": 5346
    },
    {
      "epoch": 0.9929433611884866,
      "grad_norm": 1.6245521997345618,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8668444752693176,
      "num_tokens": 186094162.0,
      "step": 5347
    },
    {
      "epoch": 0.9931290622098422,
      "grad_norm": 1.4485145622095768,
      "learning_rate": 1e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8580599427223206,
      "num_tokens": 186132649.0,
      "step": 5348
    },
    {
      "epoch": 0.9933147632311977,
      "grad_norm": 1.392921407681395,
      "learning_rate": 1e-06,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8696988821029663,
      "num_tokens": 186170556.0,
      "step": 5349
    },
    {
      "epoch": 0.9935004642525533,
      "grad_norm": 1.58940203335785,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.867275595664978,
      "num_tokens": 186203112.0,
      "step": 5350
    },
    {
      "epoch": 0.993686165273909,
      "grad_norm": 1.6660304741846577,
      "learning_rate": 1e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8454746007919312,
      "num_tokens": 186236001.0,
      "step": 5351
    },
    {
      "epoch": 0.9938718662952646,
      "grad_norm": 1.5477147391794182,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8669290542602539,
      "num_tokens": 186269532.0,
      "step": 5352
    },
    {
      "epoch": 0.9940575673166202,
      "grad_norm": 1.2770230784542402,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.873397946357727,
      "num_tokens": 186309795.0,
      "step": 5353
    },
    {
      "epoch": 0.9942432683379758,
      "grad_norm": 1.4401104792602877,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8829611539840698,
      "num_tokens": 186340886.0,
      "step": 5354
    },
    {
      "epoch": 0.9944289693593314,
      "grad_norm": 1.3689573528461387,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8744583129882812,
      "num_tokens": 186380338.0,
      "step": 5355
    },
    {
      "epoch": 0.9946146703806871,
      "grad_norm": 1.4686609129757207,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8835290670394897,
      "num_tokens": 186409961.0,
      "step": 5356
    },
    {
      "epoch": 0.9948003714020427,
      "grad_norm": 1.5808629472832978,
      "learning_rate": 1e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8631321787834167,
      "num_tokens": 186440687.0,
      "step": 5357
    },
    {
      "epoch": 0.9949860724233983,
      "grad_norm": 1.4984111158597975,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8699551224708557,
      "num_tokens": 186473491.0,
      "step": 5358
    },
    {
      "epoch": 0.9951717734447539,
      "grad_norm": 1.494706224011065,
      "learning_rate": 1e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8576444387435913,
      "num_tokens": 186508351.0,
      "step": 5359
    },
    {
      "epoch": 0.9953574744661096,
      "grad_norm": 1.5276953042814683,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.873548150062561,
      "num_tokens": 186541964.0,
      "step": 5360
    },
    {
      "epoch": 0.9955431754874652,
      "grad_norm": 1.5570902995151543,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8768213987350464,
      "num_tokens": 186572277.0,
      "step": 5361
    },
    {
      "epoch": 0.9957288765088208,
      "grad_norm": 1.3580754623363378,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8897202610969543,
      "num_tokens": 186603989.0,
      "step": 5362
    },
    {
      "epoch": 0.9959145775301764,
      "grad_norm": 1.3944368023217357,
      "learning_rate": 1e-06,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8624963164329529,
      "num_tokens": 186642123.0,
      "step": 5363
    },
    {
      "epoch": 0.996100278551532,
      "grad_norm": 1.4945450440525483,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.867000162601471,
      "num_tokens": 186676030.0,
      "step": 5364
    },
    {
      "epoch": 0.9962859795728877,
      "grad_norm": 1.518560915113813,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8714302778244019,
      "num_tokens": 186706802.0,
      "step": 5365
    },
    {
      "epoch": 0.9964716805942433,
      "grad_norm": 1.649946068127402,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8687866926193237,
      "num_tokens": 186734174.0,
      "step": 5366
    },
    {
      "epoch": 0.9966573816155989,
      "grad_norm": 1.4470020058567254,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8743871450424194,
      "num_tokens": 186769153.0,
      "step": 5367
    },
    {
      "epoch": 0.9968430826369545,
      "grad_norm": 1.4616701178138818,
      "learning_rate": 1e-06,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8706605434417725,
      "num_tokens": 186802384.0,
      "step": 5368
    },
    {
      "epoch": 0.9970287836583102,
      "grad_norm": 1.5171483058184738,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8691132068634033,
      "num_tokens": 186836615.0,
      "step": 5369
    },
    {
      "epoch": 0.9972144846796658,
      "grad_norm": 1.5579318901542567,
      "learning_rate": 1e-06,
      "loss": 0.3866,
      "mean_token_accuracy": 0.868015706539154,
      "num_tokens": 186868702.0,
      "step": 5370
    },
    {
      "epoch": 0.9974001857010214,
      "grad_norm": 1.4220940471165955,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8764626979827881,
      "num_tokens": 186903793.0,
      "step": 5371
    },
    {
      "epoch": 0.997585886722377,
      "grad_norm": 1.522659303403885,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8672789931297302,
      "num_tokens": 186940743.0,
      "step": 5372
    },
    {
      "epoch": 0.9977715877437325,
      "grad_norm": 1.473567655653102,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.853444516658783,
      "num_tokens": 186976262.0,
      "step": 5373
    },
    {
      "epoch": 0.9979572887650882,
      "grad_norm": 1.4312032447098797,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8725038170814514,
      "num_tokens": 187012958.0,
      "step": 5374
    },
    {
      "epoch": 0.9981429897864438,
      "grad_norm": 1.447585434438601,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8829196691513062,
      "num_tokens": 187045430.0,
      "step": 5375
    },
    {
      "epoch": 0.9983286908077994,
      "grad_norm": 1.4427189831409275,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8730471134185791,
      "num_tokens": 187079753.0,
      "step": 5376
    },
    {
      "epoch": 0.998514391829155,
      "grad_norm": 1.4806048092337956,
      "learning_rate": 1e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8554080724716187,
      "num_tokens": 187116291.0,
      "step": 5377
    },
    {
      "epoch": 0.9987000928505106,
      "grad_norm": 1.4494666270200327,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8675468564033508,
      "num_tokens": 187152092.0,
      "step": 5378
    },
    {
      "epoch": 0.9988857938718663,
      "grad_norm": 1.5640583194294575,
      "learning_rate": 1e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.860032320022583,
      "num_tokens": 187181513.0,
      "step": 5379
    },
    {
      "epoch": 0.9990714948932219,
      "grad_norm": 1.460229884470604,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8754336833953857,
      "num_tokens": 187215630.0,
      "step": 5380
    },
    {
      "epoch": 0.9992571959145775,
      "grad_norm": 1.4171223589061772,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8906936645507812,
      "num_tokens": 187247415.0,
      "step": 5381
    },
    {
      "epoch": 0.9994428969359331,
      "grad_norm": 1.8026806780142044,
      "learning_rate": 1e-06,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8631201386451721,
      "num_tokens": 187271008.0,
      "step": 5382
    },
    {
      "epoch": 0.9996285979572888,
      "grad_norm": 1.3931392446422957,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8792871832847595,
      "num_tokens": 187305633.0,
      "step": 5383
    },
    {
      "epoch": 0.9998142989786444,
      "grad_norm": 1.352354691987615,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8642210960388184,
      "num_tokens": 187346834.0,
      "step": 5384
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4258032144690889,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8799322247505188,
      "num_tokens": 187381029.0,
      "step": 5385
    },
    {
      "epoch": 1.0001857010213555,
      "grad_norm": 1.384891352308735,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8831203579902649,
      "num_tokens": 187417179.0,
      "step": 5386
    },
    {
      "epoch": 1.0003714020427112,
      "grad_norm": 1.3573985876328107,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8806570768356323,
      "num_tokens": 187454300.0,
      "step": 5387
    },
    {
      "epoch": 1.0005571030640668,
      "grad_norm": 1.333946465825328,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.880804181098938,
      "num_tokens": 187492376.0,
      "step": 5388
    },
    {
      "epoch": 1.0007428040854225,
      "grad_norm": 1.4676313534756305,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8740615248680115,
      "num_tokens": 187524709.0,
      "step": 5389
    },
    {
      "epoch": 1.000928505106778,
      "grad_norm": 1.3940431648787799,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8758909702301025,
      "num_tokens": 187562454.0,
      "step": 5390
    },
    {
      "epoch": 1.0011142061281337,
      "grad_norm": 1.3110105775828866,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8772425651550293,
      "num_tokens": 187603179.0,
      "step": 5391
    },
    {
      "epoch": 1.0012999071494892,
      "grad_norm": 1.4734261896231304,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8711415529251099,
      "num_tokens": 187636705.0,
      "step": 5392
    },
    {
      "epoch": 1.001485608170845,
      "grad_norm": 1.646724686411303,
      "learning_rate": 1e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8627576231956482,
      "num_tokens": 187667147.0,
      "step": 5393
    },
    {
      "epoch": 1.0016713091922005,
      "grad_norm": 1.4619155667019346,
      "learning_rate": 1e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8630976676940918,
      "num_tokens": 187700634.0,
      "step": 5394
    },
    {
      "epoch": 1.0018570102135562,
      "grad_norm": 1.4656433907772621,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8773283958435059,
      "num_tokens": 187734974.0,
      "step": 5395
    },
    {
      "epoch": 1.0020427112349117,
      "grad_norm": 1.6143261704132985,
      "learning_rate": 1e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8526917099952698,
      "num_tokens": 187774078.0,
      "step": 5396
    },
    {
      "epoch": 1.0022284122562675,
      "grad_norm": 1.4605749046037928,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8922897577285767,
      "num_tokens": 187805157.0,
      "step": 5397
    },
    {
      "epoch": 1.002414113277623,
      "grad_norm": 1.3760175599459559,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8826181292533875,
      "num_tokens": 187843569.0,
      "step": 5398
    },
    {
      "epoch": 1.0025998142989787,
      "grad_norm": 1.5205654291849384,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8731333017349243,
      "num_tokens": 187879084.0,
      "step": 5399
    },
    {
      "epoch": 1.0027855153203342,
      "grad_norm": 1.5151658387171365,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8792580366134644,
      "num_tokens": 187911978.0,
      "step": 5400
    },
    {
      "epoch": 1.00297121634169,
      "grad_norm": 1.4681496027915801,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8848071098327637,
      "num_tokens": 187947014.0,
      "step": 5401
    },
    {
      "epoch": 1.0031569173630455,
      "grad_norm": 1.4635616336464146,
      "learning_rate": 1e-06,
      "loss": 0.3826,
      "mean_token_accuracy": 0.8717821836471558,
      "num_tokens": 187985463.0,
      "step": 5402
    },
    {
      "epoch": 1.0033426183844012,
      "grad_norm": 1.5851345821034284,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8904876708984375,
      "num_tokens": 188018140.0,
      "step": 5403
    },
    {
      "epoch": 1.0035283194057567,
      "grad_norm": 1.5733833507874424,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8795454502105713,
      "num_tokens": 188053695.0,
      "step": 5404
    },
    {
      "epoch": 1.0037140204271124,
      "grad_norm": 1.4390578933085454,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8764665126800537,
      "num_tokens": 188090928.0,
      "step": 5405
    },
    {
      "epoch": 1.003899721448468,
      "grad_norm": 1.3817087462012394,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8768130540847778,
      "num_tokens": 188127456.0,
      "step": 5406
    },
    {
      "epoch": 1.0040854224698237,
      "grad_norm": 1.5748533594864256,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8804872632026672,
      "num_tokens": 188164512.0,
      "step": 5407
    },
    {
      "epoch": 1.0042711234911792,
      "grad_norm": 1.467432157042419,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.861444354057312,
      "num_tokens": 188200638.0,
      "step": 5408
    },
    {
      "epoch": 1.004456824512535,
      "grad_norm": 1.3891047400697882,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8930822014808655,
      "num_tokens": 188236031.0,
      "step": 5409
    },
    {
      "epoch": 1.0046425255338904,
      "grad_norm": 1.4603754802028455,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8781877756118774,
      "num_tokens": 188273380.0,
      "step": 5410
    },
    {
      "epoch": 1.004828226555246,
      "grad_norm": 1.3914234201385456,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8699538111686707,
      "num_tokens": 188312864.0,
      "step": 5411
    },
    {
      "epoch": 1.0050139275766017,
      "grad_norm": 1.3818541499973325,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8728896379470825,
      "num_tokens": 188354228.0,
      "step": 5412
    },
    {
      "epoch": 1.0051996285979572,
      "grad_norm": 1.6447569661547052,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8752591609954834,
      "num_tokens": 188382913.0,
      "step": 5413
    },
    {
      "epoch": 1.005385329619313,
      "grad_norm": 1.3844761930994276,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8762611150741577,
      "num_tokens": 188421035.0,
      "step": 5414
    },
    {
      "epoch": 1.0055710306406684,
      "grad_norm": 1.5178507008313915,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8680486679077148,
      "num_tokens": 188457923.0,
      "step": 5415
    },
    {
      "epoch": 1.0057567316620242,
      "grad_norm": 1.5748001884812748,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8704915642738342,
      "num_tokens": 188493750.0,
      "step": 5416
    },
    {
      "epoch": 1.0059424326833797,
      "grad_norm": 1.4546589263503342,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8904819488525391,
      "num_tokens": 188529256.0,
      "step": 5417
    },
    {
      "epoch": 1.0061281337047354,
      "grad_norm": 1.508341676039651,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8755509257316589,
      "num_tokens": 188562496.0,
      "step": 5418
    },
    {
      "epoch": 1.006313834726091,
      "grad_norm": 1.439600054245609,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8755488395690918,
      "num_tokens": 188604693.0,
      "step": 5419
    },
    {
      "epoch": 1.0064995357474467,
      "grad_norm": 1.4326789050850288,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8576366305351257,
      "num_tokens": 188644268.0,
      "step": 5420
    },
    {
      "epoch": 1.0066852367688022,
      "grad_norm": 1.4973870042663835,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8763195276260376,
      "num_tokens": 188679116.0,
      "step": 5421
    },
    {
      "epoch": 1.006870937790158,
      "grad_norm": 1.53786412380074,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8904561400413513,
      "num_tokens": 188710806.0,
      "step": 5422
    },
    {
      "epoch": 1.0070566388115134,
      "grad_norm": 1.5067765355049934,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.873640239238739,
      "num_tokens": 188743804.0,
      "step": 5423
    },
    {
      "epoch": 1.0072423398328691,
      "grad_norm": 1.53745648384832,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8735921382904053,
      "num_tokens": 188775628.0,
      "step": 5424
    },
    {
      "epoch": 1.0074280408542247,
      "grad_norm": 1.5090045978365279,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8703871965408325,
      "num_tokens": 188811653.0,
      "step": 5425
    },
    {
      "epoch": 1.0076137418755804,
      "grad_norm": 1.5652899442112373,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8761060237884521,
      "num_tokens": 188845478.0,
      "step": 5426
    },
    {
      "epoch": 1.007799442896936,
      "grad_norm": 1.5704337492342815,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.869988739490509,
      "num_tokens": 188877206.0,
      "step": 5427
    },
    {
      "epoch": 1.0079851439182916,
      "grad_norm": 1.5031631194532478,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8738652467727661,
      "num_tokens": 188912724.0,
      "step": 5428
    },
    {
      "epoch": 1.0081708449396471,
      "grad_norm": 1.3440216946521517,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8896874785423279,
      "num_tokens": 188949968.0,
      "step": 5429
    },
    {
      "epoch": 1.0083565459610029,
      "grad_norm": 1.6639577511763262,
      "learning_rate": 1e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8646592497825623,
      "num_tokens": 188982248.0,
      "step": 5430
    },
    {
      "epoch": 1.0085422469823584,
      "grad_norm": 1.5959752141572654,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8667605519294739,
      "num_tokens": 189017601.0,
      "step": 5431
    },
    {
      "epoch": 1.0087279480037141,
      "grad_norm": 1.4865337257421296,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8665112853050232,
      "num_tokens": 189051477.0,
      "step": 5432
    },
    {
      "epoch": 1.0089136490250696,
      "grad_norm": 1.471776092675321,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8863089680671692,
      "num_tokens": 189086887.0,
      "step": 5433
    },
    {
      "epoch": 1.0090993500464251,
      "grad_norm": 1.2688703769106826,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8883622884750366,
      "num_tokens": 189127420.0,
      "step": 5434
    },
    {
      "epoch": 1.0092850510677809,
      "grad_norm": 1.610233927965728,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8655463457107544,
      "num_tokens": 189156246.0,
      "step": 5435
    },
    {
      "epoch": 1.0094707520891364,
      "grad_norm": 1.5963927155313042,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8780941963195801,
      "num_tokens": 189184566.0,
      "step": 5436
    },
    {
      "epoch": 1.0096564531104921,
      "grad_norm": 1.4925444135451529,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8599342107772827,
      "num_tokens": 189223414.0,
      "step": 5437
    },
    {
      "epoch": 1.0098421541318476,
      "grad_norm": 1.3199968352476439,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8834221363067627,
      "num_tokens": 189267487.0,
      "step": 5438
    },
    {
      "epoch": 1.0100278551532034,
      "grad_norm": 1.6341417261398612,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.880820631980896,
      "num_tokens": 189295330.0,
      "step": 5439
    },
    {
      "epoch": 1.0102135561745589,
      "grad_norm": 1.59879479196224,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8686582446098328,
      "num_tokens": 189330023.0,
      "step": 5440
    },
    {
      "epoch": 1.0103992571959146,
      "grad_norm": 1.4060448908006853,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8635585308074951,
      "num_tokens": 189371798.0,
      "step": 5441
    },
    {
      "epoch": 1.0105849582172701,
      "grad_norm": 1.5350146672369787,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8731484413146973,
      "num_tokens": 189404283.0,
      "step": 5442
    },
    {
      "epoch": 1.0107706592386259,
      "grad_norm": 1.6643153292407584,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8669312596321106,
      "num_tokens": 189434295.0,
      "step": 5443
    },
    {
      "epoch": 1.0109563602599814,
      "grad_norm": 1.6344566990860345,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8709850311279297,
      "num_tokens": 189464703.0,
      "step": 5444
    },
    {
      "epoch": 1.011142061281337,
      "grad_norm": 1.4932541500216272,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8615347146987915,
      "num_tokens": 189500291.0,
      "step": 5445
    },
    {
      "epoch": 1.0113277623026926,
      "grad_norm": 1.524476581552034,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8760664463043213,
      "num_tokens": 189534179.0,
      "step": 5446
    },
    {
      "epoch": 1.0115134633240483,
      "grad_norm": 1.564675527245939,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8670043349266052,
      "num_tokens": 189565020.0,
      "step": 5447
    },
    {
      "epoch": 1.0116991643454039,
      "grad_norm": 1.5544655969301608,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8823996782302856,
      "num_tokens": 189594904.0,
      "step": 5448
    },
    {
      "epoch": 1.0118848653667596,
      "grad_norm": 1.639159313989633,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8863165974617004,
      "num_tokens": 189623874.0,
      "step": 5449
    },
    {
      "epoch": 1.012070566388115,
      "grad_norm": 1.5808985149413375,
      "learning_rate": 1e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8649477958679199,
      "num_tokens": 189655798.0,
      "step": 5450
    },
    {
      "epoch": 1.0122562674094708,
      "grad_norm": 1.469590766985143,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8761333227157593,
      "num_tokens": 189689265.0,
      "step": 5451
    },
    {
      "epoch": 1.0124419684308263,
      "grad_norm": 1.3926208695243896,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.879631519317627,
      "num_tokens": 189724545.0,
      "step": 5452
    },
    {
      "epoch": 1.012627669452182,
      "grad_norm": 1.4106604274776828,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8785465955734253,
      "num_tokens": 189762682.0,
      "step": 5453
    },
    {
      "epoch": 1.0128133704735376,
      "grad_norm": 1.354794468749866,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8837412595748901,
      "num_tokens": 189803505.0,
      "step": 5454
    },
    {
      "epoch": 1.0129990714948933,
      "grad_norm": 1.4598065992274556,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8753484487533569,
      "num_tokens": 189841998.0,
      "step": 5455
    },
    {
      "epoch": 1.0131847725162488,
      "grad_norm": 1.4312786268232023,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8835963010787964,
      "num_tokens": 189877902.0,
      "step": 5456
    },
    {
      "epoch": 1.0133704735376046,
      "grad_norm": 1.5892487400551913,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.870722770690918,
      "num_tokens": 189910586.0,
      "step": 5457
    },
    {
      "epoch": 1.01355617455896,
      "grad_norm": 1.4708009183007413,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8746852874755859,
      "num_tokens": 189945091.0,
      "step": 5458
    },
    {
      "epoch": 1.0137418755803156,
      "grad_norm": 1.3550643317382798,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8809393644332886,
      "num_tokens": 189985293.0,
      "step": 5459
    },
    {
      "epoch": 1.0139275766016713,
      "grad_norm": 1.4557452450531998,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.874344527721405,
      "num_tokens": 190019221.0,
      "step": 5460
    },
    {
      "epoch": 1.0141132776230268,
      "grad_norm": 1.4295458726115875,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8824917078018188,
      "num_tokens": 190054002.0,
      "step": 5461
    },
    {
      "epoch": 1.0142989786443826,
      "grad_norm": 1.6681131287627746,
      "learning_rate": 1e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8529481291770935,
      "num_tokens": 190087074.0,
      "step": 5462
    },
    {
      "epoch": 1.014484679665738,
      "grad_norm": 1.433515590006609,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8798010349273682,
      "num_tokens": 190122764.0,
      "step": 5463
    },
    {
      "epoch": 1.0146703806870938,
      "grad_norm": 1.9212925892213653,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8743084669113159,
      "num_tokens": 190156253.0,
      "step": 5464
    },
    {
      "epoch": 1.0148560817084493,
      "grad_norm": 1.7086398727793284,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8779007196426392,
      "num_tokens": 190187319.0,
      "step": 5465
    },
    {
      "epoch": 1.015041782729805,
      "grad_norm": 1.5567862022870986,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8728312849998474,
      "num_tokens": 190221673.0,
      "step": 5466
    },
    {
      "epoch": 1.0152274837511606,
      "grad_norm": 1.6340777898660848,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8771852254867554,
      "num_tokens": 190251275.0,
      "step": 5467
    },
    {
      "epoch": 1.0154131847725163,
      "grad_norm": 1.5363868347374465,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8764075040817261,
      "num_tokens": 190289682.0,
      "step": 5468
    },
    {
      "epoch": 1.0155988857938718,
      "grad_norm": 1.6100035331319205,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8872018456459045,
      "num_tokens": 190320690.0,
      "step": 5469
    },
    {
      "epoch": 1.0157845868152275,
      "grad_norm": 1.6167515715839795,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8816812038421631,
      "num_tokens": 190349170.0,
      "step": 5470
    },
    {
      "epoch": 1.015970287836583,
      "grad_norm": 1.4406400466298257,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8796216249465942,
      "num_tokens": 190384794.0,
      "step": 5471
    },
    {
      "epoch": 1.0161559888579388,
      "grad_norm": 1.7529175559613288,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8768744468688965,
      "num_tokens": 190418576.0,
      "step": 5472
    },
    {
      "epoch": 1.0163416898792943,
      "grad_norm": 1.5037766437933198,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8774095177650452,
      "num_tokens": 190450578.0,
      "step": 5473
    },
    {
      "epoch": 1.01652739090065,
      "grad_norm": 1.7530863943135548,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8785715699195862,
      "num_tokens": 190474761.0,
      "step": 5474
    },
    {
      "epoch": 1.0167130919220055,
      "grad_norm": 1.4921656215927073,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.874452531337738,
      "num_tokens": 190511832.0,
      "step": 5475
    },
    {
      "epoch": 1.0168987929433613,
      "grad_norm": 1.5089015837109,
      "learning_rate": 1e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8511818647384644,
      "num_tokens": 190551936.0,
      "step": 5476
    },
    {
      "epoch": 1.0170844939647168,
      "grad_norm": 1.4767338046664364,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8686044216156006,
      "num_tokens": 190587918.0,
      "step": 5477
    },
    {
      "epoch": 1.0172701949860725,
      "grad_norm": 1.4890967371565549,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8691906929016113,
      "num_tokens": 190626433.0,
      "step": 5478
    },
    {
      "epoch": 1.017455896007428,
      "grad_norm": 1.4250380167395997,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8860892057418823,
      "num_tokens": 190664160.0,
      "step": 5479
    },
    {
      "epoch": 1.0176415970287838,
      "grad_norm": 1.6861399698549855,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8670910596847534,
      "num_tokens": 190691511.0,
      "step": 5480
    },
    {
      "epoch": 1.0178272980501393,
      "grad_norm": 1.4377010059741386,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8803905844688416,
      "num_tokens": 190733734.0,
      "step": 5481
    },
    {
      "epoch": 1.0180129990714948,
      "grad_norm": 1.5086152058792246,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8867887258529663,
      "num_tokens": 190768012.0,
      "step": 5482
    },
    {
      "epoch": 1.0181987000928505,
      "grad_norm": 1.5765279443275275,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8828127980232239,
      "num_tokens": 190799329.0,
      "step": 5483
    },
    {
      "epoch": 1.018384401114206,
      "grad_norm": 1.729924453972478,
      "learning_rate": 1e-06,
      "loss": 0.3826,
      "mean_token_accuracy": 0.8742697238922119,
      "num_tokens": 190831149.0,
      "step": 5484
    },
    {
      "epoch": 1.0185701021355618,
      "grad_norm": 1.484477716584912,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8709107041358948,
      "num_tokens": 190869591.0,
      "step": 5485
    },
    {
      "epoch": 1.0187558031569173,
      "grad_norm": 1.463108515848865,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8744317293167114,
      "num_tokens": 190906413.0,
      "step": 5486
    },
    {
      "epoch": 1.018941504178273,
      "grad_norm": 1.4962663570249661,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8889515399932861,
      "num_tokens": 190938317.0,
      "step": 5487
    },
    {
      "epoch": 1.0191272051996285,
      "grad_norm": 1.4966665449706968,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8780263066291809,
      "num_tokens": 190977375.0,
      "step": 5488
    },
    {
      "epoch": 1.0193129062209842,
      "grad_norm": 1.4617916120707026,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8883042335510254,
      "num_tokens": 191009927.0,
      "step": 5489
    },
    {
      "epoch": 1.0194986072423398,
      "grad_norm": 1.4420101748888554,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8850394487380981,
      "num_tokens": 191044298.0,
      "step": 5490
    },
    {
      "epoch": 1.0196843082636955,
      "grad_norm": 1.4850369028716583,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8854778409004211,
      "num_tokens": 191080308.0,
      "step": 5491
    },
    {
      "epoch": 1.019870009285051,
      "grad_norm": 1.4805172307430017,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8930332660675049,
      "num_tokens": 191109807.0,
      "step": 5492
    },
    {
      "epoch": 1.0200557103064067,
      "grad_norm": 1.592395514179499,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.864529550075531,
      "num_tokens": 191148570.0,
      "step": 5493
    },
    {
      "epoch": 1.0202414113277622,
      "grad_norm": 1.4979838081150627,
      "learning_rate": 1e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8489953279495239,
      "num_tokens": 191188459.0,
      "step": 5494
    },
    {
      "epoch": 1.020427112349118,
      "grad_norm": 1.4836632863346348,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8747385740280151,
      "num_tokens": 191224604.0,
      "step": 5495
    },
    {
      "epoch": 1.0206128133704735,
      "grad_norm": 1.5231946453513785,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8745846748352051,
      "num_tokens": 191259707.0,
      "step": 5496
    },
    {
      "epoch": 1.0207985143918292,
      "grad_norm": 1.4361411574373382,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8783644437789917,
      "num_tokens": 191297456.0,
      "step": 5497
    },
    {
      "epoch": 1.0209842154131847,
      "grad_norm": 1.294317578871592,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8847545385360718,
      "num_tokens": 191340039.0,
      "step": 5498
    },
    {
      "epoch": 1.0211699164345405,
      "grad_norm": 1.5622639496730855,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.88990718126297,
      "num_tokens": 191369551.0,
      "step": 5499
    },
    {
      "epoch": 1.021355617455896,
      "grad_norm": 1.497466537891964,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.871431291103363,
      "num_tokens": 191405559.0,
      "step": 5500
    },
    {
      "epoch": 1.0215413184772517,
      "grad_norm": 1.4419145798809243,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8724179267883301,
      "num_tokens": 191441493.0,
      "step": 5501
    },
    {
      "epoch": 1.0217270194986072,
      "grad_norm": 1.473082476842649,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.871476411819458,
      "num_tokens": 191477685.0,
      "step": 5502
    },
    {
      "epoch": 1.021912720519963,
      "grad_norm": 1.5248511727402179,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8737855553627014,
      "num_tokens": 191510529.0,
      "step": 5503
    },
    {
      "epoch": 1.0220984215413185,
      "grad_norm": 1.4401662399375725,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8826748728752136,
      "num_tokens": 191546721.0,
      "step": 5504
    },
    {
      "epoch": 1.0222841225626742,
      "grad_norm": 1.5550992499713558,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8728259205818176,
      "num_tokens": 191578989.0,
      "step": 5505
    },
    {
      "epoch": 1.0224698235840297,
      "grad_norm": 1.4488194566132588,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8859587907791138,
      "num_tokens": 191612932.0,
      "step": 5506
    },
    {
      "epoch": 1.0226555246053852,
      "grad_norm": 1.4022599819644577,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8683009743690491,
      "num_tokens": 191654404.0,
      "step": 5507
    },
    {
      "epoch": 1.022841225626741,
      "grad_norm": 1.4306813296800576,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8795918226242065,
      "num_tokens": 191691601.0,
      "step": 5508
    },
    {
      "epoch": 1.0230269266480965,
      "grad_norm": 1.661057036037126,
      "learning_rate": 1e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8574178218841553,
      "num_tokens": 191723661.0,
      "step": 5509
    },
    {
      "epoch": 1.0232126276694522,
      "grad_norm": 1.6611003969241007,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8611118197441101,
      "num_tokens": 191757243.0,
      "step": 5510
    },
    {
      "epoch": 1.0233983286908077,
      "grad_norm": 1.6248775525741515,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8741840124130249,
      "num_tokens": 191784208.0,
      "step": 5511
    },
    {
      "epoch": 1.0235840297121634,
      "grad_norm": 1.434938259003509,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8869732618331909,
      "num_tokens": 191822100.0,
      "step": 5512
    },
    {
      "epoch": 1.023769730733519,
      "grad_norm": 1.4414991364287457,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8721249103546143,
      "num_tokens": 191858367.0,
      "step": 5513
    },
    {
      "epoch": 1.0239554317548747,
      "grad_norm": 1.4908091392396894,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.879765510559082,
      "num_tokens": 191893507.0,
      "step": 5514
    },
    {
      "epoch": 1.0241411327762302,
      "grad_norm": 1.462386432899453,
      "learning_rate": 1e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8587620258331299,
      "num_tokens": 191931395.0,
      "step": 5515
    },
    {
      "epoch": 1.024326833797586,
      "grad_norm": 1.456021027917355,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8741523623466492,
      "num_tokens": 191968996.0,
      "step": 5516
    },
    {
      "epoch": 1.0245125348189414,
      "grad_norm": 1.474958259379106,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8765007257461548,
      "num_tokens": 192003984.0,
      "step": 5517
    },
    {
      "epoch": 1.0246982358402972,
      "grad_norm": 1.4372464307166588,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8713836669921875,
      "num_tokens": 192045370.0,
      "step": 5518
    },
    {
      "epoch": 1.0248839368616527,
      "grad_norm": 1.5230290927800567,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8780449032783508,
      "num_tokens": 192076973.0,
      "step": 5519
    },
    {
      "epoch": 1.0250696378830084,
      "grad_norm": 1.5294511375087534,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8672890067100525,
      "num_tokens": 192113169.0,
      "step": 5520
    },
    {
      "epoch": 1.025255338904364,
      "grad_norm": 1.4409808136319155,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8866040110588074,
      "num_tokens": 192143904.0,
      "step": 5521
    },
    {
      "epoch": 1.0254410399257197,
      "grad_norm": 1.4813961137867913,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8796662092208862,
      "num_tokens": 192177291.0,
      "step": 5522
    },
    {
      "epoch": 1.0256267409470752,
      "grad_norm": 1.3953779295555608,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8660730123519897,
      "num_tokens": 192213911.0,
      "step": 5523
    },
    {
      "epoch": 1.025812441968431,
      "grad_norm": 1.439851257778171,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8855515718460083,
      "num_tokens": 192245763.0,
      "step": 5524
    },
    {
      "epoch": 1.0259981429897864,
      "grad_norm": 1.5471117639449024,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8727204203605652,
      "num_tokens": 192282160.0,
      "step": 5525
    },
    {
      "epoch": 1.0261838440111422,
      "grad_norm": 1.505259663604954,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.893731951713562,
      "num_tokens": 192309317.0,
      "step": 5526
    },
    {
      "epoch": 1.0263695450324977,
      "grad_norm": 1.5259006380277929,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8780463933944702,
      "num_tokens": 192346205.0,
      "step": 5527
    },
    {
      "epoch": 1.0265552460538534,
      "grad_norm": 1.4350167955047441,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8713870644569397,
      "num_tokens": 192388984.0,
      "step": 5528
    },
    {
      "epoch": 1.026740947075209,
      "grad_norm": 1.505793083105542,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8811304569244385,
      "num_tokens": 192421505.0,
      "step": 5529
    },
    {
      "epoch": 1.0269266480965644,
      "grad_norm": 1.5709903203630708,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8616049289703369,
      "num_tokens": 192454919.0,
      "step": 5530
    },
    {
      "epoch": 1.0271123491179202,
      "grad_norm": 1.6029458900678053,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8701840043067932,
      "num_tokens": 192489571.0,
      "step": 5531
    },
    {
      "epoch": 1.0272980501392757,
      "grad_norm": 1.467175521719391,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8804051876068115,
      "num_tokens": 192526000.0,
      "step": 5532
    },
    {
      "epoch": 1.0274837511606314,
      "grad_norm": 1.649062869172201,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.871181070804596,
      "num_tokens": 192556549.0,
      "step": 5533
    },
    {
      "epoch": 1.027669452181987,
      "grad_norm": 1.4641805927990938,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8866367936134338,
      "num_tokens": 192594005.0,
      "step": 5534
    },
    {
      "epoch": 1.0278551532033426,
      "grad_norm": 1.7488497690974447,
      "learning_rate": 1e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8595281839370728,
      "num_tokens": 192621251.0,
      "step": 5535
    },
    {
      "epoch": 1.0280408542246982,
      "grad_norm": 1.5065569156796221,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8676177859306335,
      "num_tokens": 192654660.0,
      "step": 5536
    },
    {
      "epoch": 1.0282265552460539,
      "grad_norm": 1.5268401764814918,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8890833854675293,
      "num_tokens": 192685307.0,
      "step": 5537
    },
    {
      "epoch": 1.0284122562674094,
      "grad_norm": 1.417198876571444,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8806973099708557,
      "num_tokens": 192726345.0,
      "step": 5538
    },
    {
      "epoch": 1.0285979572887651,
      "grad_norm": 1.431529032500988,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8812158703804016,
      "num_tokens": 192762200.0,
      "step": 5539
    },
    {
      "epoch": 1.0287836583101206,
      "grad_norm": 1.6262411528040084,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8804699182510376,
      "num_tokens": 192791993.0,
      "step": 5540
    },
    {
      "epoch": 1.0289693593314764,
      "grad_norm": 1.537411147819433,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8702117204666138,
      "num_tokens": 192824353.0,
      "step": 5541
    },
    {
      "epoch": 1.0291550603528319,
      "grad_norm": 1.409031837551487,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8720259666442871,
      "num_tokens": 192865459.0,
      "step": 5542
    },
    {
      "epoch": 1.0293407613741876,
      "grad_norm": 1.5547491731991572,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8687052726745605,
      "num_tokens": 192901347.0,
      "step": 5543
    },
    {
      "epoch": 1.0295264623955431,
      "grad_norm": 1.3669507573818243,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8898611068725586,
      "num_tokens": 192945234.0,
      "step": 5544
    },
    {
      "epoch": 1.0297121634168989,
      "grad_norm": 1.5396651163444905,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8736783862113953,
      "num_tokens": 192975653.0,
      "step": 5545
    },
    {
      "epoch": 1.0298978644382544,
      "grad_norm": 1.3835370351645784,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8859498500823975,
      "num_tokens": 193014532.0,
      "step": 5546
    },
    {
      "epoch": 1.03008356545961,
      "grad_norm": 1.5553524730566817,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8666963577270508,
      "num_tokens": 193052572.0,
      "step": 5547
    },
    {
      "epoch": 1.0302692664809656,
      "grad_norm": 1.5263266100395334,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8597981929779053,
      "num_tokens": 193089670.0,
      "step": 5548
    },
    {
      "epoch": 1.0304549675023214,
      "grad_norm": 1.3628264987461112,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.89019775390625,
      "num_tokens": 193129205.0,
      "step": 5549
    },
    {
      "epoch": 1.0306406685236769,
      "grad_norm": 1.5659801716576196,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8852193355560303,
      "num_tokens": 193160493.0,
      "step": 5550
    },
    {
      "epoch": 1.0308263695450326,
      "grad_norm": 1.4145457827249854,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8843103647232056,
      "num_tokens": 193197321.0,
      "step": 5551
    },
    {
      "epoch": 1.031012070566388,
      "grad_norm": 1.5469569388862794,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8912739753723145,
      "num_tokens": 193229673.0,
      "step": 5552
    },
    {
      "epoch": 1.0311977715877438,
      "grad_norm": 1.4801436763981952,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8848143815994263,
      "num_tokens": 193262891.0,
      "step": 5553
    },
    {
      "epoch": 1.0313834726090993,
      "grad_norm": 1.5162562697382074,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.87948077917099,
      "num_tokens": 193295962.0,
      "step": 5554
    },
    {
      "epoch": 1.0315691736304549,
      "grad_norm": 1.462527355866356,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8760641813278198,
      "num_tokens": 193331701.0,
      "step": 5555
    },
    {
      "epoch": 1.0317548746518106,
      "grad_norm": 1.3785136506373212,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8890300393104553,
      "num_tokens": 193365630.0,
      "step": 5556
    },
    {
      "epoch": 1.031940575673166,
      "grad_norm": 1.677075042614892,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8785560727119446,
      "num_tokens": 193393504.0,
      "step": 5557
    },
    {
      "epoch": 1.0321262766945218,
      "grad_norm": 1.5326813203365286,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8688274025917053,
      "num_tokens": 193426947.0,
      "step": 5558
    },
    {
      "epoch": 1.0323119777158773,
      "grad_norm": 1.4099012546071201,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8800382614135742,
      "num_tokens": 193465672.0,
      "step": 5559
    },
    {
      "epoch": 1.032497678737233,
      "grad_norm": 1.4176110772265549,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8814326524734497,
      "num_tokens": 193502346.0,
      "step": 5560
    },
    {
      "epoch": 1.0326833797585886,
      "grad_norm": 1.4422917907291146,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8895230293273926,
      "num_tokens": 193537797.0,
      "step": 5561
    },
    {
      "epoch": 1.0328690807799443,
      "grad_norm": 1.5558580405104239,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.87305748462677,
      "num_tokens": 193570930.0,
      "step": 5562
    },
    {
      "epoch": 1.0330547818012998,
      "grad_norm": 1.5060002162559005,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.878576397895813,
      "num_tokens": 193604593.0,
      "step": 5563
    },
    {
      "epoch": 1.0332404828226556,
      "grad_norm": 1.4830843199992947,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8880783319473267,
      "num_tokens": 193638170.0,
      "step": 5564
    },
    {
      "epoch": 1.033426183844011,
      "grad_norm": 1.4501857108594323,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8675368428230286,
      "num_tokens": 193675640.0,
      "step": 5565
    },
    {
      "epoch": 1.0336118848653668,
      "grad_norm": 1.4848813422242872,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.883548378944397,
      "num_tokens": 193711768.0,
      "step": 5566
    },
    {
      "epoch": 1.0337975858867223,
      "grad_norm": 1.5600502434023118,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8840682506561279,
      "num_tokens": 193741703.0,
      "step": 5567
    },
    {
      "epoch": 1.033983286908078,
      "grad_norm": 1.5051925672849493,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8904763460159302,
      "num_tokens": 193773504.0,
      "step": 5568
    },
    {
      "epoch": 1.0341689879294336,
      "grad_norm": 1.3578814416470808,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8759020566940308,
      "num_tokens": 193815598.0,
      "step": 5569
    },
    {
      "epoch": 1.0343546889507893,
      "grad_norm": 1.39255436419926,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8878235220909119,
      "num_tokens": 193852109.0,
      "step": 5570
    },
    {
      "epoch": 1.0345403899721448,
      "grad_norm": 1.4551137576853268,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8764204978942871,
      "num_tokens": 193887005.0,
      "step": 5571
    },
    {
      "epoch": 1.0347260909935005,
      "grad_norm": 1.5536980263016447,
      "learning_rate": 1e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8662801384925842,
      "num_tokens": 193918443.0,
      "step": 5572
    },
    {
      "epoch": 1.034911792014856,
      "grad_norm": 1.4610494474031572,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8809006214141846,
      "num_tokens": 193952430.0,
      "step": 5573
    },
    {
      "epoch": 1.0350974930362118,
      "grad_norm": 1.5457151866226106,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8678745627403259,
      "num_tokens": 193988969.0,
      "step": 5574
    },
    {
      "epoch": 1.0352831940575673,
      "grad_norm": 1.5685583457543273,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8720178604125977,
      "num_tokens": 194017527.0,
      "step": 5575
    },
    {
      "epoch": 1.035468895078923,
      "grad_norm": 1.5184283277246815,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8823269009590149,
      "num_tokens": 194053340.0,
      "step": 5576
    },
    {
      "epoch": 1.0356545961002785,
      "grad_norm": 1.4014480207845859,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8676836490631104,
      "num_tokens": 194096612.0,
      "step": 5577
    },
    {
      "epoch": 1.0358402971216343,
      "grad_norm": 1.4722314359988036,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8837186098098755,
      "num_tokens": 194130260.0,
      "step": 5578
    },
    {
      "epoch": 1.0360259981429898,
      "grad_norm": 1.3828686575966247,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8748992681503296,
      "num_tokens": 194170769.0,
      "step": 5579
    },
    {
      "epoch": 1.0362116991643453,
      "grad_norm": 1.4994898522025477,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8680920600891113,
      "num_tokens": 194206471.0,
      "step": 5580
    },
    {
      "epoch": 1.036397400185701,
      "grad_norm": 1.6843656162368674,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8658267855644226,
      "num_tokens": 194235988.0,
      "step": 5581
    },
    {
      "epoch": 1.0365831012070565,
      "grad_norm": 1.4516359910957524,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.891891360282898,
      "num_tokens": 194270342.0,
      "step": 5582
    },
    {
      "epoch": 1.0367688022284123,
      "grad_norm": 1.4648177343664117,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8819772601127625,
      "num_tokens": 194309982.0,
      "step": 5583
    },
    {
      "epoch": 1.0369545032497678,
      "grad_norm": 1.4319844945170523,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8814374208450317,
      "num_tokens": 194348881.0,
      "step": 5584
    },
    {
      "epoch": 1.0371402042711235,
      "grad_norm": 1.4050906719840095,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8796758055686951,
      "num_tokens": 194389210.0,
      "step": 5585
    },
    {
      "epoch": 1.037325905292479,
      "grad_norm": 1.4324207727591176,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8764809966087341,
      "num_tokens": 194425725.0,
      "step": 5586
    },
    {
      "epoch": 1.0375116063138348,
      "grad_norm": 1.4625179514936966,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8739962577819824,
      "num_tokens": 194461429.0,
      "step": 5587
    },
    {
      "epoch": 1.0376973073351903,
      "grad_norm": 1.3833309566475196,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8860229849815369,
      "num_tokens": 194499490.0,
      "step": 5588
    },
    {
      "epoch": 1.037883008356546,
      "grad_norm": 1.4494713632036338,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8682108521461487,
      "num_tokens": 194539886.0,
      "step": 5589
    },
    {
      "epoch": 1.0380687093779015,
      "grad_norm": 1.510803001261382,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8841009140014648,
      "num_tokens": 194574665.0,
      "step": 5590
    },
    {
      "epoch": 1.0382544103992573,
      "grad_norm": 1.4118717132957834,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8879863619804382,
      "num_tokens": 194609824.0,
      "step": 5591
    },
    {
      "epoch": 1.0384401114206128,
      "grad_norm": 1.458157169470907,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8840477466583252,
      "num_tokens": 194645178.0,
      "step": 5592
    },
    {
      "epoch": 1.0386258124419685,
      "grad_norm": 1.4732310797743087,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8814149498939514,
      "num_tokens": 194679115.0,
      "step": 5593
    },
    {
      "epoch": 1.038811513463324,
      "grad_norm": 1.440968529560919,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8727065324783325,
      "num_tokens": 194717119.0,
      "step": 5594
    },
    {
      "epoch": 1.0389972144846797,
      "grad_norm": 1.5562512785013602,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8867087364196777,
      "num_tokens": 194748381.0,
      "step": 5595
    },
    {
      "epoch": 1.0391829155060353,
      "grad_norm": 1.4073664555421006,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8628485202789307,
      "num_tokens": 194786846.0,
      "step": 5596
    },
    {
      "epoch": 1.039368616527391,
      "grad_norm": 1.5029821296760126,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8747695088386536,
      "num_tokens": 194828496.0,
      "step": 5597
    },
    {
      "epoch": 1.0395543175487465,
      "grad_norm": 1.7031298241582975,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8713371157646179,
      "num_tokens": 194854133.0,
      "step": 5598
    },
    {
      "epoch": 1.0397400185701022,
      "grad_norm": 1.543915804057419,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8976485133171082,
      "num_tokens": 194882758.0,
      "step": 5599
    },
    {
      "epoch": 1.0399257195914577,
      "grad_norm": 1.4460104855354747,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8791341781616211,
      "num_tokens": 194921972.0,
      "step": 5600
    },
    {
      "epoch": 1.0401114206128135,
      "grad_norm": 1.5707145882208595,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8871985673904419,
      "num_tokens": 194952051.0,
      "step": 5601
    },
    {
      "epoch": 1.040297121634169,
      "grad_norm": 1.4604039187475968,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8751387596130371,
      "num_tokens": 194987698.0,
      "step": 5602
    },
    {
      "epoch": 1.0404828226555245,
      "grad_norm": 1.4351632113702888,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8896317481994629,
      "num_tokens": 195019605.0,
      "step": 5603
    },
    {
      "epoch": 1.0406685236768802,
      "grad_norm": 1.5657686408937501,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8677128553390503,
      "num_tokens": 195053183.0,
      "step": 5604
    },
    {
      "epoch": 1.0408542246982357,
      "grad_norm": 1.8018231273057737,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8749974966049194,
      "num_tokens": 195079878.0,
      "step": 5605
    },
    {
      "epoch": 1.0410399257195915,
      "grad_norm": 1.5976894907383536,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8747286200523376,
      "num_tokens": 195113077.0,
      "step": 5606
    },
    {
      "epoch": 1.041225626740947,
      "grad_norm": 1.4258430027899358,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8791562914848328,
      "num_tokens": 195150173.0,
      "step": 5607
    },
    {
      "epoch": 1.0414113277623027,
      "grad_norm": 1.5564425605304595,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8747079968452454,
      "num_tokens": 195179334.0,
      "step": 5608
    },
    {
      "epoch": 1.0415970287836582,
      "grad_norm": 1.5450507232178399,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8797370791435242,
      "num_tokens": 195215680.0,
      "step": 5609
    },
    {
      "epoch": 1.041782729805014,
      "grad_norm": 1.5228332289022997,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8675686120986938,
      "num_tokens": 195252559.0,
      "step": 5610
    },
    {
      "epoch": 1.0419684308263695,
      "grad_norm": 1.5903885706425787,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8749729990959167,
      "num_tokens": 195285785.0,
      "step": 5611
    },
    {
      "epoch": 1.0421541318477252,
      "grad_norm": 1.411011102407428,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8765037059783936,
      "num_tokens": 195324419.0,
      "step": 5612
    },
    {
      "epoch": 1.0423398328690807,
      "grad_norm": 1.6081920323045027,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8633895516395569,
      "num_tokens": 195357952.0,
      "step": 5613
    },
    {
      "epoch": 1.0425255338904365,
      "grad_norm": 1.5006641328516568,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8660749197006226,
      "num_tokens": 195397793.0,
      "step": 5614
    },
    {
      "epoch": 1.042711234911792,
      "grad_norm": 1.5492956390139976,
      "learning_rate": 1e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.8567671179771423,
      "num_tokens": 195432709.0,
      "step": 5615
    },
    {
      "epoch": 1.0428969359331477,
      "grad_norm": 1.6139947403126726,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8743270039558411,
      "num_tokens": 195462477.0,
      "step": 5616
    },
    {
      "epoch": 1.0430826369545032,
      "grad_norm": 1.4599888267222572,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8753238916397095,
      "num_tokens": 195498837.0,
      "step": 5617
    },
    {
      "epoch": 1.043268337975859,
      "grad_norm": 1.5135730454776317,
      "learning_rate": 1e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.866223156452179,
      "num_tokens": 195536248.0,
      "step": 5618
    },
    {
      "epoch": 1.0434540389972145,
      "grad_norm": 1.5833582120323952,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8732408285140991,
      "num_tokens": 195569906.0,
      "step": 5619
    },
    {
      "epoch": 1.0436397400185702,
      "grad_norm": 1.563997799443293,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.866473376750946,
      "num_tokens": 195602259.0,
      "step": 5620
    },
    {
      "epoch": 1.0438254410399257,
      "grad_norm": 1.3950861057043749,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8835517168045044,
      "num_tokens": 195641449.0,
      "step": 5621
    },
    {
      "epoch": 1.0440111420612814,
      "grad_norm": 1.4283666149683654,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8802543878555298,
      "num_tokens": 195677348.0,
      "step": 5622
    },
    {
      "epoch": 1.044196843082637,
      "grad_norm": 1.4510569506934508,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8844830393791199,
      "num_tokens": 195714461.0,
      "step": 5623
    },
    {
      "epoch": 1.0443825441039927,
      "grad_norm": 1.4153890140025283,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8791635036468506,
      "num_tokens": 195753745.0,
      "step": 5624
    },
    {
      "epoch": 1.0445682451253482,
      "grad_norm": 1.5121414344868824,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8659759759902954,
      "num_tokens": 195790155.0,
      "step": 5625
    },
    {
      "epoch": 1.0447539461467037,
      "grad_norm": 1.5495763882040603,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8733752965927124,
      "num_tokens": 195825387.0,
      "step": 5626
    },
    {
      "epoch": 1.0449396471680594,
      "grad_norm": 1.530080306799313,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8771019577980042,
      "num_tokens": 195857973.0,
      "step": 5627
    },
    {
      "epoch": 1.045125348189415,
      "grad_norm": 1.6491156561130036,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8835507035255432,
      "num_tokens": 195886179.0,
      "step": 5628
    },
    {
      "epoch": 1.0453110492107707,
      "grad_norm": 1.3514766367813076,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8758854866027832,
      "num_tokens": 195927215.0,
      "step": 5629
    },
    {
      "epoch": 1.0454967502321262,
      "grad_norm": 1.519985495109226,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8808087110519409,
      "num_tokens": 195961108.0,
      "step": 5630
    },
    {
      "epoch": 1.045682451253482,
      "grad_norm": 1.4159314377725805,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8807505965232849,
      "num_tokens": 196000263.0,
      "step": 5631
    },
    {
      "epoch": 1.0458681522748374,
      "grad_norm": 1.5370390804414191,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8710846900939941,
      "num_tokens": 196033478.0,
      "step": 5632
    },
    {
      "epoch": 1.0460538532961932,
      "grad_norm": 1.7471221490124453,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8797616362571716,
      "num_tokens": 196058581.0,
      "step": 5633
    },
    {
      "epoch": 1.0462395543175487,
      "grad_norm": 1.486563071078982,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8930273056030273,
      "num_tokens": 196091518.0,
      "step": 5634
    },
    {
      "epoch": 1.0464252553389044,
      "grad_norm": 1.562416867140614,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8930248618125916,
      "num_tokens": 196120734.0,
      "step": 5635
    },
    {
      "epoch": 1.04661095636026,
      "grad_norm": 1.555817092894301,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8714455366134644,
      "num_tokens": 196155448.0,
      "step": 5636
    },
    {
      "epoch": 1.0467966573816156,
      "grad_norm": 1.3511329005781163,
      "learning_rate": 1e-06,
      "loss": 0.2901,
      "mean_token_accuracy": 0.8969951868057251,
      "num_tokens": 196192265.0,
      "step": 5637
    },
    {
      "epoch": 1.0469823584029712,
      "grad_norm": 1.5063880745642604,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8761892914772034,
      "num_tokens": 196228205.0,
      "step": 5638
    },
    {
      "epoch": 1.047168059424327,
      "grad_norm": 1.4686809592964827,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8837548494338989,
      "num_tokens": 196260690.0,
      "step": 5639
    },
    {
      "epoch": 1.0473537604456824,
      "grad_norm": 1.4705821381386144,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.880580484867096,
      "num_tokens": 196298572.0,
      "step": 5640
    },
    {
      "epoch": 1.0475394614670381,
      "grad_norm": 1.5363314373042358,
      "learning_rate": 1e-06,
      "loss": 0.2862,
      "mean_token_accuracy": 0.898507833480835,
      "num_tokens": 196333874.0,
      "step": 5641
    },
    {
      "epoch": 1.0477251624883936,
      "grad_norm": 1.536213787794197,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8818580508232117,
      "num_tokens": 196369438.0,
      "step": 5642
    },
    {
      "epoch": 1.0479108635097494,
      "grad_norm": 1.4923234176552505,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8823080062866211,
      "num_tokens": 196400735.0,
      "step": 5643
    },
    {
      "epoch": 1.048096564531105,
      "grad_norm": 1.527905019829511,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8743024468421936,
      "num_tokens": 196435235.0,
      "step": 5644
    },
    {
      "epoch": 1.0482822655524606,
      "grad_norm": 1.750491240564293,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8723119497299194,
      "num_tokens": 196461520.0,
      "step": 5645
    },
    {
      "epoch": 1.0484679665738161,
      "grad_norm": 1.4658159594695297,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8698840141296387,
      "num_tokens": 196500477.0,
      "step": 5646
    },
    {
      "epoch": 1.0486536675951719,
      "grad_norm": 1.5781608099265276,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8772011995315552,
      "num_tokens": 196537932.0,
      "step": 5647
    },
    {
      "epoch": 1.0488393686165274,
      "grad_norm": 1.6394384164366753,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8737643361091614,
      "num_tokens": 196569510.0,
      "step": 5648
    },
    {
      "epoch": 1.0490250696378831,
      "grad_norm": 1.455706483174583,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8811852335929871,
      "num_tokens": 196606922.0,
      "step": 5649
    },
    {
      "epoch": 1.0492107706592386,
      "grad_norm": 1.5478919736855128,
      "learning_rate": 1e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8516944050788879,
      "num_tokens": 196645872.0,
      "step": 5650
    },
    {
      "epoch": 1.0493964716805944,
      "grad_norm": 1.4083534071401675,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8774805068969727,
      "num_tokens": 196683491.0,
      "step": 5651
    },
    {
      "epoch": 1.0495821727019499,
      "grad_norm": 1.6775201414712266,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.86320561170578,
      "num_tokens": 196715035.0,
      "step": 5652
    },
    {
      "epoch": 1.0497678737233054,
      "grad_norm": 1.5452906725277453,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8745983839035034,
      "num_tokens": 196753235.0,
      "step": 5653
    },
    {
      "epoch": 1.0499535747446611,
      "grad_norm": 1.4716547528925836,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8859018683433533,
      "num_tokens": 196788095.0,
      "step": 5654
    },
    {
      "epoch": 1.0501392757660166,
      "grad_norm": 1.6130095807719607,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8767677545547485,
      "num_tokens": 196823815.0,
      "step": 5655
    },
    {
      "epoch": 1.0503249767873724,
      "grad_norm": 1.4952477304744756,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8693767189979553,
      "num_tokens": 196855781.0,
      "step": 5656
    },
    {
      "epoch": 1.0505106778087279,
      "grad_norm": 1.4878607745496057,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8646106123924255,
      "num_tokens": 196892123.0,
      "step": 5657
    },
    {
      "epoch": 1.0506963788300836,
      "grad_norm": 1.4211710506084005,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8734867572784424,
      "num_tokens": 196927623.0,
      "step": 5658
    },
    {
      "epoch": 1.050882079851439,
      "grad_norm": 1.5711954221397353,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8592681288719177,
      "num_tokens": 196968027.0,
      "step": 5659
    },
    {
      "epoch": 1.0510677808727948,
      "grad_norm": 1.5953999380075978,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8744691610336304,
      "num_tokens": 197000415.0,
      "step": 5660
    },
    {
      "epoch": 1.0512534818941504,
      "grad_norm": 1.4178879870979693,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.889972448348999,
      "num_tokens": 197037965.0,
      "step": 5661
    },
    {
      "epoch": 1.051439182915506,
      "grad_norm": 1.5668494284360923,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8686234951019287,
      "num_tokens": 197072055.0,
      "step": 5662
    },
    {
      "epoch": 1.0516248839368616,
      "grad_norm": 1.5413970477024121,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8814666867256165,
      "num_tokens": 197105804.0,
      "step": 5663
    },
    {
      "epoch": 1.0518105849582173,
      "grad_norm": 1.5089509530827232,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8818032741546631,
      "num_tokens": 197138920.0,
      "step": 5664
    },
    {
      "epoch": 1.0519962859795728,
      "grad_norm": 1.4061907246599517,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8766844272613525,
      "num_tokens": 197175556.0,
      "step": 5665
    },
    {
      "epoch": 1.0521819870009286,
      "grad_norm": 1.42862940266609,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8790798783302307,
      "num_tokens": 197215480.0,
      "step": 5666
    },
    {
      "epoch": 1.052367688022284,
      "grad_norm": 1.4505923964978387,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.867276132106781,
      "num_tokens": 197251875.0,
      "step": 5667
    },
    {
      "epoch": 1.0525533890436398,
      "grad_norm": 1.5482269158294255,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8727157115936279,
      "num_tokens": 197286357.0,
      "step": 5668
    },
    {
      "epoch": 1.0527390900649953,
      "grad_norm": 1.4582517565185373,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8615565299987793,
      "num_tokens": 197330042.0,
      "step": 5669
    },
    {
      "epoch": 1.052924791086351,
      "grad_norm": 1.518385234199787,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8776870369911194,
      "num_tokens": 197362899.0,
      "step": 5670
    },
    {
      "epoch": 1.0531104921077066,
      "grad_norm": 1.535665915714376,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8843300342559814,
      "num_tokens": 197393559.0,
      "step": 5671
    },
    {
      "epoch": 1.0532961931290623,
      "grad_norm": 1.413959891080473,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8811776638031006,
      "num_tokens": 197431208.0,
      "step": 5672
    },
    {
      "epoch": 1.0534818941504178,
      "grad_norm": 1.3192938205964087,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8705853819847107,
      "num_tokens": 197475181.0,
      "step": 5673
    },
    {
      "epoch": 1.0536675951717736,
      "grad_norm": 1.6721905664698056,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8647192716598511,
      "num_tokens": 197505362.0,
      "step": 5674
    },
    {
      "epoch": 1.053853296193129,
      "grad_norm": 1.3761516539157057,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8810189366340637,
      "num_tokens": 197543822.0,
      "step": 5675
    },
    {
      "epoch": 1.0540389972144846,
      "grad_norm": 1.5049474030640404,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8721086978912354,
      "num_tokens": 197581019.0,
      "step": 5676
    },
    {
      "epoch": 1.0542246982358403,
      "grad_norm": 1.3993249033635584,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8759903907775879,
      "num_tokens": 197622761.0,
      "step": 5677
    },
    {
      "epoch": 1.0544103992571958,
      "grad_norm": 1.543251258993103,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8651330471038818,
      "num_tokens": 197658696.0,
      "step": 5678
    },
    {
      "epoch": 1.0545961002785516,
      "grad_norm": 1.3068024956551185,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8847019672393799,
      "num_tokens": 197698772.0,
      "step": 5679
    },
    {
      "epoch": 1.054781801299907,
      "grad_norm": 1.527440500099589,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8734220266342163,
      "num_tokens": 197733641.0,
      "step": 5680
    },
    {
      "epoch": 1.0549675023212628,
      "grad_norm": 1.3786081789095825,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8779208064079285,
      "num_tokens": 197771768.0,
      "step": 5681
    },
    {
      "epoch": 1.0551532033426183,
      "grad_norm": 1.3906918600274454,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8865128755569458,
      "num_tokens": 197807309.0,
      "step": 5682
    },
    {
      "epoch": 1.055338904363974,
      "grad_norm": 1.472995349482008,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8859080672264099,
      "num_tokens": 197846199.0,
      "step": 5683
    },
    {
      "epoch": 1.0555246053853296,
      "grad_norm": 1.4911765017880296,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8870731592178345,
      "num_tokens": 197880377.0,
      "step": 5684
    },
    {
      "epoch": 1.0557103064066853,
      "grad_norm": 1.4828537328092222,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8883353471755981,
      "num_tokens": 197914965.0,
      "step": 5685
    },
    {
      "epoch": 1.0558960074280408,
      "grad_norm": 1.5742325792409104,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.873867392539978,
      "num_tokens": 197944083.0,
      "step": 5686
    },
    {
      "epoch": 1.0560817084493965,
      "grad_norm": 1.4981113696560917,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8873875737190247,
      "num_tokens": 197977181.0,
      "step": 5687
    },
    {
      "epoch": 1.056267409470752,
      "grad_norm": 1.50189353899142,
      "learning_rate": 1e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8617530465126038,
      "num_tokens": 198014616.0,
      "step": 5688
    },
    {
      "epoch": 1.0564531104921078,
      "grad_norm": 1.540311888435286,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8804789781570435,
      "num_tokens": 198046743.0,
      "step": 5689
    },
    {
      "epoch": 1.0566388115134633,
      "grad_norm": 1.509065528487574,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.861058235168457,
      "num_tokens": 198088843.0,
      "step": 5690
    },
    {
      "epoch": 1.056824512534819,
      "grad_norm": 1.5243830318129155,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8739793300628662,
      "num_tokens": 198122913.0,
      "step": 5691
    },
    {
      "epoch": 1.0570102135561745,
      "grad_norm": 1.4026128769583885,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8879392147064209,
      "num_tokens": 198159089.0,
      "step": 5692
    },
    {
      "epoch": 1.0571959145775303,
      "grad_norm": 1.6299342235204397,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8729816675186157,
      "num_tokens": 198191767.0,
      "step": 5693
    },
    {
      "epoch": 1.0573816155988858,
      "grad_norm": 1.593481351333642,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8703059554100037,
      "num_tokens": 198221781.0,
      "step": 5694
    },
    {
      "epoch": 1.0575673166202415,
      "grad_norm": 1.5328631699159083,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8926223516464233,
      "num_tokens": 198250731.0,
      "step": 5695
    },
    {
      "epoch": 1.057753017641597,
      "grad_norm": 1.658137416005786,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8837382793426514,
      "num_tokens": 198276341.0,
      "step": 5696
    },
    {
      "epoch": 1.0579387186629527,
      "grad_norm": 1.523523018834525,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8720628619194031,
      "num_tokens": 198308376.0,
      "step": 5697
    },
    {
      "epoch": 1.0581244196843083,
      "grad_norm": 1.4848434558967394,
      "learning_rate": 1e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8634272813796997,
      "num_tokens": 198346636.0,
      "step": 5698
    },
    {
      "epoch": 1.0583101207056638,
      "grad_norm": 1.4088428392885994,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8896214962005615,
      "num_tokens": 198381431.0,
      "step": 5699
    },
    {
      "epoch": 1.0584958217270195,
      "grad_norm": 1.4358705207237061,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8778903484344482,
      "num_tokens": 198418984.0,
      "step": 5700
    },
    {
      "epoch": 1.058681522748375,
      "grad_norm": 1.4856530288340937,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8825306296348572,
      "num_tokens": 198451715.0,
      "step": 5701
    },
    {
      "epoch": 1.0588672237697307,
      "grad_norm": 1.7291213523109354,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8757084608078003,
      "num_tokens": 198478007.0,
      "step": 5702
    },
    {
      "epoch": 1.0590529247910863,
      "grad_norm": 1.5978475709376476,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8882818222045898,
      "num_tokens": 198508526.0,
      "step": 5703
    },
    {
      "epoch": 1.059238625812442,
      "grad_norm": 1.4937404732355744,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8653501868247986,
      "num_tokens": 198547423.0,
      "step": 5704
    },
    {
      "epoch": 1.0594243268337975,
      "grad_norm": 1.5534805568462025,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8749818801879883,
      "num_tokens": 198580860.0,
      "step": 5705
    },
    {
      "epoch": 1.0596100278551532,
      "grad_norm": 1.5675823035438594,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8813587427139282,
      "num_tokens": 198610757.0,
      "step": 5706
    },
    {
      "epoch": 1.0597957288765087,
      "grad_norm": 1.678002412715747,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8720455765724182,
      "num_tokens": 198642929.0,
      "step": 5707
    },
    {
      "epoch": 1.0599814298978645,
      "grad_norm": 1.5584794038036507,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8804251551628113,
      "num_tokens": 198674849.0,
      "step": 5708
    },
    {
      "epoch": 1.06016713091922,
      "grad_norm": 1.3919276304442545,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8785672187805176,
      "num_tokens": 198714394.0,
      "step": 5709
    },
    {
      "epoch": 1.0603528319405757,
      "grad_norm": 1.499192546322716,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8739681839942932,
      "num_tokens": 198749866.0,
      "step": 5710
    },
    {
      "epoch": 1.0605385329619312,
      "grad_norm": 1.6921306757207575,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8754757046699524,
      "num_tokens": 198775662.0,
      "step": 5711
    },
    {
      "epoch": 1.060724233983287,
      "grad_norm": 1.3451643557547244,
      "learning_rate": 1e-06,
      "loss": 0.2809,
      "mean_token_accuracy": 0.8987432718276978,
      "num_tokens": 198810257.0,
      "step": 5712
    },
    {
      "epoch": 1.0609099350046425,
      "grad_norm": 1.4837054021001443,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8769341707229614,
      "num_tokens": 198848434.0,
      "step": 5713
    },
    {
      "epoch": 1.0610956360259982,
      "grad_norm": 1.4383497570090855,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8683171272277832,
      "num_tokens": 198884985.0,
      "step": 5714
    },
    {
      "epoch": 1.0612813370473537,
      "grad_norm": 1.5824276099782792,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8742704391479492,
      "num_tokens": 198916009.0,
      "step": 5715
    },
    {
      "epoch": 1.0614670380687095,
      "grad_norm": 1.5104100159699088,
      "learning_rate": 1e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8604154586791992,
      "num_tokens": 198953697.0,
      "step": 5716
    },
    {
      "epoch": 1.061652739090065,
      "grad_norm": 1.5659328246155517,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8671623468399048,
      "num_tokens": 198987189.0,
      "step": 5717
    },
    {
      "epoch": 1.0618384401114207,
      "grad_norm": 1.5922263603250402,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8772349953651428,
      "num_tokens": 199019099.0,
      "step": 5718
    },
    {
      "epoch": 1.0620241411327762,
      "grad_norm": 1.5774336929814214,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8688030242919922,
      "num_tokens": 199051784.0,
      "step": 5719
    },
    {
      "epoch": 1.062209842154132,
      "grad_norm": 1.5486589245404894,
      "learning_rate": 1e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8647150993347168,
      "num_tokens": 199083044.0,
      "step": 5720
    },
    {
      "epoch": 1.0623955431754875,
      "grad_norm": 1.5064865726769243,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8792797923088074,
      "num_tokens": 199118536.0,
      "step": 5721
    },
    {
      "epoch": 1.062581244196843,
      "grad_norm": 1.664935439612868,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.86881023645401,
      "num_tokens": 199148703.0,
      "step": 5722
    },
    {
      "epoch": 1.0627669452181987,
      "grad_norm": 1.4136767082382458,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8696718215942383,
      "num_tokens": 199188665.0,
      "step": 5723
    },
    {
      "epoch": 1.0629526462395544,
      "grad_norm": 1.5245435217831835,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.874750018119812,
      "num_tokens": 199220860.0,
      "step": 5724
    },
    {
      "epoch": 1.06313834726091,
      "grad_norm": 1.553474865562472,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.880476713180542,
      "num_tokens": 199251852.0,
      "step": 5725
    },
    {
      "epoch": 1.0633240482822655,
      "grad_norm": 1.580235281262064,
      "learning_rate": 1e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8705753684043884,
      "num_tokens": 199283106.0,
      "step": 5726
    },
    {
      "epoch": 1.0635097493036212,
      "grad_norm": 1.5978495880283445,
      "learning_rate": 1e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8693846464157104,
      "num_tokens": 199316361.0,
      "step": 5727
    },
    {
      "epoch": 1.0636954503249767,
      "grad_norm": 1.5474143678810561,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8726017475128174,
      "num_tokens": 199350237.0,
      "step": 5728
    },
    {
      "epoch": 1.0638811513463324,
      "grad_norm": 1.5096401413162444,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8764841556549072,
      "num_tokens": 199385903.0,
      "step": 5729
    },
    {
      "epoch": 1.064066852367688,
      "grad_norm": 1.393139117637904,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8685401082038879,
      "num_tokens": 199429449.0,
      "step": 5730
    },
    {
      "epoch": 1.0642525533890437,
      "grad_norm": 1.3935888096296596,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.8958108425140381,
      "num_tokens": 199469811.0,
      "step": 5731
    },
    {
      "epoch": 1.0644382544103992,
      "grad_norm": 1.392033170225477,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8817603588104248,
      "num_tokens": 199507976.0,
      "step": 5732
    },
    {
      "epoch": 1.064623955431755,
      "grad_norm": 1.5145449922514012,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8750653862953186,
      "num_tokens": 199544580.0,
      "step": 5733
    },
    {
      "epoch": 1.0648096564531104,
      "grad_norm": 1.549967488257986,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8775404691696167,
      "num_tokens": 199574784.0,
      "step": 5734
    },
    {
      "epoch": 1.0649953574744662,
      "grad_norm": 1.4319164801250788,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.877802848815918,
      "num_tokens": 199611370.0,
      "step": 5735
    },
    {
      "epoch": 1.0651810584958217,
      "grad_norm": 1.5594221688709804,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8888629674911499,
      "num_tokens": 199644045.0,
      "step": 5736
    },
    {
      "epoch": 1.0653667595171774,
      "grad_norm": 1.4211389273712134,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8891140222549438,
      "num_tokens": 199680897.0,
      "step": 5737
    },
    {
      "epoch": 1.065552460538533,
      "grad_norm": 1.4792179475400087,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8743820786476135,
      "num_tokens": 199717831.0,
      "step": 5738
    },
    {
      "epoch": 1.0657381615598887,
      "grad_norm": 1.5893822688084995,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8664900064468384,
      "num_tokens": 199750771.0,
      "step": 5739
    },
    {
      "epoch": 1.0659238625812442,
      "grad_norm": 1.5247131620739556,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8756415843963623,
      "num_tokens": 199786850.0,
      "step": 5740
    },
    {
      "epoch": 1.0661095636026,
      "grad_norm": 1.4836578335143704,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8728967308998108,
      "num_tokens": 199823460.0,
      "step": 5741
    },
    {
      "epoch": 1.0662952646239554,
      "grad_norm": 1.4174795834517495,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.881159782409668,
      "num_tokens": 199867417.0,
      "step": 5742
    },
    {
      "epoch": 1.0664809656453111,
      "grad_norm": 1.4815262694175917,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.877688467502594,
      "num_tokens": 199902148.0,
      "step": 5743
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 1.373476383973294,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8736631870269775,
      "num_tokens": 199946028.0,
      "step": 5744
    },
    {
      "epoch": 1.0668523676880224,
      "grad_norm": 1.4670596912758191,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8837748765945435,
      "num_tokens": 199984506.0,
      "step": 5745
    },
    {
      "epoch": 1.067038068709378,
      "grad_norm": 1.6065191188309569,
      "learning_rate": 1e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8580009937286377,
      "num_tokens": 200021680.0,
      "step": 5746
    },
    {
      "epoch": 1.0672237697307336,
      "grad_norm": 1.498042734112233,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8688144087791443,
      "num_tokens": 200056555.0,
      "step": 5747
    },
    {
      "epoch": 1.0674094707520891,
      "grad_norm": 1.6878488817855668,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8664239048957825,
      "num_tokens": 200085655.0,
      "step": 5748
    },
    {
      "epoch": 1.0675951717734447,
      "grad_norm": 1.4290152462600514,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8825894594192505,
      "num_tokens": 200118335.0,
      "step": 5749
    },
    {
      "epoch": 1.0677808727948004,
      "grad_norm": 1.4607171994319053,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8813523650169373,
      "num_tokens": 200152528.0,
      "step": 5750
    },
    {
      "epoch": 1.067966573816156,
      "grad_norm": 1.3448131822174931,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8740427494049072,
      "num_tokens": 200194651.0,
      "step": 5751
    },
    {
      "epoch": 1.0681522748375116,
      "grad_norm": 1.453562382505303,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8861889243125916,
      "num_tokens": 200228843.0,
      "step": 5752
    },
    {
      "epoch": 1.0683379758588671,
      "grad_norm": 1.4454657361609977,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8832190632820129,
      "num_tokens": 200262757.0,
      "step": 5753
    },
    {
      "epoch": 1.0685236768802229,
      "grad_norm": 1.4064270998288253,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.883236289024353,
      "num_tokens": 200300400.0,
      "step": 5754
    },
    {
      "epoch": 1.0687093779015784,
      "grad_norm": 1.586147651442737,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8689430952072144,
      "num_tokens": 200336066.0,
      "step": 5755
    },
    {
      "epoch": 1.0688950789229341,
      "grad_norm": 1.5150587107485067,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8710464239120483,
      "num_tokens": 200372883.0,
      "step": 5756
    },
    {
      "epoch": 1.0690807799442896,
      "grad_norm": 1.4108853690989367,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8921976685523987,
      "num_tokens": 200409400.0,
      "step": 5757
    },
    {
      "epoch": 1.0692664809656454,
      "grad_norm": 1.4249313511674566,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8861362338066101,
      "num_tokens": 200447952.0,
      "step": 5758
    },
    {
      "epoch": 1.0694521819870009,
      "grad_norm": 1.385215479936225,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8865562677383423,
      "num_tokens": 200484557.0,
      "step": 5759
    },
    {
      "epoch": 1.0696378830083566,
      "grad_norm": 1.4882419749226465,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8693552017211914,
      "num_tokens": 200523882.0,
      "step": 5760
    },
    {
      "epoch": 1.0698235840297121,
      "grad_norm": 1.4501796572777534,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8793411254882812,
      "num_tokens": 200562670.0,
      "step": 5761
    },
    {
      "epoch": 1.0700092850510678,
      "grad_norm": 1.5677251975282025,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8760400414466858,
      "num_tokens": 200594634.0,
      "step": 5762
    },
    {
      "epoch": 1.0701949860724234,
      "grad_norm": 1.3871095122272015,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8807491064071655,
      "num_tokens": 200631905.0,
      "step": 5763
    },
    {
      "epoch": 1.070380687093779,
      "grad_norm": 1.5066116540114485,
      "learning_rate": 1e-06,
      "loss": 0.3145,
      "mean_token_accuracy": 0.8890306949615479,
      "num_tokens": 200664975.0,
      "step": 5764
    },
    {
      "epoch": 1.0705663881151346,
      "grad_norm": 1.5006332930900608,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8831413388252258,
      "num_tokens": 200698711.0,
      "step": 5765
    },
    {
      "epoch": 1.0707520891364903,
      "grad_norm": 1.5044895172565351,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8749099969863892,
      "num_tokens": 200732235.0,
      "step": 5766
    },
    {
      "epoch": 1.0709377901578458,
      "grad_norm": 1.3844911604220467,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8875724077224731,
      "num_tokens": 200768538.0,
      "step": 5767
    },
    {
      "epoch": 1.0711234911792016,
      "grad_norm": 1.4670437376415728,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8743637800216675,
      "num_tokens": 200801263.0,
      "step": 5768
    },
    {
      "epoch": 1.071309192200557,
      "grad_norm": 1.7590713561421212,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.882591962814331,
      "num_tokens": 200826957.0,
      "step": 5769
    },
    {
      "epoch": 1.0714948932219128,
      "grad_norm": 1.551179918141849,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.8740721940994263,
      "num_tokens": 200859671.0,
      "step": 5770
    },
    {
      "epoch": 1.0716805942432683,
      "grad_norm": 1.5035983498883168,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8845033645629883,
      "num_tokens": 200891845.0,
      "step": 5771
    },
    {
      "epoch": 1.0718662952646238,
      "grad_norm": 1.5944630368123915,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8820585012435913,
      "num_tokens": 200920519.0,
      "step": 5772
    },
    {
      "epoch": 1.0720519962859796,
      "grad_norm": 1.6135017657554462,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8730822205543518,
      "num_tokens": 200951217.0,
      "step": 5773
    },
    {
      "epoch": 1.072237697307335,
      "grad_norm": 1.4861162554597431,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.869632363319397,
      "num_tokens": 200987854.0,
      "step": 5774
    },
    {
      "epoch": 1.0724233983286908,
      "grad_norm": 1.370844845653352,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8845227360725403,
      "num_tokens": 201028627.0,
      "step": 5775
    },
    {
      "epoch": 1.0726090993500463,
      "grad_norm": 1.7480879057206677,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8638752698898315,
      "num_tokens": 201057707.0,
      "step": 5776
    },
    {
      "epoch": 1.072794800371402,
      "grad_norm": 1.5135609567367814,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8867998123168945,
      "num_tokens": 201089638.0,
      "step": 5777
    },
    {
      "epoch": 1.0729805013927576,
      "grad_norm": 1.5845795820835689,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8765646815299988,
      "num_tokens": 201120030.0,
      "step": 5778
    },
    {
      "epoch": 1.0731662024141133,
      "grad_norm": 1.818796024617807,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8721103072166443,
      "num_tokens": 201143962.0,
      "step": 5779
    },
    {
      "epoch": 1.0733519034354688,
      "grad_norm": 1.5496708246055533,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.860895037651062,
      "num_tokens": 201178327.0,
      "step": 5780
    },
    {
      "epoch": 1.0735376044568246,
      "grad_norm": 1.6422596596690253,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8779138922691345,
      "num_tokens": 201207140.0,
      "step": 5781
    },
    {
      "epoch": 1.07372330547818,
      "grad_norm": 1.6928273101946614,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8706270456314087,
      "num_tokens": 201233711.0,
      "step": 5782
    },
    {
      "epoch": 1.0739090064995358,
      "grad_norm": 1.4863643395633517,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8821451663970947,
      "num_tokens": 201272251.0,
      "step": 5783
    },
    {
      "epoch": 1.0740947075208913,
      "grad_norm": 1.5057338446572046,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.880179762840271,
      "num_tokens": 201304116.0,
      "step": 5784
    },
    {
      "epoch": 1.074280408542247,
      "grad_norm": 1.5348298791972628,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8706944584846497,
      "num_tokens": 201337935.0,
      "step": 5785
    },
    {
      "epoch": 1.0744661095636026,
      "grad_norm": 1.605586632481421,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8855672478675842,
      "num_tokens": 201370095.0,
      "step": 5786
    },
    {
      "epoch": 1.0746518105849583,
      "grad_norm": 1.7243877291972554,
      "learning_rate": 1e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8677758574485779,
      "num_tokens": 201397933.0,
      "step": 5787
    },
    {
      "epoch": 1.0748375116063138,
      "grad_norm": 1.6047194953525397,
      "learning_rate": 1e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8594557046890259,
      "num_tokens": 201428248.0,
      "step": 5788
    },
    {
      "epoch": 1.0750232126276695,
      "grad_norm": 1.515675336298002,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8872759342193604,
      "num_tokens": 201459234.0,
      "step": 5789
    },
    {
      "epoch": 1.075208913649025,
      "grad_norm": 1.6401262429031094,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.863857626914978,
      "num_tokens": 201492713.0,
      "step": 5790
    },
    {
      "epoch": 1.0753946146703808,
      "grad_norm": 1.4950618068182964,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8827641010284424,
      "num_tokens": 201523288.0,
      "step": 5791
    },
    {
      "epoch": 1.0755803156917363,
      "grad_norm": 1.4897157583385292,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.89028000831604,
      "num_tokens": 201554867.0,
      "step": 5792
    },
    {
      "epoch": 1.075766016713092,
      "grad_norm": 1.516718568072758,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8661177158355713,
      "num_tokens": 201596136.0,
      "step": 5793
    },
    {
      "epoch": 1.0759517177344475,
      "grad_norm": 1.3471238924383262,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8907631635665894,
      "num_tokens": 201634286.0,
      "step": 5794
    },
    {
      "epoch": 1.076137418755803,
      "grad_norm": 1.535515935013563,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8811217546463013,
      "num_tokens": 201669556.0,
      "step": 5795
    },
    {
      "epoch": 1.0763231197771588,
      "grad_norm": 1.623236085136414,
      "learning_rate": 1e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8609024286270142,
      "num_tokens": 201702284.0,
      "step": 5796
    },
    {
      "epoch": 1.0765088207985145,
      "grad_norm": 1.3804368878808344,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8768707513809204,
      "num_tokens": 201743282.0,
      "step": 5797
    },
    {
      "epoch": 1.07669452181987,
      "grad_norm": 1.3883496550904972,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8801267147064209,
      "num_tokens": 201783748.0,
      "step": 5798
    },
    {
      "epoch": 1.0768802228412255,
      "grad_norm": 1.6228090233814492,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.8913609385490417,
      "num_tokens": 201812114.0,
      "step": 5799
    },
    {
      "epoch": 1.0770659238625813,
      "grad_norm": 1.6577671114693464,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8722820281982422,
      "num_tokens": 201838759.0,
      "step": 5800
    },
    {
      "epoch": 1.0772516248839368,
      "grad_norm": 1.5904459580349561,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8699444532394409,
      "num_tokens": 201873384.0,
      "step": 5801
    },
    {
      "epoch": 1.0774373259052925,
      "grad_norm": 1.655342320222992,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8618113994598389,
      "num_tokens": 201903695.0,
      "step": 5802
    },
    {
      "epoch": 1.077623026926648,
      "grad_norm": 1.4693751640897226,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8732123970985413,
      "num_tokens": 201942306.0,
      "step": 5803
    },
    {
      "epoch": 1.0778087279480038,
      "grad_norm": 1.5463187062564094,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8704664707183838,
      "num_tokens": 201976476.0,
      "step": 5804
    },
    {
      "epoch": 1.0779944289693593,
      "grad_norm": 1.4352953973325193,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8951737880706787,
      "num_tokens": 202012462.0,
      "step": 5805
    },
    {
      "epoch": 1.078180129990715,
      "grad_norm": 1.4145445092863054,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8837010860443115,
      "num_tokens": 202047263.0,
      "step": 5806
    },
    {
      "epoch": 1.0783658310120705,
      "grad_norm": 1.463708198076408,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8756380677223206,
      "num_tokens": 202088910.0,
      "step": 5807
    },
    {
      "epoch": 1.0785515320334262,
      "grad_norm": 1.575318963887088,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8899489641189575,
      "num_tokens": 202121008.0,
      "step": 5808
    },
    {
      "epoch": 1.0787372330547818,
      "grad_norm": 1.5503577790089615,
      "learning_rate": 1e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8645731210708618,
      "num_tokens": 202158213.0,
      "step": 5809
    },
    {
      "epoch": 1.0789229340761375,
      "grad_norm": 1.4736763930988537,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.879805862903595,
      "num_tokens": 202191848.0,
      "step": 5810
    },
    {
      "epoch": 1.079108635097493,
      "grad_norm": 1.5759595643048214,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8721928596496582,
      "num_tokens": 202227205.0,
      "step": 5811
    },
    {
      "epoch": 1.0792943361188487,
      "grad_norm": 1.6134508860773553,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8747335076332092,
      "num_tokens": 202257433.0,
      "step": 5812
    },
    {
      "epoch": 1.0794800371402042,
      "grad_norm": 1.6024428422975552,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8801259398460388,
      "num_tokens": 202294840.0,
      "step": 5813
    },
    {
      "epoch": 1.07966573816156,
      "grad_norm": 1.6074657552308547,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8799679279327393,
      "num_tokens": 202323719.0,
      "step": 5814
    },
    {
      "epoch": 1.0798514391829155,
      "grad_norm": 1.7183374281085262,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8736329078674316,
      "num_tokens": 202352482.0,
      "step": 5815
    },
    {
      "epoch": 1.0800371402042712,
      "grad_norm": 1.5577054816365217,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8694249391555786,
      "num_tokens": 202386808.0,
      "step": 5816
    },
    {
      "epoch": 1.0802228412256267,
      "grad_norm": 1.5339360758037082,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.87509685754776,
      "num_tokens": 202423455.0,
      "step": 5817
    },
    {
      "epoch": 1.0804085422469825,
      "grad_norm": 1.4068241516181732,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.890268087387085,
      "num_tokens": 202459543.0,
      "step": 5818
    },
    {
      "epoch": 1.080594243268338,
      "grad_norm": 1.5467532403590438,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8759080171585083,
      "num_tokens": 202495905.0,
      "step": 5819
    },
    {
      "epoch": 1.0807799442896937,
      "grad_norm": 1.4484550661512943,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8752374649047852,
      "num_tokens": 202533563.0,
      "step": 5820
    },
    {
      "epoch": 1.0809656453110492,
      "grad_norm": 1.5970813868008396,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8772878646850586,
      "num_tokens": 202567077.0,
      "step": 5821
    },
    {
      "epoch": 1.0811513463324047,
      "grad_norm": 1.4867724397881525,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8861560821533203,
      "num_tokens": 202597301.0,
      "step": 5822
    },
    {
      "epoch": 1.0813370473537605,
      "grad_norm": 1.3641680467401391,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8752074241638184,
      "num_tokens": 202643116.0,
      "step": 5823
    },
    {
      "epoch": 1.081522748375116,
      "grad_norm": 1.4933550565967282,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8807057738304138,
      "num_tokens": 202675917.0,
      "step": 5824
    },
    {
      "epoch": 1.0817084493964717,
      "grad_norm": 1.5187720879312474,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8712164759635925,
      "num_tokens": 202707701.0,
      "step": 5825
    },
    {
      "epoch": 1.0818941504178272,
      "grad_norm": 1.5675219369443574,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8833876848220825,
      "num_tokens": 202740878.0,
      "step": 5826
    },
    {
      "epoch": 1.082079851439183,
      "grad_norm": 1.5161883293832779,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8705039024353027,
      "num_tokens": 202776150.0,
      "step": 5827
    },
    {
      "epoch": 1.0822655524605385,
      "grad_norm": 1.3458935991947292,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8864374160766602,
      "num_tokens": 202816562.0,
      "step": 5828
    },
    {
      "epoch": 1.0824512534818942,
      "grad_norm": 1.5929195516435652,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8807390332221985,
      "num_tokens": 202848253.0,
      "step": 5829
    },
    {
      "epoch": 1.0826369545032497,
      "grad_norm": 1.5000439675150687,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8681552410125732,
      "num_tokens": 202883975.0,
      "step": 5830
    },
    {
      "epoch": 1.0828226555246054,
      "grad_norm": 1.5517604155426836,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8853568434715271,
      "num_tokens": 202916978.0,
      "step": 5831
    },
    {
      "epoch": 1.083008356545961,
      "grad_norm": 1.3775628115321275,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.889563262462616,
      "num_tokens": 202954692.0,
      "step": 5832
    },
    {
      "epoch": 1.0831940575673167,
      "grad_norm": 1.5624543204100192,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8806387782096863,
      "num_tokens": 202986298.0,
      "step": 5833
    },
    {
      "epoch": 1.0833797585886722,
      "grad_norm": 1.5115158599953633,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8815667629241943,
      "num_tokens": 203016767.0,
      "step": 5834
    },
    {
      "epoch": 1.083565459610028,
      "grad_norm": 1.5963307416133534,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8638724684715271,
      "num_tokens": 203053974.0,
      "step": 5835
    },
    {
      "epoch": 1.0837511606313834,
      "grad_norm": 1.3847945455900899,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8886747360229492,
      "num_tokens": 203091089.0,
      "step": 5836
    },
    {
      "epoch": 1.0839368616527392,
      "grad_norm": 1.4638679575542635,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8771153688430786,
      "num_tokens": 203126887.0,
      "step": 5837
    },
    {
      "epoch": 1.0841225626740947,
      "grad_norm": 1.4085825586152605,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8966398239135742,
      "num_tokens": 203159671.0,
      "step": 5838
    },
    {
      "epoch": 1.0843082636954504,
      "grad_norm": 1.4505633749261662,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8836071491241455,
      "num_tokens": 203193822.0,
      "step": 5839
    },
    {
      "epoch": 1.084493964716806,
      "grad_norm": 1.4179631313838532,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8811044692993164,
      "num_tokens": 203229785.0,
      "step": 5840
    },
    {
      "epoch": 1.0846796657381617,
      "grad_norm": 1.402448788670461,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8838338255882263,
      "num_tokens": 203266121.0,
      "step": 5841
    },
    {
      "epoch": 1.0848653667595172,
      "grad_norm": 1.4704453249684573,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8721588253974915,
      "num_tokens": 203301666.0,
      "step": 5842
    },
    {
      "epoch": 1.085051067780873,
      "grad_norm": 1.4675241958621859,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8942056894302368,
      "num_tokens": 203335239.0,
      "step": 5843
    },
    {
      "epoch": 1.0852367688022284,
      "grad_norm": 1.4163390117225994,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8851308822631836,
      "num_tokens": 203369259.0,
      "step": 5844
    },
    {
      "epoch": 1.085422469823584,
      "grad_norm": 1.5672039730096816,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8820390105247498,
      "num_tokens": 203400461.0,
      "step": 5845
    },
    {
      "epoch": 1.0856081708449397,
      "grad_norm": 1.5820535409954757,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8638793230056763,
      "num_tokens": 203434107.0,
      "step": 5846
    },
    {
      "epoch": 1.0857938718662952,
      "grad_norm": 1.5436079173460733,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8769218325614929,
      "num_tokens": 203467215.0,
      "step": 5847
    },
    {
      "epoch": 1.085979572887651,
      "grad_norm": 1.764084854449045,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8630696535110474,
      "num_tokens": 203493438.0,
      "step": 5848
    },
    {
      "epoch": 1.0861652739090064,
      "grad_norm": 1.5999609099990904,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8861913084983826,
      "num_tokens": 203522143.0,
      "step": 5849
    },
    {
      "epoch": 1.0863509749303621,
      "grad_norm": 1.5160483502428723,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8738410472869873,
      "num_tokens": 203557598.0,
      "step": 5850
    },
    {
      "epoch": 1.0865366759517177,
      "grad_norm": 1.4659835498555691,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8869340419769287,
      "num_tokens": 203591486.0,
      "step": 5851
    },
    {
      "epoch": 1.0867223769730734,
      "grad_norm": 1.500265124764864,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8706611394882202,
      "num_tokens": 203628566.0,
      "step": 5852
    },
    {
      "epoch": 1.086908077994429,
      "grad_norm": 1.3542858415014931,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8881931900978088,
      "num_tokens": 203665886.0,
      "step": 5853
    },
    {
      "epoch": 1.0870937790157846,
      "grad_norm": 1.5077054930257023,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8769665956497192,
      "num_tokens": 203698534.0,
      "step": 5854
    },
    {
      "epoch": 1.0872794800371401,
      "grad_norm": 1.5947930995431203,
      "learning_rate": 1e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8577888607978821,
      "num_tokens": 203733133.0,
      "step": 5855
    },
    {
      "epoch": 1.0874651810584959,
      "grad_norm": 1.444685490474273,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.873309850692749,
      "num_tokens": 203772846.0,
      "step": 5856
    },
    {
      "epoch": 1.0876508820798514,
      "grad_norm": 1.540618804106706,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8736628293991089,
      "num_tokens": 203805096.0,
      "step": 5857
    },
    {
      "epoch": 1.0878365831012071,
      "grad_norm": 1.498220601179472,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8691720962524414,
      "num_tokens": 203841416.0,
      "step": 5858
    },
    {
      "epoch": 1.0880222841225626,
      "grad_norm": 1.5689978157346458,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8664026856422424,
      "num_tokens": 203875127.0,
      "step": 5859
    },
    {
      "epoch": 1.0882079851439184,
      "grad_norm": 1.65201865733498,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8749057054519653,
      "num_tokens": 203904955.0,
      "step": 5860
    },
    {
      "epoch": 1.0883936861652739,
      "grad_norm": 1.4167708716216543,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8906725645065308,
      "num_tokens": 203940117.0,
      "step": 5861
    },
    {
      "epoch": 1.0885793871866296,
      "grad_norm": 1.559227290748245,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8762173652648926,
      "num_tokens": 203974496.0,
      "step": 5862
    },
    {
      "epoch": 1.0887650882079851,
      "grad_norm": 1.4256014920300246,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8828102946281433,
      "num_tokens": 204013063.0,
      "step": 5863
    },
    {
      "epoch": 1.0889507892293409,
      "grad_norm": 1.4655917118086255,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8666466474533081,
      "num_tokens": 204052718.0,
      "step": 5864
    },
    {
      "epoch": 1.0891364902506964,
      "grad_norm": 1.4989500568644543,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8874646425247192,
      "num_tokens": 204083367.0,
      "step": 5865
    },
    {
      "epoch": 1.089322191272052,
      "grad_norm": 1.4116841128927395,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8834688663482666,
      "num_tokens": 204123041.0,
      "step": 5866
    },
    {
      "epoch": 1.0895078922934076,
      "grad_norm": 1.481813265121859,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8806049823760986,
      "num_tokens": 204157614.0,
      "step": 5867
    },
    {
      "epoch": 1.0896935933147631,
      "grad_norm": 1.529680973166231,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8567203879356384,
      "num_tokens": 204194855.0,
      "step": 5868
    },
    {
      "epoch": 1.0898792943361189,
      "grad_norm": 1.5796812343096023,
      "learning_rate": 1e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8613976240158081,
      "num_tokens": 204229320.0,
      "step": 5869
    },
    {
      "epoch": 1.0900649953574744,
      "grad_norm": 1.5241823356424602,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8712360858917236,
      "num_tokens": 204265942.0,
      "step": 5870
    },
    {
      "epoch": 1.09025069637883,
      "grad_norm": 1.4290591073479262,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8771302700042725,
      "num_tokens": 204306895.0,
      "step": 5871
    },
    {
      "epoch": 1.0904363974001856,
      "grad_norm": 1.532523973012078,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8769551515579224,
      "num_tokens": 204341459.0,
      "step": 5872
    },
    {
      "epoch": 1.0906220984215413,
      "grad_norm": 1.34547721519794,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8777385950088501,
      "num_tokens": 204385820.0,
      "step": 5873
    },
    {
      "epoch": 1.0908077994428969,
      "grad_norm": 1.6058957617512675,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8653125762939453,
      "num_tokens": 204420480.0,
      "step": 5874
    },
    {
      "epoch": 1.0909935004642526,
      "grad_norm": 1.811067558887719,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8723812103271484,
      "num_tokens": 204446526.0,
      "step": 5875
    },
    {
      "epoch": 1.091179201485608,
      "grad_norm": 1.439346783700605,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.875486433506012,
      "num_tokens": 204479214.0,
      "step": 5876
    },
    {
      "epoch": 1.0913649025069638,
      "grad_norm": 1.416786566816867,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8767672181129456,
      "num_tokens": 204516139.0,
      "step": 5877
    },
    {
      "epoch": 1.0915506035283193,
      "grad_norm": 1.467211175042625,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8845599293708801,
      "num_tokens": 204549680.0,
      "step": 5878
    },
    {
      "epoch": 1.091736304549675,
      "grad_norm": 1.3944649773755509,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8903825283050537,
      "num_tokens": 204584559.0,
      "step": 5879
    },
    {
      "epoch": 1.0919220055710306,
      "grad_norm": 1.581773819018903,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8806499242782593,
      "num_tokens": 204617754.0,
      "step": 5880
    },
    {
      "epoch": 1.0921077065923863,
      "grad_norm": 1.5697268332674246,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8772866725921631,
      "num_tokens": 204652025.0,
      "step": 5881
    },
    {
      "epoch": 1.0922934076137418,
      "grad_norm": 1.4893438187100985,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8961046934127808,
      "num_tokens": 204683848.0,
      "step": 5882
    },
    {
      "epoch": 1.0924791086350976,
      "grad_norm": 1.6335888218945747,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8847703337669373,
      "num_tokens": 204709909.0,
      "step": 5883
    },
    {
      "epoch": 1.092664809656453,
      "grad_norm": 1.3469412099259344,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8729183673858643,
      "num_tokens": 204751302.0,
      "step": 5884
    },
    {
      "epoch": 1.0928505106778088,
      "grad_norm": 1.5100329546861293,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8688081502914429,
      "num_tokens": 204785350.0,
      "step": 5885
    },
    {
      "epoch": 1.0930362116991643,
      "grad_norm": 1.4690460236501122,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.875335156917572,
      "num_tokens": 204821249.0,
      "step": 5886
    },
    {
      "epoch": 1.09322191272052,
      "grad_norm": 1.3469528484968476,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.889523446559906,
      "num_tokens": 204861372.0,
      "step": 5887
    },
    {
      "epoch": 1.0934076137418756,
      "grad_norm": 1.4363744420668696,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8737555742263794,
      "num_tokens": 204900412.0,
      "step": 5888
    },
    {
      "epoch": 1.0935933147632313,
      "grad_norm": 1.6016778679258303,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.86965012550354,
      "num_tokens": 204932422.0,
      "step": 5889
    },
    {
      "epoch": 1.0937790157845868,
      "grad_norm": 1.655507787206979,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8861005306243896,
      "num_tokens": 204959838.0,
      "step": 5890
    },
    {
      "epoch": 1.0939647168059423,
      "grad_norm": 1.5220098045331303,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8846505880355835,
      "num_tokens": 204995068.0,
      "step": 5891
    },
    {
      "epoch": 1.094150417827298,
      "grad_norm": 1.4001200187020995,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8760586977005005,
      "num_tokens": 205034671.0,
      "step": 5892
    },
    {
      "epoch": 1.0943361188486538,
      "grad_norm": 1.461216511462813,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8755711317062378,
      "num_tokens": 205068504.0,
      "step": 5893
    },
    {
      "epoch": 1.0945218198700093,
      "grad_norm": 1.437849464259874,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8715904951095581,
      "num_tokens": 205104105.0,
      "step": 5894
    },
    {
      "epoch": 1.0947075208913648,
      "grad_norm": 1.6021882712106275,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8739341497421265,
      "num_tokens": 205136979.0,
      "step": 5895
    },
    {
      "epoch": 1.0948932219127205,
      "grad_norm": 1.4856998084231576,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8730670809745789,
      "num_tokens": 205171820.0,
      "step": 5896
    },
    {
      "epoch": 1.095078922934076,
      "grad_norm": 1.582664790099438,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8852376937866211,
      "num_tokens": 205202081.0,
      "step": 5897
    },
    {
      "epoch": 1.0952646239554318,
      "grad_norm": 1.4891230563657498,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8707934617996216,
      "num_tokens": 205241708.0,
      "step": 5898
    },
    {
      "epoch": 1.0954503249767873,
      "grad_norm": 1.5448423450060658,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8727946281433105,
      "num_tokens": 205275990.0,
      "step": 5899
    },
    {
      "epoch": 1.095636025998143,
      "grad_norm": 1.6936576622704649,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8678710460662842,
      "num_tokens": 205308438.0,
      "step": 5900
    },
    {
      "epoch": 1.0958217270194985,
      "grad_norm": 1.4759426475695927,
      "learning_rate": 1e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8654986619949341,
      "num_tokens": 205344510.0,
      "step": 5901
    },
    {
      "epoch": 1.0960074280408543,
      "grad_norm": 1.4509863248654256,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8719092011451721,
      "num_tokens": 205382099.0,
      "step": 5902
    },
    {
      "epoch": 1.0961931290622098,
      "grad_norm": 1.4976088498922522,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.878496527671814,
      "num_tokens": 205414945.0,
      "step": 5903
    },
    {
      "epoch": 1.0963788300835655,
      "grad_norm": 1.5115076669446397,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8856276273727417,
      "num_tokens": 205450484.0,
      "step": 5904
    },
    {
      "epoch": 1.096564531104921,
      "grad_norm": 1.576774441573001,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8772373199462891,
      "num_tokens": 205482011.0,
      "step": 5905
    },
    {
      "epoch": 1.0967502321262768,
      "grad_norm": 1.5271280541258951,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8653374314308167,
      "num_tokens": 205516863.0,
      "step": 5906
    },
    {
      "epoch": 1.0969359331476323,
      "grad_norm": 1.6551982204150044,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8665928840637207,
      "num_tokens": 205545090.0,
      "step": 5907
    },
    {
      "epoch": 1.097121634168988,
      "grad_norm": 1.5175975990910584,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.885573148727417,
      "num_tokens": 205577507.0,
      "step": 5908
    },
    {
      "epoch": 1.0973073351903435,
      "grad_norm": 1.5397692365388196,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.879338264465332,
      "num_tokens": 205609578.0,
      "step": 5909
    },
    {
      "epoch": 1.0974930362116992,
      "grad_norm": 1.8864637717418506,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8692121505737305,
      "num_tokens": 205637378.0,
      "step": 5910
    },
    {
      "epoch": 1.0976787372330548,
      "grad_norm": 1.5824686766214422,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8696684241294861,
      "num_tokens": 205672817.0,
      "step": 5911
    },
    {
      "epoch": 1.0978644382544105,
      "grad_norm": 1.4163851010453044,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8760733008384705,
      "num_tokens": 205710103.0,
      "step": 5912
    },
    {
      "epoch": 1.098050139275766,
      "grad_norm": 1.4143788507131503,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8849941492080688,
      "num_tokens": 205745107.0,
      "step": 5913
    },
    {
      "epoch": 1.0982358402971217,
      "grad_norm": 1.466501416227474,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.879319429397583,
      "num_tokens": 205779297.0,
      "step": 5914
    },
    {
      "epoch": 1.0984215413184772,
      "grad_norm": 1.5300764767648316,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8712202310562134,
      "num_tokens": 205814313.0,
      "step": 5915
    },
    {
      "epoch": 1.098607242339833,
      "grad_norm": 1.5663723007549577,
      "learning_rate": 1e-06,
      "loss": 0.3564,
      "mean_token_accuracy": 0.8763245344161987,
      "num_tokens": 205846638.0,
      "step": 5916
    },
    {
      "epoch": 1.0987929433611885,
      "grad_norm": 1.4452334110054803,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8848576545715332,
      "num_tokens": 205882702.0,
      "step": 5917
    },
    {
      "epoch": 1.098978644382544,
      "grad_norm": 1.549129739407019,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8859184384346008,
      "num_tokens": 205917179.0,
      "step": 5918
    },
    {
      "epoch": 1.0991643454038997,
      "grad_norm": 1.4961000895395813,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8919198513031006,
      "num_tokens": 205948351.0,
      "step": 5919
    },
    {
      "epoch": 1.0993500464252552,
      "grad_norm": 1.4520415085563934,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8837672472000122,
      "num_tokens": 205982339.0,
      "step": 5920
    },
    {
      "epoch": 1.099535747446611,
      "grad_norm": 1.4141652407334846,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8750413060188293,
      "num_tokens": 206025343.0,
      "step": 5921
    },
    {
      "epoch": 1.0997214484679665,
      "grad_norm": 1.5351097265529272,
      "learning_rate": 1e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8561115264892578,
      "num_tokens": 206059511.0,
      "step": 5922
    },
    {
      "epoch": 1.0999071494893222,
      "grad_norm": 1.5357152339319367,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8670454621315002,
      "num_tokens": 206092272.0,
      "step": 5923
    },
    {
      "epoch": 1.1000928505106777,
      "grad_norm": 1.3660703277242288,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.882132887840271,
      "num_tokens": 206135171.0,
      "step": 5924
    },
    {
      "epoch": 1.1002785515320335,
      "grad_norm": 1.4509618003148335,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8882405161857605,
      "num_tokens": 206173405.0,
      "step": 5925
    },
    {
      "epoch": 1.100464252553389,
      "grad_norm": 1.5096199248004794,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8765113353729248,
      "num_tokens": 206211981.0,
      "step": 5926
    },
    {
      "epoch": 1.1006499535747447,
      "grad_norm": 1.5156721057518974,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8717286586761475,
      "num_tokens": 206247879.0,
      "step": 5927
    },
    {
      "epoch": 1.1008356545961002,
      "grad_norm": 1.5528630521936337,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8755003213882446,
      "num_tokens": 206280527.0,
      "step": 5928
    },
    {
      "epoch": 1.101021355617456,
      "grad_norm": 1.6145162271733535,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8846081495285034,
      "num_tokens": 206308189.0,
      "step": 5929
    },
    {
      "epoch": 1.1012070566388115,
      "grad_norm": 1.4210877461959377,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.876392662525177,
      "num_tokens": 206347214.0,
      "step": 5930
    },
    {
      "epoch": 1.1013927576601672,
      "grad_norm": 1.556774499230922,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8737044334411621,
      "num_tokens": 206377164.0,
      "step": 5931
    },
    {
      "epoch": 1.1015784586815227,
      "grad_norm": 1.5092079257476745,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8688466548919678,
      "num_tokens": 206414445.0,
      "step": 5932
    },
    {
      "epoch": 1.1017641597028784,
      "grad_norm": 1.4934775681685906,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8794342875480652,
      "num_tokens": 206448690.0,
      "step": 5933
    },
    {
      "epoch": 1.101949860724234,
      "grad_norm": 1.4456002640526544,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8745023012161255,
      "num_tokens": 206483807.0,
      "step": 5934
    },
    {
      "epoch": 1.1021355617455897,
      "grad_norm": 1.5921742686138844,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8729463815689087,
      "num_tokens": 206515264.0,
      "step": 5935
    },
    {
      "epoch": 1.1023212627669452,
      "grad_norm": 1.5520161151240863,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8762947916984558,
      "num_tokens": 206547061.0,
      "step": 5936
    },
    {
      "epoch": 1.102506963788301,
      "grad_norm": 1.4397374115574875,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8858246803283691,
      "num_tokens": 206582292.0,
      "step": 5937
    },
    {
      "epoch": 1.1026926648096564,
      "grad_norm": 1.4733224527670414,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8722050189971924,
      "num_tokens": 206618087.0,
      "step": 5938
    },
    {
      "epoch": 1.1028783658310122,
      "grad_norm": 1.3897085399292295,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8840562105178833,
      "num_tokens": 206655469.0,
      "step": 5939
    },
    {
      "epoch": 1.1030640668523677,
      "grad_norm": 1.5889623919300504,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8684073686599731,
      "num_tokens": 206686231.0,
      "step": 5940
    },
    {
      "epoch": 1.1032497678737232,
      "grad_norm": 1.493651666434071,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8844759464263916,
      "num_tokens": 206718760.0,
      "step": 5941
    },
    {
      "epoch": 1.103435468895079,
      "grad_norm": 1.3947650477424463,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8737540245056152,
      "num_tokens": 206758634.0,
      "step": 5942
    },
    {
      "epoch": 1.1036211699164344,
      "grad_norm": 1.5912970968594367,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8621883392333984,
      "num_tokens": 206793688.0,
      "step": 5943
    },
    {
      "epoch": 1.1038068709377902,
      "grad_norm": 1.3978686841882464,
      "learning_rate": 1e-06,
      "loss": 0.305,
      "mean_token_accuracy": 0.8930243253707886,
      "num_tokens": 206827556.0,
      "step": 5944
    },
    {
      "epoch": 1.1039925719591457,
      "grad_norm": 1.6645560614722983,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8768930435180664,
      "num_tokens": 206856172.0,
      "step": 5945
    },
    {
      "epoch": 1.1041782729805014,
      "grad_norm": 1.5320886149493103,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8922973871231079,
      "num_tokens": 206889986.0,
      "step": 5946
    },
    {
      "epoch": 1.104363974001857,
      "grad_norm": 1.4077576243101466,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8681489825248718,
      "num_tokens": 206932419.0,
      "step": 5947
    },
    {
      "epoch": 1.1045496750232127,
      "grad_norm": 1.4836289495534751,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8694393038749695,
      "num_tokens": 206969301.0,
      "step": 5948
    },
    {
      "epoch": 1.1047353760445682,
      "grad_norm": 1.6633307062600802,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8804352283477783,
      "num_tokens": 206999765.0,
      "step": 5949
    },
    {
      "epoch": 1.104921077065924,
      "grad_norm": 1.536763420568933,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8733588457107544,
      "num_tokens": 207031730.0,
      "step": 5950
    },
    {
      "epoch": 1.1051067780872794,
      "grad_norm": 1.562173162016109,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8664432764053345,
      "num_tokens": 207066450.0,
      "step": 5951
    },
    {
      "epoch": 1.1052924791086352,
      "grad_norm": 1.4529668977174042,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.86127769947052,
      "num_tokens": 207103944.0,
      "step": 5952
    },
    {
      "epoch": 1.1054781801299907,
      "grad_norm": 1.5812194796925934,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8882050514221191,
      "num_tokens": 207134010.0,
      "step": 5953
    },
    {
      "epoch": 1.1056638811513464,
      "grad_norm": 1.5315314317886628,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8777902722358704,
      "num_tokens": 207171311.0,
      "step": 5954
    },
    {
      "epoch": 1.105849582172702,
      "grad_norm": 1.4746432723174145,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.866136372089386,
      "num_tokens": 207205865.0,
      "step": 5955
    },
    {
      "epoch": 1.1060352831940576,
      "grad_norm": 1.5464343698408471,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8780847787857056,
      "num_tokens": 207238645.0,
      "step": 5956
    },
    {
      "epoch": 1.1062209842154132,
      "grad_norm": 1.688150885037684,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8781469464302063,
      "num_tokens": 207264981.0,
      "step": 5957
    },
    {
      "epoch": 1.1064066852367689,
      "grad_norm": 1.5312854522421107,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8662064671516418,
      "num_tokens": 207299775.0,
      "step": 5958
    },
    {
      "epoch": 1.1065923862581244,
      "grad_norm": 1.508797474513159,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8772974014282227,
      "num_tokens": 207334917.0,
      "step": 5959
    },
    {
      "epoch": 1.1067780872794801,
      "grad_norm": 1.4657355952115356,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8700616955757141,
      "num_tokens": 207376196.0,
      "step": 5960
    },
    {
      "epoch": 1.1069637883008356,
      "grad_norm": 1.625640926291882,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8661926984786987,
      "num_tokens": 207405913.0,
      "step": 5961
    },
    {
      "epoch": 1.1071494893221914,
      "grad_norm": 1.5571253535839362,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8822850584983826,
      "num_tokens": 207437423.0,
      "step": 5962
    },
    {
      "epoch": 1.1073351903435469,
      "grad_norm": 1.3915597045746588,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8749252557754517,
      "num_tokens": 207475274.0,
      "step": 5963
    },
    {
      "epoch": 1.1075208913649024,
      "grad_norm": 1.4809258917839103,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8729767799377441,
      "num_tokens": 207507466.0,
      "step": 5964
    },
    {
      "epoch": 1.1077065923862581,
      "grad_norm": 1.52245832893065,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.885901689529419,
      "num_tokens": 207545210.0,
      "step": 5965
    },
    {
      "epoch": 1.1078922934076139,
      "grad_norm": 1.678304982897542,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8701775074005127,
      "num_tokens": 207574204.0,
      "step": 5966
    },
    {
      "epoch": 1.1080779944289694,
      "grad_norm": 1.7028920280387405,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8725394010543823,
      "num_tokens": 207605156.0,
      "step": 5967
    },
    {
      "epoch": 1.1082636954503249,
      "grad_norm": 1.630333961436783,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8754032254219055,
      "num_tokens": 207634561.0,
      "step": 5968
    },
    {
      "epoch": 1.1084493964716806,
      "grad_norm": 1.6210973179779595,
      "learning_rate": 1e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8586191534996033,
      "num_tokens": 207664637.0,
      "step": 5969
    },
    {
      "epoch": 1.1086350974930361,
      "grad_norm": 1.4582039570082983,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8900171518325806,
      "num_tokens": 207701570.0,
      "step": 5970
    },
    {
      "epoch": 1.1088207985143919,
      "grad_norm": 1.5377622280725844,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8753820657730103,
      "num_tokens": 207737927.0,
      "step": 5971
    },
    {
      "epoch": 1.1090064995357474,
      "grad_norm": 1.6274734541755382,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8600550889968872,
      "num_tokens": 207770367.0,
      "step": 5972
    },
    {
      "epoch": 1.109192200557103,
      "grad_norm": 1.7350588226027397,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.884401798248291,
      "num_tokens": 207802048.0,
      "step": 5973
    },
    {
      "epoch": 1.1093779015784586,
      "grad_norm": 1.5846153312826738,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8712001442909241,
      "num_tokens": 207833282.0,
      "step": 5974
    },
    {
      "epoch": 1.1095636025998143,
      "grad_norm": 1.4695667869201732,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.887206494808197,
      "num_tokens": 207873047.0,
      "step": 5975
    },
    {
      "epoch": 1.1097493036211699,
      "grad_norm": 1.4377912651780262,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8757919073104858,
      "num_tokens": 207909125.0,
      "step": 5976
    },
    {
      "epoch": 1.1099350046425256,
      "grad_norm": 1.5708795054567453,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8637540340423584,
      "num_tokens": 207943651.0,
      "step": 5977
    },
    {
      "epoch": 1.110120705663881,
      "grad_norm": 1.6422656758459229,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8675570487976074,
      "num_tokens": 207978013.0,
      "step": 5978
    },
    {
      "epoch": 1.1103064066852368,
      "grad_norm": 1.633179778438278,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.889668345451355,
      "num_tokens": 208005730.0,
      "step": 5979
    },
    {
      "epoch": 1.1104921077065923,
      "grad_norm": 1.5094541570167994,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8793636560440063,
      "num_tokens": 208045542.0,
      "step": 5980
    },
    {
      "epoch": 1.110677808727948,
      "grad_norm": 1.5653586984326147,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8770207166671753,
      "num_tokens": 208080372.0,
      "step": 5981
    },
    {
      "epoch": 1.1108635097493036,
      "grad_norm": 1.5170194951022067,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8879805207252502,
      "num_tokens": 208110478.0,
      "step": 5982
    },
    {
      "epoch": 1.1110492107706593,
      "grad_norm": 1.5753938298693628,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8719385266304016,
      "num_tokens": 208144508.0,
      "step": 5983
    },
    {
      "epoch": 1.1112349117920148,
      "grad_norm": 1.500952410216923,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8925693035125732,
      "num_tokens": 208178084.0,
      "step": 5984
    },
    {
      "epoch": 1.1114206128133706,
      "grad_norm": 1.4884058309815584,
      "learning_rate": 1e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8657059073448181,
      "num_tokens": 208220309.0,
      "step": 5985
    },
    {
      "epoch": 1.111606313834726,
      "grad_norm": 2.323911515268005,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8652900457382202,
      "num_tokens": 208248991.0,
      "step": 5986
    },
    {
      "epoch": 1.1117920148560818,
      "grad_norm": 1.5746769194308647,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8862327337265015,
      "num_tokens": 208278625.0,
      "step": 5987
    },
    {
      "epoch": 1.1119777158774373,
      "grad_norm": 1.5224772163007594,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8691943883895874,
      "num_tokens": 208313410.0,
      "step": 5988
    },
    {
      "epoch": 1.112163416898793,
      "grad_norm": 1.537341229464868,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8831755518913269,
      "num_tokens": 208346067.0,
      "step": 5989
    },
    {
      "epoch": 1.1123491179201486,
      "grad_norm": 1.4228618953444028,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8805931210517883,
      "num_tokens": 208386696.0,
      "step": 5990
    },
    {
      "epoch": 1.112534818941504,
      "grad_norm": 1.610785761998778,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8734512329101562,
      "num_tokens": 208418030.0,
      "step": 5991
    },
    {
      "epoch": 1.1127205199628598,
      "grad_norm": 1.388945149794084,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8814071416854858,
      "num_tokens": 208454136.0,
      "step": 5992
    },
    {
      "epoch": 1.1129062209842153,
      "grad_norm": 1.4587740477581992,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.875011682510376,
      "num_tokens": 208492246.0,
      "step": 5993
    },
    {
      "epoch": 1.113091922005571,
      "grad_norm": 1.3868739143575093,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8840168118476868,
      "num_tokens": 208530874.0,
      "step": 5994
    },
    {
      "epoch": 1.1132776230269266,
      "grad_norm": 1.6331165536314582,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8756924867630005,
      "num_tokens": 208561056.0,
      "step": 5995
    },
    {
      "epoch": 1.1134633240482823,
      "grad_norm": 1.477728427162915,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.879396915435791,
      "num_tokens": 208601996.0,
      "step": 5996
    },
    {
      "epoch": 1.1136490250696378,
      "grad_norm": 1.5387858260619613,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.883651614189148,
      "num_tokens": 208631980.0,
      "step": 5997
    },
    {
      "epoch": 1.1138347260909935,
      "grad_norm": 1.4218688595845168,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8836930990219116,
      "num_tokens": 208666515.0,
      "step": 5998
    },
    {
      "epoch": 1.114020427112349,
      "grad_norm": 1.5437261786791385,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8807774186134338,
      "num_tokens": 208700144.0,
      "step": 5999
    },
    {
      "epoch": 1.1142061281337048,
      "grad_norm": 1.5892926635323434,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8773446083068848,
      "num_tokens": 208729726.0,
      "step": 6000
    },
    {
      "epoch": 1.1143918291550603,
      "grad_norm": 1.5434305035008276,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8789619207382202,
      "num_tokens": 208764495.0,
      "step": 6001
    },
    {
      "epoch": 1.114577530176416,
      "grad_norm": 1.5710756967883777,
      "learning_rate": 1e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8593476414680481,
      "num_tokens": 208800561.0,
      "step": 6002
    },
    {
      "epoch": 1.1147632311977715,
      "grad_norm": 1.4639529425180873,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8726935386657715,
      "num_tokens": 208837500.0,
      "step": 6003
    },
    {
      "epoch": 1.1149489322191273,
      "grad_norm": 1.4412001294307426,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8743305206298828,
      "num_tokens": 208875670.0,
      "step": 6004
    },
    {
      "epoch": 1.1151346332404828,
      "grad_norm": 1.5239793755623063,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8764983415603638,
      "num_tokens": 208911046.0,
      "step": 6005
    },
    {
      "epoch": 1.1153203342618385,
      "grad_norm": 1.470206197568518,
      "learning_rate": 1e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8563770055770874,
      "num_tokens": 208952329.0,
      "step": 6006
    },
    {
      "epoch": 1.115506035283194,
      "grad_norm": 1.7102357840269147,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8801068067550659,
      "num_tokens": 208980067.0,
      "step": 6007
    },
    {
      "epoch": 1.1156917363045498,
      "grad_norm": 1.3814274081461082,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8829541206359863,
      "num_tokens": 209016979.0,
      "step": 6008
    },
    {
      "epoch": 1.1158774373259053,
      "grad_norm": 1.5249010787878092,
      "learning_rate": 1e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8644770979881287,
      "num_tokens": 209051958.0,
      "step": 6009
    },
    {
      "epoch": 1.116063138347261,
      "grad_norm": 1.5844418801928084,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8808447122573853,
      "num_tokens": 209084579.0,
      "step": 6010
    },
    {
      "epoch": 1.1162488393686165,
      "grad_norm": 1.6325800331212785,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8866803646087646,
      "num_tokens": 209114224.0,
      "step": 6011
    },
    {
      "epoch": 1.1164345403899723,
      "grad_norm": 1.509294808929709,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8688081502914429,
      "num_tokens": 209151021.0,
      "step": 6012
    },
    {
      "epoch": 1.1166202414113278,
      "grad_norm": 1.5420258559358202,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8663853406906128,
      "num_tokens": 209187030.0,
      "step": 6013
    },
    {
      "epoch": 1.1168059424326833,
      "grad_norm": 1.4909016577175311,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.878456711769104,
      "num_tokens": 209220165.0,
      "step": 6014
    },
    {
      "epoch": 1.116991643454039,
      "grad_norm": 1.4795635028478702,
      "learning_rate": 1e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8646248579025269,
      "num_tokens": 209256109.0,
      "step": 6015
    },
    {
      "epoch": 1.1171773444753945,
      "grad_norm": 1.3886364932941566,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8833630084991455,
      "num_tokens": 209293572.0,
      "step": 6016
    },
    {
      "epoch": 1.1173630454967503,
      "grad_norm": 1.4226273136564007,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8845064043998718,
      "num_tokens": 209329544.0,
      "step": 6017
    },
    {
      "epoch": 1.1175487465181058,
      "grad_norm": 1.48607222258785,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8740063905715942,
      "num_tokens": 209364114.0,
      "step": 6018
    },
    {
      "epoch": 1.1177344475394615,
      "grad_norm": 1.569957870810149,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8781148791313171,
      "num_tokens": 209394669.0,
      "step": 6019
    },
    {
      "epoch": 1.117920148560817,
      "grad_norm": 1.4719082602332587,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.878135085105896,
      "num_tokens": 209430297.0,
      "step": 6020
    },
    {
      "epoch": 1.1181058495821727,
      "grad_norm": 1.5606266767433783,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8565640449523926,
      "num_tokens": 209466060.0,
      "step": 6021
    },
    {
      "epoch": 1.1182915506035283,
      "grad_norm": 1.4514651269048087,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8837834000587463,
      "num_tokens": 209503053.0,
      "step": 6022
    },
    {
      "epoch": 1.118477251624884,
      "grad_norm": 1.4457647053232598,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8741559386253357,
      "num_tokens": 209540484.0,
      "step": 6023
    },
    {
      "epoch": 1.1186629526462395,
      "grad_norm": 1.4262080196454088,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8729444146156311,
      "num_tokens": 209580189.0,
      "step": 6024
    },
    {
      "epoch": 1.1188486536675952,
      "grad_norm": 1.705073328750231,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8747955560684204,
      "num_tokens": 209607659.0,
      "step": 6025
    },
    {
      "epoch": 1.1190343546889507,
      "grad_norm": 1.5461601025959362,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8730827569961548,
      "num_tokens": 209638199.0,
      "step": 6026
    },
    {
      "epoch": 1.1192200557103065,
      "grad_norm": 1.531315662677108,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8817952275276184,
      "num_tokens": 209670245.0,
      "step": 6027
    },
    {
      "epoch": 1.119405756731662,
      "grad_norm": 1.5654953724279457,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8816331624984741,
      "num_tokens": 209702505.0,
      "step": 6028
    },
    {
      "epoch": 1.1195914577530177,
      "grad_norm": 1.4336354439684285,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8768457174301147,
      "num_tokens": 209741998.0,
      "step": 6029
    },
    {
      "epoch": 1.1197771587743732,
      "grad_norm": 1.3510943395148074,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8863657712936401,
      "num_tokens": 209780406.0,
      "step": 6030
    },
    {
      "epoch": 1.119962859795729,
      "grad_norm": 1.5378338689706315,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8781048059463501,
      "num_tokens": 209813137.0,
      "step": 6031
    },
    {
      "epoch": 1.1201485608170845,
      "grad_norm": 1.633568248078696,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8688093423843384,
      "num_tokens": 209842372.0,
      "step": 6032
    },
    {
      "epoch": 1.1203342618384402,
      "grad_norm": 1.515817650142286,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8739979267120361,
      "num_tokens": 209876065.0,
      "step": 6033
    },
    {
      "epoch": 1.1205199628597957,
      "grad_norm": 1.5029224056367976,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8748182058334351,
      "num_tokens": 209914743.0,
      "step": 6034
    },
    {
      "epoch": 1.1207056638811514,
      "grad_norm": 1.5504644984606277,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8872017860412598,
      "num_tokens": 209947550.0,
      "step": 6035
    },
    {
      "epoch": 1.120891364902507,
      "grad_norm": 1.558364714857516,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8804049491882324,
      "num_tokens": 209977242.0,
      "step": 6036
    },
    {
      "epoch": 1.1210770659238625,
      "grad_norm": 1.5922681518360797,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8867891430854797,
      "num_tokens": 210006073.0,
      "step": 6037
    },
    {
      "epoch": 1.1212627669452182,
      "grad_norm": 1.4939498635375021,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8868997097015381,
      "num_tokens": 210040299.0,
      "step": 6038
    },
    {
      "epoch": 1.1214484679665737,
      "grad_norm": 1.5176581590351432,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8841451406478882,
      "num_tokens": 210077328.0,
      "step": 6039
    },
    {
      "epoch": 1.1216341689879294,
      "grad_norm": 1.5308872866848582,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8744204044342041,
      "num_tokens": 210112187.0,
      "step": 6040
    },
    {
      "epoch": 1.121819870009285,
      "grad_norm": 1.7128258696118175,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8572656512260437,
      "num_tokens": 210144582.0,
      "step": 6041
    },
    {
      "epoch": 1.1220055710306407,
      "grad_norm": 1.5282872140880042,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8768211603164673,
      "num_tokens": 210180116.0,
      "step": 6042
    },
    {
      "epoch": 1.1221912720519962,
      "grad_norm": 1.532676540860364,
      "learning_rate": 1e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8530613780021667,
      "num_tokens": 210220823.0,
      "step": 6043
    },
    {
      "epoch": 1.122376973073352,
      "grad_norm": 1.5430920404578443,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8708754777908325,
      "num_tokens": 210257406.0,
      "step": 6044
    },
    {
      "epoch": 1.1225626740947074,
      "grad_norm": 1.4815797281692569,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8798420429229736,
      "num_tokens": 210296934.0,
      "step": 6045
    },
    {
      "epoch": 1.1227483751160632,
      "grad_norm": 1.5042286866786545,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8801043629646301,
      "num_tokens": 210334501.0,
      "step": 6046
    },
    {
      "epoch": 1.1229340761374187,
      "grad_norm": 1.5711562670233437,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8783655166625977,
      "num_tokens": 210365797.0,
      "step": 6047
    },
    {
      "epoch": 1.1231197771587744,
      "grad_norm": 1.7240137804571385,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8651968240737915,
      "num_tokens": 210395560.0,
      "step": 6048
    },
    {
      "epoch": 1.12330547818013,
      "grad_norm": 1.4977131589465738,
      "learning_rate": 1e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8575877547264099,
      "num_tokens": 210434112.0,
      "step": 6049
    },
    {
      "epoch": 1.1234911792014857,
      "grad_norm": 1.565488529619506,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8776178359985352,
      "num_tokens": 210466238.0,
      "step": 6050
    },
    {
      "epoch": 1.1236768802228412,
      "grad_norm": 1.6723881708942185,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8791708946228027,
      "num_tokens": 210497047.0,
      "step": 6051
    },
    {
      "epoch": 1.123862581244197,
      "grad_norm": 1.595978174874807,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.866782009601593,
      "num_tokens": 210526833.0,
      "step": 6052
    },
    {
      "epoch": 1.1240482822655524,
      "grad_norm": 1.6171406517569145,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8671032190322876,
      "num_tokens": 210560450.0,
      "step": 6053
    },
    {
      "epoch": 1.1242339832869082,
      "grad_norm": 1.5101634832645736,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.877208411693573,
      "num_tokens": 210598146.0,
      "step": 6054
    },
    {
      "epoch": 1.1244196843082637,
      "grad_norm": 1.5104308888461389,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8933441042900085,
      "num_tokens": 210629575.0,
      "step": 6055
    },
    {
      "epoch": 1.1246053853296194,
      "grad_norm": 1.6000773913762005,
      "learning_rate": 1e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.847987174987793,
      "num_tokens": 210668354.0,
      "step": 6056
    },
    {
      "epoch": 1.124791086350975,
      "grad_norm": 1.5537385678392641,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8808220028877258,
      "num_tokens": 210699935.0,
      "step": 6057
    },
    {
      "epoch": 1.1249767873723306,
      "grad_norm": 1.6284257128605444,
      "learning_rate": 1e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8601495027542114,
      "num_tokens": 210732988.0,
      "step": 6058
    },
    {
      "epoch": 1.1251624883936862,
      "grad_norm": 1.4963258195582734,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8702349066734314,
      "num_tokens": 210773723.0,
      "step": 6059
    },
    {
      "epoch": 1.1253481894150417,
      "grad_norm": 1.550678738374587,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8871867060661316,
      "num_tokens": 210806264.0,
      "step": 6060
    },
    {
      "epoch": 1.1255338904363974,
      "grad_norm": 1.5522582341837383,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8709474205970764,
      "num_tokens": 210841028.0,
      "step": 6061
    },
    {
      "epoch": 1.1257195914577531,
      "grad_norm": 1.5509899565959322,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8774361610412598,
      "num_tokens": 210876137.0,
      "step": 6062
    },
    {
      "epoch": 1.1259052924791086,
      "grad_norm": 1.5973085110028789,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.867951512336731,
      "num_tokens": 210909657.0,
      "step": 6063
    },
    {
      "epoch": 1.1260909935004642,
      "grad_norm": 1.5493961525551316,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8840477466583252,
      "num_tokens": 210940628.0,
      "step": 6064
    },
    {
      "epoch": 1.1262766945218199,
      "grad_norm": 1.4781982730360925,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8643096685409546,
      "num_tokens": 210979316.0,
      "step": 6065
    },
    {
      "epoch": 1.1264623955431754,
      "grad_norm": 1.5372745159044867,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8660932779312134,
      "num_tokens": 211015063.0,
      "step": 6066
    },
    {
      "epoch": 1.1266480965645311,
      "grad_norm": 1.5861534333081384,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8749301433563232,
      "num_tokens": 211048966.0,
      "step": 6067
    },
    {
      "epoch": 1.1268337975858866,
      "grad_norm": 1.4443406680577358,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8693313002586365,
      "num_tokens": 211086952.0,
      "step": 6068
    },
    {
      "epoch": 1.1270194986072424,
      "grad_norm": 1.5326349834657844,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8855473399162292,
      "num_tokens": 211119333.0,
      "step": 6069
    },
    {
      "epoch": 1.1272051996285979,
      "grad_norm": 1.4070494954967625,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8806473016738892,
      "num_tokens": 211157009.0,
      "step": 6070
    },
    {
      "epoch": 1.1273909006499536,
      "grad_norm": 1.5217418373679952,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8701847791671753,
      "num_tokens": 211194423.0,
      "step": 6071
    },
    {
      "epoch": 1.1275766016713091,
      "grad_norm": 1.4995006831275635,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8778382539749146,
      "num_tokens": 211227882.0,
      "step": 6072
    },
    {
      "epoch": 1.1277623026926649,
      "grad_norm": 1.4863455193274586,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8588711023330688,
      "num_tokens": 211268663.0,
      "step": 6073
    },
    {
      "epoch": 1.1279480037140204,
      "grad_norm": 1.367033113446459,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.882498562335968,
      "num_tokens": 211308977.0,
      "step": 6074
    },
    {
      "epoch": 1.128133704735376,
      "grad_norm": 1.4020469623572047,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8863292932510376,
      "num_tokens": 211346593.0,
      "step": 6075
    },
    {
      "epoch": 1.1283194057567316,
      "grad_norm": 1.5545787532322326,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8713591694831848,
      "num_tokens": 211380947.0,
      "step": 6076
    },
    {
      "epoch": 1.1285051067780874,
      "grad_norm": 1.5074035590236414,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8707317113876343,
      "num_tokens": 211420424.0,
      "step": 6077
    },
    {
      "epoch": 1.1286908077994429,
      "grad_norm": 1.659715755783203,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8786025047302246,
      "num_tokens": 211446282.0,
      "step": 6078
    },
    {
      "epoch": 1.1288765088207986,
      "grad_norm": 1.4352937974132536,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8642290830612183,
      "num_tokens": 211489393.0,
      "step": 6079
    },
    {
      "epoch": 1.129062209842154,
      "grad_norm": 1.5157972191479254,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8741236925125122,
      "num_tokens": 211522805.0,
      "step": 6080
    },
    {
      "epoch": 1.1292479108635098,
      "grad_norm": 1.686362922738272,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8645654320716858,
      "num_tokens": 211552276.0,
      "step": 6081
    },
    {
      "epoch": 1.1294336118848654,
      "grad_norm": 1.5693400836705522,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8826712369918823,
      "num_tokens": 211584166.0,
      "step": 6082
    },
    {
      "epoch": 1.1296193129062209,
      "grad_norm": 1.3592100773866258,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8940132856369019,
      "num_tokens": 211618826.0,
      "step": 6083
    },
    {
      "epoch": 1.1298050139275766,
      "grad_norm": 1.5270249702057184,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.869163990020752,
      "num_tokens": 211655917.0,
      "step": 6084
    },
    {
      "epoch": 1.1299907149489323,
      "grad_norm": 1.4186658875170692,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8732534050941467,
      "num_tokens": 211696962.0,
      "step": 6085
    },
    {
      "epoch": 1.1301764159702878,
      "grad_norm": 1.4540118623667078,
      "learning_rate": 1e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.86855149269104,
      "num_tokens": 211737979.0,
      "step": 6086
    },
    {
      "epoch": 1.1303621169916434,
      "grad_norm": 1.5791279009873538,
      "learning_rate": 1e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8722708225250244,
      "num_tokens": 211775180.0,
      "step": 6087
    },
    {
      "epoch": 1.130547818012999,
      "grad_norm": 1.5563875236167009,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8724502325057983,
      "num_tokens": 211807201.0,
      "step": 6088
    },
    {
      "epoch": 1.1307335190343546,
      "grad_norm": 1.5897553933777269,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8802608251571655,
      "num_tokens": 211837528.0,
      "step": 6089
    },
    {
      "epoch": 1.1309192200557103,
      "grad_norm": 1.5257649438872218,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.878666877746582,
      "num_tokens": 211871871.0,
      "step": 6090
    },
    {
      "epoch": 1.1311049210770658,
      "grad_norm": 1.574729901153273,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8688637614250183,
      "num_tokens": 211903204.0,
      "step": 6091
    },
    {
      "epoch": 1.1312906220984216,
      "grad_norm": 1.5012324388913945,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8765600919723511,
      "num_tokens": 211936835.0,
      "step": 6092
    },
    {
      "epoch": 1.131476323119777,
      "grad_norm": 1.518346750029448,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8701604604721069,
      "num_tokens": 211972601.0,
      "step": 6093
    },
    {
      "epoch": 1.1316620241411328,
      "grad_norm": 1.5060465993753687,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8849096298217773,
      "num_tokens": 212005334.0,
      "step": 6094
    },
    {
      "epoch": 1.1318477251624883,
      "grad_norm": 1.5915119622129064,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8749154806137085,
      "num_tokens": 212035165.0,
      "step": 6095
    },
    {
      "epoch": 1.132033426183844,
      "grad_norm": 1.5216921067440188,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8612954020500183,
      "num_tokens": 212072145.0,
      "step": 6096
    },
    {
      "epoch": 1.1322191272051996,
      "grad_norm": 1.6199818340215624,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8711937665939331,
      "num_tokens": 212105458.0,
      "step": 6097
    },
    {
      "epoch": 1.1324048282265553,
      "grad_norm": 1.5330588591581327,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8893284797668457,
      "num_tokens": 212135083.0,
      "step": 6098
    },
    {
      "epoch": 1.1325905292479108,
      "grad_norm": 1.4910804459353342,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8829840421676636,
      "num_tokens": 212169262.0,
      "step": 6099
    },
    {
      "epoch": 1.1327762302692665,
      "grad_norm": 1.5861431458290103,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.871512234210968,
      "num_tokens": 212203390.0,
      "step": 6100
    },
    {
      "epoch": 1.132961931290622,
      "grad_norm": 1.5443012191354561,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8819824457168579,
      "num_tokens": 212237711.0,
      "step": 6101
    },
    {
      "epoch": 1.1331476323119778,
      "grad_norm": 1.4741952845212842,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8766077756881714,
      "num_tokens": 212277335.0,
      "step": 6102
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 1.6792264470020848,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8812329769134521,
      "num_tokens": 212309287.0,
      "step": 6103
    },
    {
      "epoch": 1.133519034354689,
      "grad_norm": 1.5433578586343464,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8761271238327026,
      "num_tokens": 212342595.0,
      "step": 6104
    },
    {
      "epoch": 1.1337047353760445,
      "grad_norm": 1.584516452666333,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8840345740318298,
      "num_tokens": 212378710.0,
      "step": 6105
    },
    {
      "epoch": 1.1338904363974003,
      "grad_norm": 1.5893339113567144,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8693040013313293,
      "num_tokens": 212412326.0,
      "step": 6106
    },
    {
      "epoch": 1.1340761374187558,
      "grad_norm": 1.4827174563739989,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8872565031051636,
      "num_tokens": 212446045.0,
      "step": 6107
    },
    {
      "epoch": 1.1342618384401115,
      "grad_norm": 1.4942050538928942,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8751553297042847,
      "num_tokens": 212477984.0,
      "step": 6108
    },
    {
      "epoch": 1.134447539461467,
      "grad_norm": 1.5705559195038086,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8733320236206055,
      "num_tokens": 212509385.0,
      "step": 6109
    },
    {
      "epoch": 1.1346332404828225,
      "grad_norm": 1.6623223102754492,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8707432746887207,
      "num_tokens": 212538784.0,
      "step": 6110
    },
    {
      "epoch": 1.1348189415041783,
      "grad_norm": 1.5713132023218093,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8685038089752197,
      "num_tokens": 212577968.0,
      "step": 6111
    },
    {
      "epoch": 1.135004642525534,
      "grad_norm": 1.5831150205446742,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8767291307449341,
      "num_tokens": 212613498.0,
      "step": 6112
    },
    {
      "epoch": 1.1351903435468895,
      "grad_norm": 1.475493336189597,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8827844262123108,
      "num_tokens": 212648465.0,
      "step": 6113
    },
    {
      "epoch": 1.135376044568245,
      "grad_norm": 1.4096951224418788,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8856298923492432,
      "num_tokens": 212686092.0,
      "step": 6114
    },
    {
      "epoch": 1.1355617455896008,
      "grad_norm": 1.3786529002171677,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8876991271972656,
      "num_tokens": 212727457.0,
      "step": 6115
    },
    {
      "epoch": 1.1357474466109563,
      "grad_norm": 1.7566932929950323,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.873677670955658,
      "num_tokens": 212757992.0,
      "step": 6116
    },
    {
      "epoch": 1.135933147632312,
      "grad_norm": 1.6229518140257726,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8828699588775635,
      "num_tokens": 212785810.0,
      "step": 6117
    },
    {
      "epoch": 1.1361188486536675,
      "grad_norm": 1.4120582533870214,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8688099384307861,
      "num_tokens": 212827465.0,
      "step": 6118
    },
    {
      "epoch": 1.1363045496750233,
      "grad_norm": 1.5339198438412585,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8856762647628784,
      "num_tokens": 212858278.0,
      "step": 6119
    },
    {
      "epoch": 1.1364902506963788,
      "grad_norm": 1.533576267446573,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8884932398796082,
      "num_tokens": 212886329.0,
      "step": 6120
    },
    {
      "epoch": 1.1366759517177345,
      "grad_norm": 1.4483067653434627,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.881129264831543,
      "num_tokens": 212925289.0,
      "step": 6121
    },
    {
      "epoch": 1.13686165273909,
      "grad_norm": 1.5665743469930553,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8708974123001099,
      "num_tokens": 212958090.0,
      "step": 6122
    },
    {
      "epoch": 1.1370473537604457,
      "grad_norm": 1.2980189306837666,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8865944147109985,
      "num_tokens": 212999242.0,
      "step": 6123
    },
    {
      "epoch": 1.1372330547818013,
      "grad_norm": 1.401112792599496,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8785266876220703,
      "num_tokens": 213037213.0,
      "step": 6124
    },
    {
      "epoch": 1.137418755803157,
      "grad_norm": 1.6436471504790584,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8759862184524536,
      "num_tokens": 213068705.0,
      "step": 6125
    },
    {
      "epoch": 1.1376044568245125,
      "grad_norm": 1.5277714952409098,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8792415261268616,
      "num_tokens": 213102160.0,
      "step": 6126
    },
    {
      "epoch": 1.1377901578458682,
      "grad_norm": 1.5165588101406644,
      "learning_rate": 1e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8640901446342468,
      "num_tokens": 213137941.0,
      "step": 6127
    },
    {
      "epoch": 1.1379758588672237,
      "grad_norm": 1.349028173642914,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8852275609970093,
      "num_tokens": 213177669.0,
      "step": 6128
    },
    {
      "epoch": 1.1381615598885795,
      "grad_norm": 1.4873707377447036,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8766853213310242,
      "num_tokens": 213215837.0,
      "step": 6129
    },
    {
      "epoch": 1.138347260909935,
      "grad_norm": 1.4165580111069724,
      "learning_rate": 1e-06,
      "loss": 0.3246,
      "mean_token_accuracy": 0.8854432702064514,
      "num_tokens": 213255275.0,
      "step": 6130
    },
    {
      "epoch": 1.1385329619312907,
      "grad_norm": 1.423926737274186,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8812825679779053,
      "num_tokens": 213292355.0,
      "step": 6131
    },
    {
      "epoch": 1.1387186629526462,
      "grad_norm": 1.46194395571855,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8722748756408691,
      "num_tokens": 213329756.0,
      "step": 6132
    },
    {
      "epoch": 1.1389043639740017,
      "grad_norm": 1.4179032951441128,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.8809865117073059,
      "num_tokens": 213368228.0,
      "step": 6133
    },
    {
      "epoch": 1.1390900649953575,
      "grad_norm": 1.6710068068649926,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8725070953369141,
      "num_tokens": 213400546.0,
      "step": 6134
    },
    {
      "epoch": 1.1392757660167132,
      "grad_norm": 1.5498594071335035,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8763773441314697,
      "num_tokens": 213438765.0,
      "step": 6135
    },
    {
      "epoch": 1.1394614670380687,
      "grad_norm": 1.5990379315693337,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8753570318222046,
      "num_tokens": 213472174.0,
      "step": 6136
    },
    {
      "epoch": 1.1396471680594242,
      "grad_norm": 1.5956487790005156,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.871910572052002,
      "num_tokens": 213504249.0,
      "step": 6137
    },
    {
      "epoch": 1.13983286908078,
      "grad_norm": 1.5915296424156067,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8771576881408691,
      "num_tokens": 213539917.0,
      "step": 6138
    },
    {
      "epoch": 1.1400185701021355,
      "grad_norm": 1.4998665434898326,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8709178566932678,
      "num_tokens": 213577648.0,
      "step": 6139
    },
    {
      "epoch": 1.1402042711234912,
      "grad_norm": 1.4206034154758014,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.881891131401062,
      "num_tokens": 213616215.0,
      "step": 6140
    },
    {
      "epoch": 1.1403899721448467,
      "grad_norm": 1.576454300530814,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8727301359176636,
      "num_tokens": 213647487.0,
      "step": 6141
    },
    {
      "epoch": 1.1405756731662025,
      "grad_norm": 1.4809022738226163,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8759714365005493,
      "num_tokens": 213682045.0,
      "step": 6142
    },
    {
      "epoch": 1.140761374187558,
      "grad_norm": 1.5150324941481477,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8760901689529419,
      "num_tokens": 213721282.0,
      "step": 6143
    },
    {
      "epoch": 1.1409470752089137,
      "grad_norm": 1.3282062748374606,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8832252025604248,
      "num_tokens": 213765586.0,
      "step": 6144
    },
    {
      "epoch": 1.1411327762302692,
      "grad_norm": 1.569929472093239,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8898738622665405,
      "num_tokens": 213796117.0,
      "step": 6145
    },
    {
      "epoch": 1.141318477251625,
      "grad_norm": 1.4731142213997916,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.880927324295044,
      "num_tokens": 213833609.0,
      "step": 6146
    },
    {
      "epoch": 1.1415041782729805,
      "grad_norm": 1.432382031019945,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8837906122207642,
      "num_tokens": 213871420.0,
      "step": 6147
    },
    {
      "epoch": 1.1416898792943362,
      "grad_norm": 1.576464635974266,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8791934251785278,
      "num_tokens": 213906569.0,
      "step": 6148
    },
    {
      "epoch": 1.1418755803156917,
      "grad_norm": 1.4017366440909664,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8902738094329834,
      "num_tokens": 213942537.0,
      "step": 6149
    },
    {
      "epoch": 1.1420612813370474,
      "grad_norm": 1.518450015939752,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8843653202056885,
      "num_tokens": 213973071.0,
      "step": 6150
    },
    {
      "epoch": 1.142246982358403,
      "grad_norm": 1.50296468744069,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8643966317176819,
      "num_tokens": 214008609.0,
      "step": 6151
    },
    {
      "epoch": 1.1424326833797587,
      "grad_norm": 1.5204329244853343,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8706402778625488,
      "num_tokens": 214044470.0,
      "step": 6152
    },
    {
      "epoch": 1.1426183844011142,
      "grad_norm": 1.5156131184800667,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8748541474342346,
      "num_tokens": 214078280.0,
      "step": 6153
    },
    {
      "epoch": 1.14280408542247,
      "grad_norm": 1.55827281264963,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8819712996482849,
      "num_tokens": 214109576.0,
      "step": 6154
    },
    {
      "epoch": 1.1429897864438254,
      "grad_norm": 1.6628246933638762,
      "learning_rate": 1e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8627318739891052,
      "num_tokens": 214145759.0,
      "step": 6155
    },
    {
      "epoch": 1.143175487465181,
      "grad_norm": 1.3389684213081001,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8863329887390137,
      "num_tokens": 214184855.0,
      "step": 6156
    },
    {
      "epoch": 1.1433611884865367,
      "grad_norm": 1.6374486692429118,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8780666589736938,
      "num_tokens": 214213687.0,
      "step": 6157
    },
    {
      "epoch": 1.1435468895078924,
      "grad_norm": 1.5832574189923794,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8710503578186035,
      "num_tokens": 214247708.0,
      "step": 6158
    },
    {
      "epoch": 1.143732590529248,
      "grad_norm": 1.610717124697872,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8826676607131958,
      "num_tokens": 214278708.0,
      "step": 6159
    },
    {
      "epoch": 1.1439182915506034,
      "grad_norm": 1.499745574220487,
      "learning_rate": 1e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8552888631820679,
      "num_tokens": 214320100.0,
      "step": 6160
    },
    {
      "epoch": 1.1441039925719592,
      "grad_norm": 1.4696417748898156,
      "learning_rate": 1e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.859489917755127,
      "num_tokens": 214358027.0,
      "step": 6161
    },
    {
      "epoch": 1.1442896935933147,
      "grad_norm": 1.7097158987609062,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8667315244674683,
      "num_tokens": 214384291.0,
      "step": 6162
    },
    {
      "epoch": 1.1444753946146704,
      "grad_norm": 1.6699373678815939,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8713945150375366,
      "num_tokens": 214412193.0,
      "step": 6163
    },
    {
      "epoch": 1.144661095636026,
      "grad_norm": 1.3725689898099303,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8827325105667114,
      "num_tokens": 214448986.0,
      "step": 6164
    },
    {
      "epoch": 1.1448467966573816,
      "grad_norm": 1.469002378935148,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8767174482345581,
      "num_tokens": 214486045.0,
      "step": 6165
    },
    {
      "epoch": 1.1450324976787372,
      "grad_norm": 1.4421056274999007,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8756842017173767,
      "num_tokens": 214521535.0,
      "step": 6166
    },
    {
      "epoch": 1.145218198700093,
      "grad_norm": 1.4289168856252212,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8769775629043579,
      "num_tokens": 214560130.0,
      "step": 6167
    },
    {
      "epoch": 1.1454038997214484,
      "grad_norm": 1.4831469167554008,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8826728463172913,
      "num_tokens": 214595279.0,
      "step": 6168
    },
    {
      "epoch": 1.1455896007428041,
      "grad_norm": 1.5551241726498424,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8718037605285645,
      "num_tokens": 214627319.0,
      "step": 6169
    },
    {
      "epoch": 1.1457753017641596,
      "grad_norm": 1.470930590630006,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8880244493484497,
      "num_tokens": 214658043.0,
      "step": 6170
    },
    {
      "epoch": 1.1459610027855154,
      "grad_norm": 1.4522944068591765,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8722920417785645,
      "num_tokens": 214696352.0,
      "step": 6171
    },
    {
      "epoch": 1.146146703806871,
      "grad_norm": 1.3254785405175926,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8871829509735107,
      "num_tokens": 214738171.0,
      "step": 6172
    },
    {
      "epoch": 1.1463324048282266,
      "grad_norm": 1.6825856813315674,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8807663321495056,
      "num_tokens": 214767989.0,
      "step": 6173
    },
    {
      "epoch": 1.1465181058495821,
      "grad_norm": 1.7415343322044023,
      "learning_rate": 1e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.869297206401825,
      "num_tokens": 214795334.0,
      "step": 6174
    },
    {
      "epoch": 1.1467038068709379,
      "grad_norm": 1.6137635019885175,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8811118602752686,
      "num_tokens": 214830119.0,
      "step": 6175
    },
    {
      "epoch": 1.1468895078922934,
      "grad_norm": 1.6649276949947902,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8606129884719849,
      "num_tokens": 214864514.0,
      "step": 6176
    },
    {
      "epoch": 1.1470752089136491,
      "grad_norm": 1.4965146516665757,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8835231065750122,
      "num_tokens": 214898026.0,
      "step": 6177
    },
    {
      "epoch": 1.1472609099350046,
      "grad_norm": 1.5453611309056612,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8618850708007812,
      "num_tokens": 214935125.0,
      "step": 6178
    },
    {
      "epoch": 1.1474466109563601,
      "grad_norm": 1.5666778793004117,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8800263404846191,
      "num_tokens": 214968482.0,
      "step": 6179
    },
    {
      "epoch": 1.1476323119777159,
      "grad_norm": 1.6701037242886916,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8614789247512817,
      "num_tokens": 214999507.0,
      "step": 6180
    },
    {
      "epoch": 1.1478180129990716,
      "grad_norm": 1.423901904604476,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.878303050994873,
      "num_tokens": 215036835.0,
      "step": 6181
    },
    {
      "epoch": 1.1480037140204271,
      "grad_norm": 1.4594702025261985,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8807199001312256,
      "num_tokens": 215075022.0,
      "step": 6182
    },
    {
      "epoch": 1.1481894150417826,
      "grad_norm": 1.3870538196858084,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8762366771697998,
      "num_tokens": 215114265.0,
      "step": 6183
    },
    {
      "epoch": 1.1483751160631384,
      "grad_norm": 1.4526387946357264,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8874486088752747,
      "num_tokens": 215147753.0,
      "step": 6184
    },
    {
      "epoch": 1.1485608170844939,
      "grad_norm": 1.3876274441850283,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.886528491973877,
      "num_tokens": 215185567.0,
      "step": 6185
    },
    {
      "epoch": 1.1487465181058496,
      "grad_norm": 1.6369014954432073,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8695342540740967,
      "num_tokens": 215215596.0,
      "step": 6186
    },
    {
      "epoch": 1.1489322191272051,
      "grad_norm": 1.5141293008292405,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8647669553756714,
      "num_tokens": 215251510.0,
      "step": 6187
    },
    {
      "epoch": 1.1491179201485608,
      "grad_norm": 1.3616790308673483,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8803578615188599,
      "num_tokens": 215289187.0,
      "step": 6188
    },
    {
      "epoch": 1.1493036211699164,
      "grad_norm": 1.3766107834868864,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8867359757423401,
      "num_tokens": 215327864.0,
      "step": 6189
    },
    {
      "epoch": 1.149489322191272,
      "grad_norm": 1.4381343080481668,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8882436752319336,
      "num_tokens": 215363712.0,
      "step": 6190
    },
    {
      "epoch": 1.1496750232126276,
      "grad_norm": 1.5670313198327654,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8778846263885498,
      "num_tokens": 215395252.0,
      "step": 6191
    },
    {
      "epoch": 1.1498607242339833,
      "grad_norm": 1.7024026231381126,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8646393418312073,
      "num_tokens": 215426613.0,
      "step": 6192
    },
    {
      "epoch": 1.1500464252553388,
      "grad_norm": 1.4369000654978368,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8731542229652405,
      "num_tokens": 215463373.0,
      "step": 6193
    },
    {
      "epoch": 1.1502321262766946,
      "grad_norm": 1.2992845086820854,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8875508308410645,
      "num_tokens": 215503006.0,
      "step": 6194
    },
    {
      "epoch": 1.15041782729805,
      "grad_norm": 1.4497262165001814,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8886855840682983,
      "num_tokens": 215537658.0,
      "step": 6195
    },
    {
      "epoch": 1.1506035283194058,
      "grad_norm": 1.5871508232475209,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8790018558502197,
      "num_tokens": 215568683.0,
      "step": 6196
    },
    {
      "epoch": 1.1507892293407613,
      "grad_norm": 1.4702068966225588,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8715096712112427,
      "num_tokens": 215607297.0,
      "step": 6197
    },
    {
      "epoch": 1.150974930362117,
      "grad_norm": 1.4285968636944262,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.879282534122467,
      "num_tokens": 215642867.0,
      "step": 6198
    },
    {
      "epoch": 1.1511606313834726,
      "grad_norm": 1.5269530076589608,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8845954537391663,
      "num_tokens": 215678427.0,
      "step": 6199
    },
    {
      "epoch": 1.1513463324048283,
      "grad_norm": 1.4813236754733914,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8739910125732422,
      "num_tokens": 215713114.0,
      "step": 6200
    },
    {
      "epoch": 1.1515320334261838,
      "grad_norm": 1.510610401897965,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8902754783630371,
      "num_tokens": 215743717.0,
      "step": 6201
    },
    {
      "epoch": 1.1517177344475396,
      "grad_norm": 1.47200373587115,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.868154764175415,
      "num_tokens": 215784999.0,
      "step": 6202
    },
    {
      "epoch": 1.151903435468895,
      "grad_norm": 1.478446322886375,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8708362579345703,
      "num_tokens": 215821632.0,
      "step": 6203
    },
    {
      "epoch": 1.1520891364902508,
      "grad_norm": 1.7010659218367223,
      "learning_rate": 1e-06,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8742424845695496,
      "num_tokens": 215848860.0,
      "step": 6204
    },
    {
      "epoch": 1.1522748375116063,
      "grad_norm": 1.5814556141838145,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8717974424362183,
      "num_tokens": 215883568.0,
      "step": 6205
    },
    {
      "epoch": 1.1524605385329618,
      "grad_norm": 1.6222317522407237,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8847537040710449,
      "num_tokens": 215912339.0,
      "step": 6206
    },
    {
      "epoch": 1.1526462395543176,
      "grad_norm": 1.461957630576745,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8787828683853149,
      "num_tokens": 215947943.0,
      "step": 6207
    },
    {
      "epoch": 1.1528319405756733,
      "grad_norm": 1.4601335744338195,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8753629326820374,
      "num_tokens": 215982474.0,
      "step": 6208
    },
    {
      "epoch": 1.1530176415970288,
      "grad_norm": 1.4220682836272214,
      "learning_rate": 1e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8544923663139343,
      "num_tokens": 216024509.0,
      "step": 6209
    },
    {
      "epoch": 1.1532033426183843,
      "grad_norm": 1.3825527895316272,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8933801651000977,
      "num_tokens": 216061218.0,
      "step": 6210
    },
    {
      "epoch": 1.15338904363974,
      "grad_norm": 1.4505194461624233,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8730047941207886,
      "num_tokens": 216100630.0,
      "step": 6211
    },
    {
      "epoch": 1.1535747446610956,
      "grad_norm": 1.508115422323027,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8671858310699463,
      "num_tokens": 216136942.0,
      "step": 6212
    },
    {
      "epoch": 1.1537604456824513,
      "grad_norm": 1.4301229821798767,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8894174098968506,
      "num_tokens": 216171332.0,
      "step": 6213
    },
    {
      "epoch": 1.1539461467038068,
      "grad_norm": 1.300590220311354,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8895439505577087,
      "num_tokens": 216215569.0,
      "step": 6214
    },
    {
      "epoch": 1.1541318477251625,
      "grad_norm": 1.421778252135055,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8811262845993042,
      "num_tokens": 216251474.0,
      "step": 6215
    },
    {
      "epoch": 1.154317548746518,
      "grad_norm": 1.3971592624060072,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8662463426589966,
      "num_tokens": 216291846.0,
      "step": 6216
    },
    {
      "epoch": 1.1545032497678738,
      "grad_norm": 1.4832376148092008,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8836561441421509,
      "num_tokens": 216330659.0,
      "step": 6217
    },
    {
      "epoch": 1.1546889507892293,
      "grad_norm": 1.3881028913344782,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.886309802532196,
      "num_tokens": 216371516.0,
      "step": 6218
    },
    {
      "epoch": 1.154874651810585,
      "grad_norm": 1.5054320556985772,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8654529452323914,
      "num_tokens": 216409290.0,
      "step": 6219
    },
    {
      "epoch": 1.1550603528319405,
      "grad_norm": 1.4034017518689033,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8707813620567322,
      "num_tokens": 216449067.0,
      "step": 6220
    },
    {
      "epoch": 1.1552460538532963,
      "grad_norm": 1.551515771959678,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8748883008956909,
      "num_tokens": 216483178.0,
      "step": 6221
    },
    {
      "epoch": 1.1554317548746518,
      "grad_norm": 1.459800077006411,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8831274509429932,
      "num_tokens": 216517212.0,
      "step": 6222
    },
    {
      "epoch": 1.1556174558960075,
      "grad_norm": 1.5361635281180521,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8648472428321838,
      "num_tokens": 216554758.0,
      "step": 6223
    },
    {
      "epoch": 1.155803156917363,
      "grad_norm": 1.421493129967895,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.874302327632904,
      "num_tokens": 216595328.0,
      "step": 6224
    },
    {
      "epoch": 1.1559888579387188,
      "grad_norm": 1.4414474450418322,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8767815828323364,
      "num_tokens": 216631214.0,
      "step": 6225
    },
    {
      "epoch": 1.1561745589600743,
      "grad_norm": 1.6242738348944126,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8732228875160217,
      "num_tokens": 216666047.0,
      "step": 6226
    },
    {
      "epoch": 1.15636025998143,
      "grad_norm": 1.3932594575634127,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.893949568271637,
      "num_tokens": 216704268.0,
      "step": 6227
    },
    {
      "epoch": 1.1565459610027855,
      "grad_norm": 1.4978948080661387,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8879499435424805,
      "num_tokens": 216738251.0,
      "step": 6228
    },
    {
      "epoch": 1.156731662024141,
      "grad_norm": 1.4140989765531466,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8896777033805847,
      "num_tokens": 216774684.0,
      "step": 6229
    },
    {
      "epoch": 1.1569173630454967,
      "grad_norm": 1.586843975501399,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8704585433006287,
      "num_tokens": 216807030.0,
      "step": 6230
    },
    {
      "epoch": 1.1571030640668525,
      "grad_norm": 1.5787376892887295,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8790676593780518,
      "num_tokens": 216837525.0,
      "step": 6231
    },
    {
      "epoch": 1.157288765088208,
      "grad_norm": 1.3882992032140302,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8788022994995117,
      "num_tokens": 216875610.0,
      "step": 6232
    },
    {
      "epoch": 1.1574744661095635,
      "grad_norm": 1.5597726613651763,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8763149380683899,
      "num_tokens": 216908757.0,
      "step": 6233
    },
    {
      "epoch": 1.1576601671309192,
      "grad_norm": 1.6942205429630943,
      "learning_rate": 1e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.859154224395752,
      "num_tokens": 216945827.0,
      "step": 6234
    },
    {
      "epoch": 1.1578458681522747,
      "grad_norm": 1.4768565775204436,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8849384784698486,
      "num_tokens": 216979939.0,
      "step": 6235
    },
    {
      "epoch": 1.1580315691736305,
      "grad_norm": 1.3720469244823352,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8938555717468262,
      "num_tokens": 217015791.0,
      "step": 6236
    },
    {
      "epoch": 1.158217270194986,
      "grad_norm": 1.5672991239290646,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8878222703933716,
      "num_tokens": 217044774.0,
      "step": 6237
    },
    {
      "epoch": 1.1584029712163417,
      "grad_norm": 1.5314492360904761,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8833078145980835,
      "num_tokens": 217080428.0,
      "step": 6238
    },
    {
      "epoch": 1.1585886722376972,
      "grad_norm": 1.5534949853794497,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8790809512138367,
      "num_tokens": 217112305.0,
      "step": 6239
    },
    {
      "epoch": 1.158774373259053,
      "grad_norm": 1.6445527630409853,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.873259425163269,
      "num_tokens": 217141988.0,
      "step": 6240
    },
    {
      "epoch": 1.1589600742804085,
      "grad_norm": 1.415904795320033,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8813565969467163,
      "num_tokens": 217177190.0,
      "step": 6241
    },
    {
      "epoch": 1.1591457753017642,
      "grad_norm": 1.55789084247272,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8667645454406738,
      "num_tokens": 217211753.0,
      "step": 6242
    },
    {
      "epoch": 1.1593314763231197,
      "grad_norm": 1.389428221545594,
      "learning_rate": 1e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.8780562877655029,
      "num_tokens": 217252057.0,
      "step": 6243
    },
    {
      "epoch": 1.1595171773444755,
      "grad_norm": 1.5955333193908547,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8777042627334595,
      "num_tokens": 217284281.0,
      "step": 6244
    },
    {
      "epoch": 1.159702878365831,
      "grad_norm": 1.480888976072886,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8830997943878174,
      "num_tokens": 217320875.0,
      "step": 6245
    },
    {
      "epoch": 1.1598885793871867,
      "grad_norm": 1.4702994294817384,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.868059515953064,
      "num_tokens": 217358270.0,
      "step": 6246
    },
    {
      "epoch": 1.1600742804085422,
      "grad_norm": 1.5132567834102753,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8776260018348694,
      "num_tokens": 217392000.0,
      "step": 6247
    },
    {
      "epoch": 1.160259981429898,
      "grad_norm": 1.4601935959926358,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8683332204818726,
      "num_tokens": 217427758.0,
      "step": 6248
    },
    {
      "epoch": 1.1604456824512535,
      "grad_norm": 1.6315671800225957,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8718290328979492,
      "num_tokens": 217464947.0,
      "step": 6249
    },
    {
      "epoch": 1.1606313834726092,
      "grad_norm": 1.391847834190616,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8818399906158447,
      "num_tokens": 217505912.0,
      "step": 6250
    },
    {
      "epoch": 1.1608170844939647,
      "grad_norm": 1.4877882069387969,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8656085729598999,
      "num_tokens": 217544255.0,
      "step": 6251
    },
    {
      "epoch": 1.1610027855153202,
      "grad_norm": 1.4414369263217466,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8875675201416016,
      "num_tokens": 217581715.0,
      "step": 6252
    },
    {
      "epoch": 1.161188486536676,
      "grad_norm": 1.5498704353830781,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8594294786453247,
      "num_tokens": 217617639.0,
      "step": 6253
    },
    {
      "epoch": 1.1613741875580317,
      "grad_norm": 1.5120052716258707,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8716635704040527,
      "num_tokens": 217652684.0,
      "step": 6254
    },
    {
      "epoch": 1.1615598885793872,
      "grad_norm": 1.3910877865976659,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8772807121276855,
      "num_tokens": 217691144.0,
      "step": 6255
    },
    {
      "epoch": 1.1617455896007427,
      "grad_norm": 1.3814770004179917,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8850594758987427,
      "num_tokens": 217727645.0,
      "step": 6256
    },
    {
      "epoch": 1.1619312906220984,
      "grad_norm": 1.5308759098227882,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8713399767875671,
      "num_tokens": 217764331.0,
      "step": 6257
    },
    {
      "epoch": 1.162116991643454,
      "grad_norm": 1.4688641023174782,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8777257204055786,
      "num_tokens": 217799603.0,
      "step": 6258
    },
    {
      "epoch": 1.1623026926648097,
      "grad_norm": 1.4171240619785628,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.864436686038971,
      "num_tokens": 217838958.0,
      "step": 6259
    },
    {
      "epoch": 1.1624883936861652,
      "grad_norm": 1.4651692132146432,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.884943425655365,
      "num_tokens": 217874731.0,
      "step": 6260
    },
    {
      "epoch": 1.162674094707521,
      "grad_norm": 1.533978994766772,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8877615332603455,
      "num_tokens": 217903690.0,
      "step": 6261
    },
    {
      "epoch": 1.1628597957288764,
      "grad_norm": 1.350378859808284,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8664839267730713,
      "num_tokens": 217944663.0,
      "step": 6262
    },
    {
      "epoch": 1.1630454967502322,
      "grad_norm": 1.6379182657123224,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8714200258255005,
      "num_tokens": 217974446.0,
      "step": 6263
    },
    {
      "epoch": 1.1632311977715877,
      "grad_norm": 1.5128702335436752,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8733986616134644,
      "num_tokens": 218010481.0,
      "step": 6264
    },
    {
      "epoch": 1.1634168987929434,
      "grad_norm": 1.4608716759336975,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8662502765655518,
      "num_tokens": 218049527.0,
      "step": 6265
    },
    {
      "epoch": 1.163602599814299,
      "grad_norm": 1.360476636453076,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8880169987678528,
      "num_tokens": 218086132.0,
      "step": 6266
    },
    {
      "epoch": 1.1637883008356547,
      "grad_norm": 1.5829218432365846,
      "learning_rate": 1e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8629693984985352,
      "num_tokens": 218121332.0,
      "step": 6267
    },
    {
      "epoch": 1.1639740018570102,
      "grad_norm": 1.593203945421723,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8770527839660645,
      "num_tokens": 218154576.0,
      "step": 6268
    },
    {
      "epoch": 1.164159702878366,
      "grad_norm": 1.6180463461237682,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8764728307723999,
      "num_tokens": 218187974.0,
      "step": 6269
    },
    {
      "epoch": 1.1643454038997214,
      "grad_norm": 1.4624393759101102,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8657705783843994,
      "num_tokens": 218224432.0,
      "step": 6270
    },
    {
      "epoch": 1.1645311049210771,
      "grad_norm": 1.6495621043322901,
      "learning_rate": 1e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8490889072418213,
      "num_tokens": 218257721.0,
      "step": 6271
    },
    {
      "epoch": 1.1647168059424327,
      "grad_norm": 1.502227484591265,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8765881657600403,
      "num_tokens": 218295581.0,
      "step": 6272
    },
    {
      "epoch": 1.1649025069637884,
      "grad_norm": 1.5346425711715763,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8884280920028687,
      "num_tokens": 218326347.0,
      "step": 6273
    },
    {
      "epoch": 1.165088207985144,
      "grad_norm": 1.5897583235902784,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8728877305984497,
      "num_tokens": 218357553.0,
      "step": 6274
    },
    {
      "epoch": 1.1652739090064996,
      "grad_norm": 1.5128190983828773,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8853636980056763,
      "num_tokens": 218389174.0,
      "step": 6275
    },
    {
      "epoch": 1.1654596100278551,
      "grad_norm": 1.6218178613137022,
      "learning_rate": 1e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8594174385070801,
      "num_tokens": 218420768.0,
      "step": 6276
    },
    {
      "epoch": 1.1656453110492109,
      "grad_norm": 1.501767345128329,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8853867053985596,
      "num_tokens": 218456530.0,
      "step": 6277
    },
    {
      "epoch": 1.1658310120705664,
      "grad_norm": 1.52129666592478,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8719910383224487,
      "num_tokens": 218491072.0,
      "step": 6278
    },
    {
      "epoch": 1.166016713091922,
      "grad_norm": 1.4602699202704612,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.8831025958061218,
      "num_tokens": 218526755.0,
      "step": 6279
    },
    {
      "epoch": 1.1662024141132776,
      "grad_norm": 1.4299887936072688,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8753370642662048,
      "num_tokens": 218562889.0,
      "step": 6280
    },
    {
      "epoch": 1.1663881151346334,
      "grad_norm": 1.5492449391957555,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8647125959396362,
      "num_tokens": 218597886.0,
      "step": 6281
    },
    {
      "epoch": 1.1665738161559889,
      "grad_norm": 1.542438521615538,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8657394647598267,
      "num_tokens": 218633421.0,
      "step": 6282
    },
    {
      "epoch": 1.1667595171773444,
      "grad_norm": 1.5135878001667828,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8734187483787537,
      "num_tokens": 218666657.0,
      "step": 6283
    },
    {
      "epoch": 1.1669452181987001,
      "grad_norm": 1.7393721561039912,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8672827482223511,
      "num_tokens": 218693710.0,
      "step": 6284
    },
    {
      "epoch": 1.1671309192200556,
      "grad_norm": 1.5475280879531008,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8721601963043213,
      "num_tokens": 218728309.0,
      "step": 6285
    },
    {
      "epoch": 1.1673166202414114,
      "grad_norm": 1.8198237694161117,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8653804063796997,
      "num_tokens": 218758025.0,
      "step": 6286
    },
    {
      "epoch": 1.1675023212627669,
      "grad_norm": 1.5610056891104291,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8781194686889648,
      "num_tokens": 218791881.0,
      "step": 6287
    },
    {
      "epoch": 1.1676880222841226,
      "grad_norm": 1.8256651155539114,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8618896007537842,
      "num_tokens": 218820725.0,
      "step": 6288
    },
    {
      "epoch": 1.1678737233054781,
      "grad_norm": 1.3472694741554117,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8785451650619507,
      "num_tokens": 218863443.0,
      "step": 6289
    },
    {
      "epoch": 1.1680594243268339,
      "grad_norm": 1.3818590217303692,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8952435851097107,
      "num_tokens": 218903596.0,
      "step": 6290
    },
    {
      "epoch": 1.1682451253481894,
      "grad_norm": 1.4580829409369338,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8704918622970581,
      "num_tokens": 218947218.0,
      "step": 6291
    },
    {
      "epoch": 1.168430826369545,
      "grad_norm": 1.6276700926768148,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8720958232879639,
      "num_tokens": 218979653.0,
      "step": 6292
    },
    {
      "epoch": 1.1686165273909006,
      "grad_norm": 1.4871370229112482,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8780087232589722,
      "num_tokens": 219016393.0,
      "step": 6293
    },
    {
      "epoch": 1.1688022284122563,
      "grad_norm": 1.4624259715784704,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8826828002929688,
      "num_tokens": 219052009.0,
      "step": 6294
    },
    {
      "epoch": 1.1689879294336118,
      "grad_norm": 1.4299154573487876,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8833873271942139,
      "num_tokens": 219089724.0,
      "step": 6295
    },
    {
      "epoch": 1.1691736304549676,
      "grad_norm": 1.4777795212697054,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8726285696029663,
      "num_tokens": 219127456.0,
      "step": 6296
    },
    {
      "epoch": 1.169359331476323,
      "grad_norm": 1.415179583102484,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8735772371292114,
      "num_tokens": 219163516.0,
      "step": 6297
    },
    {
      "epoch": 1.1695450324976788,
      "grad_norm": 1.5437626428202595,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8808426260948181,
      "num_tokens": 219197164.0,
      "step": 6298
    },
    {
      "epoch": 1.1697307335190343,
      "grad_norm": 1.5983349713648165,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8880240321159363,
      "num_tokens": 219227042.0,
      "step": 6299
    },
    {
      "epoch": 1.16991643454039,
      "grad_norm": 1.685295392923547,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.884436845779419,
      "num_tokens": 219254519.0,
      "step": 6300
    },
    {
      "epoch": 1.1701021355617456,
      "grad_norm": 1.5687559028402407,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8836921453475952,
      "num_tokens": 219286661.0,
      "step": 6301
    },
    {
      "epoch": 1.170287836583101,
      "grad_norm": 1.414139113467143,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.878904402256012,
      "num_tokens": 219328673.0,
      "step": 6302
    },
    {
      "epoch": 1.1704735376044568,
      "grad_norm": 1.8196366639517698,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8673344254493713,
      "num_tokens": 219363948.0,
      "step": 6303
    },
    {
      "epoch": 1.1706592386258126,
      "grad_norm": 1.5748886469968517,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8782347440719604,
      "num_tokens": 219398157.0,
      "step": 6304
    },
    {
      "epoch": 1.170844939647168,
      "grad_norm": 1.6464435800396584,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8799936175346375,
      "num_tokens": 219429028.0,
      "step": 6305
    },
    {
      "epoch": 1.1710306406685236,
      "grad_norm": 1.4541212011091063,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8684743046760559,
      "num_tokens": 219468861.0,
      "step": 6306
    },
    {
      "epoch": 1.1712163416898793,
      "grad_norm": 1.6037790070936708,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8725407719612122,
      "num_tokens": 219501239.0,
      "step": 6307
    },
    {
      "epoch": 1.1714020427112348,
      "grad_norm": 1.369250253129838,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.886974573135376,
      "num_tokens": 219537281.0,
      "step": 6308
    },
    {
      "epoch": 1.1715877437325906,
      "grad_norm": 1.5762531332264458,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8745301961898804,
      "num_tokens": 219569405.0,
      "step": 6309
    },
    {
      "epoch": 1.171773444753946,
      "grad_norm": 1.6298876967182794,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8743395805358887,
      "num_tokens": 219604556.0,
      "step": 6310
    },
    {
      "epoch": 1.1719591457753018,
      "grad_norm": 1.5418166615926039,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8730263113975525,
      "num_tokens": 219639187.0,
      "step": 6311
    },
    {
      "epoch": 1.1721448467966573,
      "grad_norm": 1.5014939248178236,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8634933233261108,
      "num_tokens": 219676696.0,
      "step": 6312
    },
    {
      "epoch": 1.172330547818013,
      "grad_norm": 1.5559724174345007,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8822685480117798,
      "num_tokens": 219710158.0,
      "step": 6313
    },
    {
      "epoch": 1.1725162488393686,
      "grad_norm": 1.6551835222892075,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8805993795394897,
      "num_tokens": 219739920.0,
      "step": 6314
    },
    {
      "epoch": 1.1727019498607243,
      "grad_norm": 1.4438617869119157,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8779996633529663,
      "num_tokens": 219776309.0,
      "step": 6315
    },
    {
      "epoch": 1.1728876508820798,
      "grad_norm": 1.3260995048414632,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8744854927062988,
      "num_tokens": 219819740.0,
      "step": 6316
    },
    {
      "epoch": 1.1730733519034355,
      "grad_norm": 1.4856137433459196,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8755521178245544,
      "num_tokens": 219855063.0,
      "step": 6317
    },
    {
      "epoch": 1.173259052924791,
      "grad_norm": 1.4311882083877316,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8744977712631226,
      "num_tokens": 219893079.0,
      "step": 6318
    },
    {
      "epoch": 1.1734447539461468,
      "grad_norm": 1.3879381389906627,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8704727292060852,
      "num_tokens": 219934186.0,
      "step": 6319
    },
    {
      "epoch": 1.1736304549675023,
      "grad_norm": 1.4102601196145135,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8864134550094604,
      "num_tokens": 219972238.0,
      "step": 6320
    },
    {
      "epoch": 1.173816155988858,
      "grad_norm": 1.6028978974297798,
      "learning_rate": 1e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8569820523262024,
      "num_tokens": 220003919.0,
      "step": 6321
    },
    {
      "epoch": 1.1740018570102135,
      "grad_norm": 1.5260764618743974,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8925981521606445,
      "num_tokens": 220036697.0,
      "step": 6322
    },
    {
      "epoch": 1.1741875580315693,
      "grad_norm": 1.5235149122616278,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.861328661441803,
      "num_tokens": 220071483.0,
      "step": 6323
    },
    {
      "epoch": 1.1743732590529248,
      "grad_norm": 1.4297118947407408,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8807364702224731,
      "num_tokens": 220108029.0,
      "step": 6324
    },
    {
      "epoch": 1.1745589600742803,
      "grad_norm": 1.4325182290086893,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.879928708076477,
      "num_tokens": 220143614.0,
      "step": 6325
    },
    {
      "epoch": 1.174744661095636,
      "grad_norm": 1.4463620904620431,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8670675158500671,
      "num_tokens": 220181422.0,
      "step": 6326
    },
    {
      "epoch": 1.1749303621169918,
      "grad_norm": 1.562812306488419,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8817057609558105,
      "num_tokens": 220210298.0,
      "step": 6327
    },
    {
      "epoch": 1.1751160631383473,
      "grad_norm": 1.3642480481851138,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8841696977615356,
      "num_tokens": 220253319.0,
      "step": 6328
    },
    {
      "epoch": 1.1753017641597028,
      "grad_norm": 1.5128739481493865,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8829762935638428,
      "num_tokens": 220283679.0,
      "step": 6329
    },
    {
      "epoch": 1.1754874651810585,
      "grad_norm": 1.577158546327914,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.858859658241272,
      "num_tokens": 220317099.0,
      "step": 6330
    },
    {
      "epoch": 1.175673166202414,
      "grad_norm": 1.5772406817171296,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8706613183021545,
      "num_tokens": 220349519.0,
      "step": 6331
    },
    {
      "epoch": 1.1758588672237698,
      "grad_norm": 1.5203353253325351,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.872466504573822,
      "num_tokens": 220387105.0,
      "step": 6332
    },
    {
      "epoch": 1.1760445682451253,
      "grad_norm": 1.5400372637980289,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8787177801132202,
      "num_tokens": 220420690.0,
      "step": 6333
    },
    {
      "epoch": 1.176230269266481,
      "grad_norm": 1.6385681835638233,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8947986364364624,
      "num_tokens": 220448247.0,
      "step": 6334
    },
    {
      "epoch": 1.1764159702878365,
      "grad_norm": 1.431839947465282,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.874330997467041,
      "num_tokens": 220485152.0,
      "step": 6335
    },
    {
      "epoch": 1.1766016713091922,
      "grad_norm": 1.6605292411864814,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8854702115058899,
      "num_tokens": 220515144.0,
      "step": 6336
    },
    {
      "epoch": 1.1767873723305478,
      "grad_norm": 1.5431040856772202,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.878720223903656,
      "num_tokens": 220547659.0,
      "step": 6337
    },
    {
      "epoch": 1.1769730733519035,
      "grad_norm": 1.4336293017641746,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8906356692314148,
      "num_tokens": 220583162.0,
      "step": 6338
    },
    {
      "epoch": 1.177158774373259,
      "grad_norm": 1.6820433104285828,
      "learning_rate": 1e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8570660352706909,
      "num_tokens": 220616949.0,
      "step": 6339
    },
    {
      "epoch": 1.1773444753946147,
      "grad_norm": 1.4457163522554954,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8879196047782898,
      "num_tokens": 220650687.0,
      "step": 6340
    },
    {
      "epoch": 1.1775301764159702,
      "grad_norm": 1.406653225918987,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.877045750617981,
      "num_tokens": 220689875.0,
      "step": 6341
    },
    {
      "epoch": 1.177715877437326,
      "grad_norm": 1.4218380700659479,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8878117799758911,
      "num_tokens": 220727709.0,
      "step": 6342
    },
    {
      "epoch": 1.1779015784586815,
      "grad_norm": 1.4552841683400721,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8774936199188232,
      "num_tokens": 220761361.0,
      "step": 6343
    },
    {
      "epoch": 1.1780872794800372,
      "grad_norm": 1.4409757980685398,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.877951979637146,
      "num_tokens": 220799778.0,
      "step": 6344
    },
    {
      "epoch": 1.1782729805013927,
      "grad_norm": 1.511150033336397,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8717328310012817,
      "num_tokens": 220833459.0,
      "step": 6345
    },
    {
      "epoch": 1.1784586815227485,
      "grad_norm": 1.5193728679871,
      "learning_rate": 1e-06,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8555448651313782,
      "num_tokens": 220870828.0,
      "step": 6346
    },
    {
      "epoch": 1.178644382544104,
      "grad_norm": 1.6147723624596968,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8872480988502502,
      "num_tokens": 220899558.0,
      "step": 6347
    },
    {
      "epoch": 1.1788300835654595,
      "grad_norm": 1.574800461954438,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8655053377151489,
      "num_tokens": 220932614.0,
      "step": 6348
    },
    {
      "epoch": 1.1790157845868152,
      "grad_norm": 1.518580009409418,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8767567873001099,
      "num_tokens": 220967078.0,
      "step": 6349
    },
    {
      "epoch": 1.179201485608171,
      "grad_norm": 1.4580110108243138,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8771138191223145,
      "num_tokens": 221006181.0,
      "step": 6350
    },
    {
      "epoch": 1.1793871866295265,
      "grad_norm": 1.5518022942099634,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8689777851104736,
      "num_tokens": 221040892.0,
      "step": 6351
    },
    {
      "epoch": 1.179572887650882,
      "grad_norm": 1.5319289227571475,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8787516355514526,
      "num_tokens": 221075446.0,
      "step": 6352
    },
    {
      "epoch": 1.1797585886722377,
      "grad_norm": 1.584337935311231,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8790172934532166,
      "num_tokens": 221107361.0,
      "step": 6353
    },
    {
      "epoch": 1.1799442896935932,
      "grad_norm": 1.628729857033381,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8849538564682007,
      "num_tokens": 221142774.0,
      "step": 6354
    },
    {
      "epoch": 1.180129990714949,
      "grad_norm": 1.4503173290258946,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8795474767684937,
      "num_tokens": 221180452.0,
      "step": 6355
    },
    {
      "epoch": 1.1803156917363045,
      "grad_norm": 1.483748357449128,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8716104030609131,
      "num_tokens": 221221483.0,
      "step": 6356
    },
    {
      "epoch": 1.1805013927576602,
      "grad_norm": 1.424877235532896,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8737737536430359,
      "num_tokens": 221259981.0,
      "step": 6357
    },
    {
      "epoch": 1.1806870937790157,
      "grad_norm": 1.5562992962718054,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.875895619392395,
      "num_tokens": 221294398.0,
      "step": 6358
    },
    {
      "epoch": 1.1808727948003714,
      "grad_norm": 1.456645849647264,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8747602105140686,
      "num_tokens": 221332752.0,
      "step": 6359
    },
    {
      "epoch": 1.181058495821727,
      "grad_norm": 1.5441643388493995,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8722733855247498,
      "num_tokens": 221366111.0,
      "step": 6360
    },
    {
      "epoch": 1.1812441968430827,
      "grad_norm": 1.5346432595888055,
      "learning_rate": 1e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8673485517501831,
      "num_tokens": 221402282.0,
      "step": 6361
    },
    {
      "epoch": 1.1814298978644382,
      "grad_norm": 1.3317241476305577,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8890963792800903,
      "num_tokens": 221441265.0,
      "step": 6362
    },
    {
      "epoch": 1.181615598885794,
      "grad_norm": 1.3718314055987197,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8801790475845337,
      "num_tokens": 221478730.0,
      "step": 6363
    },
    {
      "epoch": 1.1818012999071494,
      "grad_norm": 1.6647146384102531,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8827208280563354,
      "num_tokens": 221506565.0,
      "step": 6364
    },
    {
      "epoch": 1.1819870009285052,
      "grad_norm": 1.66641379523104,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8674592971801758,
      "num_tokens": 221539599.0,
      "step": 6365
    },
    {
      "epoch": 1.1821727019498607,
      "grad_norm": 1.674310937966129,
      "learning_rate": 1e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8618737459182739,
      "num_tokens": 221572208.0,
      "step": 6366
    },
    {
      "epoch": 1.1823584029712164,
      "grad_norm": 1.5321052350752409,
      "learning_rate": 1e-06,
      "loss": 0.3222,
      "mean_token_accuracy": 0.8869778513908386,
      "num_tokens": 221602154.0,
      "step": 6367
    },
    {
      "epoch": 1.182544103992572,
      "grad_norm": 1.5726370227857174,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8636045455932617,
      "num_tokens": 221637147.0,
      "step": 6368
    },
    {
      "epoch": 1.1827298050139277,
      "grad_norm": 1.5099136801500765,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8828141689300537,
      "num_tokens": 221670812.0,
      "step": 6369
    },
    {
      "epoch": 1.1829155060352832,
      "grad_norm": 1.8318050699449473,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8663996458053589,
      "num_tokens": 221700164.0,
      "step": 6370
    },
    {
      "epoch": 1.183101207056639,
      "grad_norm": 1.5479005787312523,
      "learning_rate": 1e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.864581823348999,
      "num_tokens": 221732014.0,
      "step": 6371
    },
    {
      "epoch": 1.1832869080779944,
      "grad_norm": 1.5619115825763417,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8807423710823059,
      "num_tokens": 221761689.0,
      "step": 6372
    },
    {
      "epoch": 1.1834726090993501,
      "grad_norm": 1.4842301388774097,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8756062984466553,
      "num_tokens": 221796560.0,
      "step": 6373
    },
    {
      "epoch": 1.1836583101207057,
      "grad_norm": 1.54177372293004,
      "learning_rate": 1e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8660790324211121,
      "num_tokens": 221835810.0,
      "step": 6374
    },
    {
      "epoch": 1.1838440111420612,
      "grad_norm": 1.4632559823396771,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8573265671730042,
      "num_tokens": 221878148.0,
      "step": 6375
    },
    {
      "epoch": 1.184029712163417,
      "grad_norm": 1.58816734645822,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8829353451728821,
      "num_tokens": 221911582.0,
      "step": 6376
    },
    {
      "epoch": 1.1842154131847726,
      "grad_norm": 1.614534438110549,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8756452798843384,
      "num_tokens": 221947969.0,
      "step": 6377
    },
    {
      "epoch": 1.1844011142061281,
      "grad_norm": 1.459549753202955,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8846290111541748,
      "num_tokens": 221983426.0,
      "step": 6378
    },
    {
      "epoch": 1.1845868152274837,
      "grad_norm": 1.4864591730721095,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.893890380859375,
      "num_tokens": 222016794.0,
      "step": 6379
    },
    {
      "epoch": 1.1847725162488394,
      "grad_norm": 1.54442455117147,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8841756582260132,
      "num_tokens": 222048128.0,
      "step": 6380
    },
    {
      "epoch": 1.184958217270195,
      "grad_norm": 1.6195014647963362,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8783649206161499,
      "num_tokens": 222078397.0,
      "step": 6381
    },
    {
      "epoch": 1.1851439182915506,
      "grad_norm": 1.817311771179718,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8761987686157227,
      "num_tokens": 222105715.0,
      "step": 6382
    },
    {
      "epoch": 1.1853296193129061,
      "grad_norm": 1.7948333409204031,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8755230903625488,
      "num_tokens": 222132980.0,
      "step": 6383
    },
    {
      "epoch": 1.1855153203342619,
      "grad_norm": 1.5928036757905515,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8714341521263123,
      "num_tokens": 222165865.0,
      "step": 6384
    },
    {
      "epoch": 1.1857010213556174,
      "grad_norm": 1.4582449392527892,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8792985081672668,
      "num_tokens": 222201040.0,
      "step": 6385
    },
    {
      "epoch": 1.1858867223769731,
      "grad_norm": 1.5534661108529164,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8692373037338257,
      "num_tokens": 222235802.0,
      "step": 6386
    },
    {
      "epoch": 1.1860724233983286,
      "grad_norm": 1.4906802369090573,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.877132773399353,
      "num_tokens": 222268991.0,
      "step": 6387
    },
    {
      "epoch": 1.1862581244196844,
      "grad_norm": 1.567636925385504,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8713564872741699,
      "num_tokens": 222304559.0,
      "step": 6388
    },
    {
      "epoch": 1.1864438254410399,
      "grad_norm": 1.5449139848810132,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8678655028343201,
      "num_tokens": 222340017.0,
      "step": 6389
    },
    {
      "epoch": 1.1866295264623956,
      "grad_norm": 1.6245722827544342,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8689830303192139,
      "num_tokens": 222370124.0,
      "step": 6390
    },
    {
      "epoch": 1.1868152274837511,
      "grad_norm": 1.4521345497483749,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.886850893497467,
      "num_tokens": 222406149.0,
      "step": 6391
    },
    {
      "epoch": 1.1870009285051069,
      "grad_norm": 1.49219157427084,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8679167032241821,
      "num_tokens": 222444956.0,
      "step": 6392
    },
    {
      "epoch": 1.1871866295264624,
      "grad_norm": 1.6489554238063504,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8706549406051636,
      "num_tokens": 222473965.0,
      "step": 6393
    },
    {
      "epoch": 1.187372330547818,
      "grad_norm": 1.398467058038134,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.881519615650177,
      "num_tokens": 222510053.0,
      "step": 6394
    },
    {
      "epoch": 1.1875580315691736,
      "grad_norm": 1.5898294317789254,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8612768650054932,
      "num_tokens": 222544948.0,
      "step": 6395
    },
    {
      "epoch": 1.1877437325905293,
      "grad_norm": 1.6746619472074706,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.887527585029602,
      "num_tokens": 222571932.0,
      "step": 6396
    },
    {
      "epoch": 1.1879294336118849,
      "grad_norm": 1.39493761250469,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8753662109375,
      "num_tokens": 222613029.0,
      "step": 6397
    },
    {
      "epoch": 1.1881151346332404,
      "grad_norm": 1.4861396131397908,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8754593133926392,
      "num_tokens": 222647040.0,
      "step": 6398
    },
    {
      "epoch": 1.188300835654596,
      "grad_norm": 1.4649757406033048,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8693224787712097,
      "num_tokens": 222686738.0,
      "step": 6399
    },
    {
      "epoch": 1.1884865366759518,
      "grad_norm": 1.749600611245533,
      "learning_rate": 1e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8634912967681885,
      "num_tokens": 222714301.0,
      "step": 6400
    },
    {
      "epoch": 1.1886722376973073,
      "grad_norm": 1.550359167963346,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8821172714233398,
      "num_tokens": 222749670.0,
      "step": 6401
    },
    {
      "epoch": 1.1888579387186629,
      "grad_norm": 1.481957782916023,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8786681890487671,
      "num_tokens": 222783134.0,
      "step": 6402
    },
    {
      "epoch": 1.1890436397400186,
      "grad_norm": 1.5888912542137614,
      "learning_rate": 1e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8614501357078552,
      "num_tokens": 222819719.0,
      "step": 6403
    },
    {
      "epoch": 1.189229340761374,
      "grad_norm": 1.509745681359789,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8771282434463501,
      "num_tokens": 222856140.0,
      "step": 6404
    },
    {
      "epoch": 1.1894150417827298,
      "grad_norm": 1.5257826224364768,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8877990245819092,
      "num_tokens": 222887254.0,
      "step": 6405
    },
    {
      "epoch": 1.1896007428040853,
      "grad_norm": 1.556093936054649,
      "learning_rate": 1e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8674670457839966,
      "num_tokens": 222921121.0,
      "step": 6406
    },
    {
      "epoch": 1.189786443825441,
      "grad_norm": 1.4108666753666466,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8824335932731628,
      "num_tokens": 222955416.0,
      "step": 6407
    },
    {
      "epoch": 1.1899721448467966,
      "grad_norm": 1.3579939277272417,
      "learning_rate": 1e-06,
      "loss": 0.2895,
      "mean_token_accuracy": 0.8987478613853455,
      "num_tokens": 222991635.0,
      "step": 6408
    },
    {
      "epoch": 1.1901578458681523,
      "grad_norm": 1.4330381525302585,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8813265562057495,
      "num_tokens": 223028707.0,
      "step": 6409
    },
    {
      "epoch": 1.1903435468895078,
      "grad_norm": 1.4413129816921353,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8708783388137817,
      "num_tokens": 223066375.0,
      "step": 6410
    },
    {
      "epoch": 1.1905292479108636,
      "grad_norm": 1.4537518738201032,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8873181343078613,
      "num_tokens": 223100915.0,
      "step": 6411
    },
    {
      "epoch": 1.190714948932219,
      "grad_norm": 1.3747297291837821,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8779162764549255,
      "num_tokens": 223141581.0,
      "step": 6412
    },
    {
      "epoch": 1.1909006499535748,
      "grad_norm": 1.4714468089097397,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8838196992874146,
      "num_tokens": 223176464.0,
      "step": 6413
    },
    {
      "epoch": 1.1910863509749303,
      "grad_norm": 1.494324480395904,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.883050262928009,
      "num_tokens": 223210919.0,
      "step": 6414
    },
    {
      "epoch": 1.191272051996286,
      "grad_norm": 1.5725690363191314,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8898857831954956,
      "num_tokens": 223244568.0,
      "step": 6415
    },
    {
      "epoch": 1.1914577530176416,
      "grad_norm": 1.4200703025265053,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8733378648757935,
      "num_tokens": 223284286.0,
      "step": 6416
    },
    {
      "epoch": 1.1916434540389973,
      "grad_norm": 1.516092162939947,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8790880441665649,
      "num_tokens": 223322187.0,
      "step": 6417
    },
    {
      "epoch": 1.1918291550603528,
      "grad_norm": 1.6651499068644482,
      "learning_rate": 1e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8591853380203247,
      "num_tokens": 223352028.0,
      "step": 6418
    },
    {
      "epoch": 1.1920148560817085,
      "grad_norm": 1.3724806756993486,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8804380893707275,
      "num_tokens": 223396175.0,
      "step": 6419
    },
    {
      "epoch": 1.192200557103064,
      "grad_norm": 1.5272857874438448,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8712577819824219,
      "num_tokens": 223429234.0,
      "step": 6420
    },
    {
      "epoch": 1.1923862581244196,
      "grad_norm": 1.6329000492782997,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8811109066009521,
      "num_tokens": 223459317.0,
      "step": 6421
    },
    {
      "epoch": 1.1925719591457753,
      "grad_norm": 1.5926466208289478,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8754183650016785,
      "num_tokens": 223493739.0,
      "step": 6422
    },
    {
      "epoch": 1.192757660167131,
      "grad_norm": 1.677708817054742,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8610934615135193,
      "num_tokens": 223524609.0,
      "step": 6423
    },
    {
      "epoch": 1.1929433611884865,
      "grad_norm": 1.4551039024448646,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8784496784210205,
      "num_tokens": 223562735.0,
      "step": 6424
    },
    {
      "epoch": 1.193129062209842,
      "grad_norm": 1.4856639794741364,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8922914862632751,
      "num_tokens": 223594231.0,
      "step": 6425
    },
    {
      "epoch": 1.1933147632311978,
      "grad_norm": 1.5178345817508192,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8662666082382202,
      "num_tokens": 223629900.0,
      "step": 6426
    },
    {
      "epoch": 1.1935004642525533,
      "grad_norm": 1.5922157023994792,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8817166090011597,
      "num_tokens": 223661074.0,
      "step": 6427
    },
    {
      "epoch": 1.193686165273909,
      "grad_norm": 1.5687209252568393,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8573737740516663,
      "num_tokens": 223693802.0,
      "step": 6428
    },
    {
      "epoch": 1.1938718662952645,
      "grad_norm": 1.5772336917241434,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8683783411979675,
      "num_tokens": 223725398.0,
      "step": 6429
    },
    {
      "epoch": 1.1940575673166203,
      "grad_norm": 1.644131000665635,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8694944977760315,
      "num_tokens": 223756103.0,
      "step": 6430
    },
    {
      "epoch": 1.1942432683379758,
      "grad_norm": 1.4442555923716984,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8717224597930908,
      "num_tokens": 223793661.0,
      "step": 6431
    },
    {
      "epoch": 1.1944289693593315,
      "grad_norm": 1.6070675537793493,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8786808252334595,
      "num_tokens": 223825599.0,
      "step": 6432
    },
    {
      "epoch": 1.194614670380687,
      "grad_norm": 1.3353570422518644,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8861790895462036,
      "num_tokens": 223867077.0,
      "step": 6433
    },
    {
      "epoch": 1.1948003714020428,
      "grad_norm": 1.5273559678487303,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8747601509094238,
      "num_tokens": 223901157.0,
      "step": 6434
    },
    {
      "epoch": 1.1949860724233983,
      "grad_norm": 1.4279177304880049,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8770849108695984,
      "num_tokens": 223937616.0,
      "step": 6435
    },
    {
      "epoch": 1.195171773444754,
      "grad_norm": 1.452137276993966,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8875380754470825,
      "num_tokens": 223970902.0,
      "step": 6436
    },
    {
      "epoch": 1.1953574744661095,
      "grad_norm": 1.6143364498603356,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8732521533966064,
      "num_tokens": 223999814.0,
      "step": 6437
    },
    {
      "epoch": 1.1955431754874652,
      "grad_norm": 1.5869291392737346,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8670641183853149,
      "num_tokens": 224031328.0,
      "step": 6438
    },
    {
      "epoch": 1.1957288765088208,
      "grad_norm": 1.5508148148202907,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8776991367340088,
      "num_tokens": 224065556.0,
      "step": 6439
    },
    {
      "epoch": 1.1959145775301765,
      "grad_norm": 1.4605611633771258,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8878498077392578,
      "num_tokens": 224101544.0,
      "step": 6440
    },
    {
      "epoch": 1.196100278551532,
      "grad_norm": 1.5183498434461338,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.889412522315979,
      "num_tokens": 224132692.0,
      "step": 6441
    },
    {
      "epoch": 1.1962859795728877,
      "grad_norm": 1.6903111461038443,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.856569230556488,
      "num_tokens": 224163108.0,
      "step": 6442
    },
    {
      "epoch": 1.1964716805942432,
      "grad_norm": 1.4509595389131456,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8820487856864929,
      "num_tokens": 224201727.0,
      "step": 6443
    },
    {
      "epoch": 1.196657381615599,
      "grad_norm": 1.4510466563054434,
      "learning_rate": 1e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8594468235969543,
      "num_tokens": 224244620.0,
      "step": 6444
    },
    {
      "epoch": 1.1968430826369545,
      "grad_norm": 1.579046195590986,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8668431043624878,
      "num_tokens": 224281587.0,
      "step": 6445
    },
    {
      "epoch": 1.1970287836583102,
      "grad_norm": 1.5325159011914085,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8841432332992554,
      "num_tokens": 224316707.0,
      "step": 6446
    },
    {
      "epoch": 1.1972144846796657,
      "grad_norm": 1.4959364286716972,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8658654689788818,
      "num_tokens": 224355022.0,
      "step": 6447
    },
    {
      "epoch": 1.1974001857010212,
      "grad_norm": 1.6626735818934675,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8696812987327576,
      "num_tokens": 224384832.0,
      "step": 6448
    },
    {
      "epoch": 1.197585886722377,
      "grad_norm": 1.2768383955740714,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8857995271682739,
      "num_tokens": 224430857.0,
      "step": 6449
    },
    {
      "epoch": 1.1977715877437327,
      "grad_norm": 1.384702407056521,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8701857328414917,
      "num_tokens": 224469091.0,
      "step": 6450
    },
    {
      "epoch": 1.1979572887650882,
      "grad_norm": 1.6471958718444315,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8837380409240723,
      "num_tokens": 224497617.0,
      "step": 6451
    },
    {
      "epoch": 1.1981429897864437,
      "grad_norm": 1.4980170436567808,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8760218620300293,
      "num_tokens": 224532167.0,
      "step": 6452
    },
    {
      "epoch": 1.1983286908077995,
      "grad_norm": 1.3717395269191153,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8749845623970032,
      "num_tokens": 224570094.0,
      "step": 6453
    },
    {
      "epoch": 1.198514391829155,
      "grad_norm": 1.4682061626509055,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8813546895980835,
      "num_tokens": 224605799.0,
      "step": 6454
    },
    {
      "epoch": 1.1987000928505107,
      "grad_norm": 1.5642939334171744,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8749542832374573,
      "num_tokens": 224639114.0,
      "step": 6455
    },
    {
      "epoch": 1.1988857938718662,
      "grad_norm": 1.4312591220398656,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8775451183319092,
      "num_tokens": 224672436.0,
      "step": 6456
    },
    {
      "epoch": 1.199071494893222,
      "grad_norm": 1.5660054503282106,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8645247220993042,
      "num_tokens": 224707244.0,
      "step": 6457
    },
    {
      "epoch": 1.1992571959145775,
      "grad_norm": 1.4465611138270882,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8738081455230713,
      "num_tokens": 224744116.0,
      "step": 6458
    },
    {
      "epoch": 1.1994428969359332,
      "grad_norm": 1.5156439272499764,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8771322965621948,
      "num_tokens": 224777909.0,
      "step": 6459
    },
    {
      "epoch": 1.1996285979572887,
      "grad_norm": 1.5877308360599085,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.8843916058540344,
      "num_tokens": 224810711.0,
      "step": 6460
    },
    {
      "epoch": 1.1998142989786444,
      "grad_norm": 1.554061610772159,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8840747475624084,
      "num_tokens": 224842194.0,
      "step": 6461
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.5919967413837635,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8693341612815857,
      "num_tokens": 224873221.0,
      "step": 6462
    },
    {
      "epoch": 1.2001857010213557,
      "grad_norm": 1.6721449216723105,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8614605665206909,
      "num_tokens": 224907613.0,
      "step": 6463
    },
    {
      "epoch": 1.2003714020427112,
      "grad_norm": 1.5636542985907427,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8672002553939819,
      "num_tokens": 224942048.0,
      "step": 6464
    },
    {
      "epoch": 1.200557103064067,
      "grad_norm": 1.3549501996358104,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.871605634689331,
      "num_tokens": 224983586.0,
      "step": 6465
    },
    {
      "epoch": 1.2007428040854224,
      "grad_norm": 1.578204235497419,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8757795095443726,
      "num_tokens": 225016986.0,
      "step": 6466
    },
    {
      "epoch": 1.2009285051067782,
      "grad_norm": 1.5631930395740288,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8587204813957214,
      "num_tokens": 225049582.0,
      "step": 6467
    },
    {
      "epoch": 1.2011142061281337,
      "grad_norm": 1.4903544909482456,
      "learning_rate": 1e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8673672080039978,
      "num_tokens": 225087403.0,
      "step": 6468
    },
    {
      "epoch": 1.2012999071494894,
      "grad_norm": 1.4083383691602274,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8800255656242371,
      "num_tokens": 225124424.0,
      "step": 6469
    },
    {
      "epoch": 1.201485608170845,
      "grad_norm": 1.367229435343246,
      "learning_rate": 1e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8632548451423645,
      "num_tokens": 225166225.0,
      "step": 6470
    },
    {
      "epoch": 1.2016713091922004,
      "grad_norm": 1.4980618256421825,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.895187258720398,
      "num_tokens": 225199832.0,
      "step": 6471
    },
    {
      "epoch": 1.2018570102135562,
      "grad_norm": 1.5060387412729477,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8809213638305664,
      "num_tokens": 225234806.0,
      "step": 6472
    },
    {
      "epoch": 1.202042711234912,
      "grad_norm": 1.47653445292369,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.876013457775116,
      "num_tokens": 225270440.0,
      "step": 6473
    },
    {
      "epoch": 1.2022284122562674,
      "grad_norm": 1.5016526709539202,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8835042715072632,
      "num_tokens": 225300908.0,
      "step": 6474
    },
    {
      "epoch": 1.202414113277623,
      "grad_norm": 1.855191308279069,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8787013292312622,
      "num_tokens": 225332912.0,
      "step": 6475
    },
    {
      "epoch": 1.2025998142989787,
      "grad_norm": 1.5246443367240732,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.869215726852417,
      "num_tokens": 225369361.0,
      "step": 6476
    },
    {
      "epoch": 1.2027855153203342,
      "grad_norm": 1.4839154696378587,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8792796730995178,
      "num_tokens": 225403568.0,
      "step": 6477
    },
    {
      "epoch": 1.20297121634169,
      "grad_norm": 1.4445288197071207,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8752955794334412,
      "num_tokens": 225441822.0,
      "step": 6478
    },
    {
      "epoch": 1.2031569173630454,
      "grad_norm": 1.4698606492389392,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8641660809516907,
      "num_tokens": 225480799.0,
      "step": 6479
    },
    {
      "epoch": 1.2033426183844012,
      "grad_norm": 1.7608812166610597,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.866279125213623,
      "num_tokens": 225518548.0,
      "step": 6480
    },
    {
      "epoch": 1.2035283194057567,
      "grad_norm": 1.5085359655409956,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8905000686645508,
      "num_tokens": 225550366.0,
      "step": 6481
    },
    {
      "epoch": 1.2037140204271124,
      "grad_norm": 1.6211282838401577,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8747762441635132,
      "num_tokens": 225581213.0,
      "step": 6482
    },
    {
      "epoch": 1.203899721448468,
      "grad_norm": 1.5972547527610184,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8750032186508179,
      "num_tokens": 225610569.0,
      "step": 6483
    },
    {
      "epoch": 1.2040854224698236,
      "grad_norm": 1.458622359066799,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8677946329116821,
      "num_tokens": 225647356.0,
      "step": 6484
    },
    {
      "epoch": 1.2042711234911792,
      "grad_norm": 1.527622942330561,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8717145919799805,
      "num_tokens": 225682805.0,
      "step": 6485
    },
    {
      "epoch": 1.2044568245125349,
      "grad_norm": 1.5575457260423589,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8793516159057617,
      "num_tokens": 225713520.0,
      "step": 6486
    },
    {
      "epoch": 1.2046425255338904,
      "grad_norm": 1.4729088911543553,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8810315132141113,
      "num_tokens": 225747834.0,
      "step": 6487
    },
    {
      "epoch": 1.2048282265552461,
      "grad_norm": 1.5656282766950989,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8620641827583313,
      "num_tokens": 225783816.0,
      "step": 6488
    },
    {
      "epoch": 1.2050139275766016,
      "grad_norm": 1.4765565093616428,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8770694732666016,
      "num_tokens": 225818827.0,
      "step": 6489
    },
    {
      "epoch": 1.2051996285979574,
      "grad_norm": 1.4356689934149196,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.878277063369751,
      "num_tokens": 225855303.0,
      "step": 6490
    },
    {
      "epoch": 1.2053853296193129,
      "grad_norm": 1.506678749623122,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8902305364608765,
      "num_tokens": 225887410.0,
      "step": 6491
    },
    {
      "epoch": 1.2055710306406686,
      "grad_norm": 1.499381154255952,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8766453862190247,
      "num_tokens": 225924018.0,
      "step": 6492
    },
    {
      "epoch": 1.2057567316620241,
      "grad_norm": 1.5858151471107924,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8809802532196045,
      "num_tokens": 225953022.0,
      "step": 6493
    },
    {
      "epoch": 1.2059424326833796,
      "grad_norm": 1.712455040928464,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8725968599319458,
      "num_tokens": 225982356.0,
      "step": 6494
    },
    {
      "epoch": 1.2061281337047354,
      "grad_norm": 1.5612667751815463,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8782898187637329,
      "num_tokens": 226014756.0,
      "step": 6495
    },
    {
      "epoch": 1.206313834726091,
      "grad_norm": 1.7660080377581058,
      "learning_rate": 1e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8674587607383728,
      "num_tokens": 226041279.0,
      "step": 6496
    },
    {
      "epoch": 1.2064995357474466,
      "grad_norm": 1.4530221641936054,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8789290189743042,
      "num_tokens": 226076591.0,
      "step": 6497
    },
    {
      "epoch": 1.2066852367688021,
      "grad_norm": 1.6309916757596201,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8829755783081055,
      "num_tokens": 226107468.0,
      "step": 6498
    },
    {
      "epoch": 1.2068709377901579,
      "grad_norm": 1.5139644007411543,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.8914477825164795,
      "num_tokens": 226139902.0,
      "step": 6499
    },
    {
      "epoch": 1.2070566388115134,
      "grad_norm": 1.608176077488565,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8883495330810547,
      "num_tokens": 226167102.0,
      "step": 6500
    },
    {
      "epoch": 1.207242339832869,
      "grad_norm": 1.414577278944724,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8852468132972717,
      "num_tokens": 226203062.0,
      "step": 6501
    },
    {
      "epoch": 1.2074280408542246,
      "grad_norm": 1.4010588788931364,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8743531703948975,
      "num_tokens": 226241663.0,
      "step": 6502
    },
    {
      "epoch": 1.2076137418755803,
      "grad_norm": 1.4404456203342177,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.867150068283081,
      "num_tokens": 226281652.0,
      "step": 6503
    },
    {
      "epoch": 1.2077994428969359,
      "grad_norm": 1.4880749393580566,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8729366064071655,
      "num_tokens": 226315477.0,
      "step": 6504
    },
    {
      "epoch": 1.2079851439182916,
      "grad_norm": 1.5989467189441937,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8751589059829712,
      "num_tokens": 226346779.0,
      "step": 6505
    },
    {
      "epoch": 1.208170844939647,
      "grad_norm": 1.7292149984685163,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8811101317405701,
      "num_tokens": 226373201.0,
      "step": 6506
    },
    {
      "epoch": 1.2083565459610028,
      "grad_norm": 1.6363263065115403,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8700180053710938,
      "num_tokens": 226402158.0,
      "step": 6507
    },
    {
      "epoch": 1.2085422469823583,
      "grad_norm": 1.535128797320105,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8713394403457642,
      "num_tokens": 226437260.0,
      "step": 6508
    },
    {
      "epoch": 1.208727948003714,
      "grad_norm": 1.442235725948333,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8733301162719727,
      "num_tokens": 226476827.0,
      "step": 6509
    },
    {
      "epoch": 1.2089136490250696,
      "grad_norm": 1.40333278704182,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.865166187286377,
      "num_tokens": 226520252.0,
      "step": 6510
    },
    {
      "epoch": 1.2090993500464253,
      "grad_norm": 1.3966570494599517,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8750991225242615,
      "num_tokens": 226558709.0,
      "step": 6511
    },
    {
      "epoch": 1.2092850510677808,
      "grad_norm": 1.4805679448936417,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8909488916397095,
      "num_tokens": 226589545.0,
      "step": 6512
    },
    {
      "epoch": 1.2094707520891366,
      "grad_norm": 1.5869103486184661,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8791550993919373,
      "num_tokens": 226623975.0,
      "step": 6513
    },
    {
      "epoch": 1.209656453110492,
      "grad_norm": 1.6516444043378569,
      "learning_rate": 1e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8660705089569092,
      "num_tokens": 226658966.0,
      "step": 6514
    },
    {
      "epoch": 1.2098421541318478,
      "grad_norm": 1.4496405369353214,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8745633959770203,
      "num_tokens": 226696707.0,
      "step": 6515
    },
    {
      "epoch": 1.2100278551532033,
      "grad_norm": 1.5402934338619578,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8673406839370728,
      "num_tokens": 226733534.0,
      "step": 6516
    },
    {
      "epoch": 1.2102135561745588,
      "grad_norm": 1.414737466085277,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8737373948097229,
      "num_tokens": 226772264.0,
      "step": 6517
    },
    {
      "epoch": 1.2103992571959146,
      "grad_norm": 1.6062807017080845,
      "learning_rate": 1e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8605968952178955,
      "num_tokens": 226805551.0,
      "step": 6518
    },
    {
      "epoch": 1.2105849582172703,
      "grad_norm": 1.6732894119016763,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.874854564666748,
      "num_tokens": 226837850.0,
      "step": 6519
    },
    {
      "epoch": 1.2107706592386258,
      "grad_norm": 1.5019874115845504,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8707746863365173,
      "num_tokens": 226875346.0,
      "step": 6520
    },
    {
      "epoch": 1.2109563602599813,
      "grad_norm": 1.4980798864431062,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8804621696472168,
      "num_tokens": 226913459.0,
      "step": 6521
    },
    {
      "epoch": 1.211142061281337,
      "grad_norm": 1.5624877502573835,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8802939653396606,
      "num_tokens": 226946172.0,
      "step": 6522
    },
    {
      "epoch": 1.2113277623026926,
      "grad_norm": 1.580801248237279,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8903499841690063,
      "num_tokens": 226973479.0,
      "step": 6523
    },
    {
      "epoch": 1.2115134633240483,
      "grad_norm": 1.7704595823740912,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.874648928642273,
      "num_tokens": 226999128.0,
      "step": 6524
    },
    {
      "epoch": 1.2116991643454038,
      "grad_norm": 1.4192037913326956,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8761828541755676,
      "num_tokens": 227037772.0,
      "step": 6525
    },
    {
      "epoch": 1.2118848653667595,
      "grad_norm": 1.5053509203762632,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8948348760604858,
      "num_tokens": 227069840.0,
      "step": 6526
    },
    {
      "epoch": 1.212070566388115,
      "grad_norm": 1.430277030474564,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8680770397186279,
      "num_tokens": 227108414.0,
      "step": 6527
    },
    {
      "epoch": 1.2122562674094708,
      "grad_norm": 1.4863362316182223,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8711196184158325,
      "num_tokens": 227144789.0,
      "step": 6528
    },
    {
      "epoch": 1.2124419684308263,
      "grad_norm": 1.3897255129777453,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8828226327896118,
      "num_tokens": 227184611.0,
      "step": 6529
    },
    {
      "epoch": 1.212627669452182,
      "grad_norm": 1.4361342132043455,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8703907132148743,
      "num_tokens": 227222122.0,
      "step": 6530
    },
    {
      "epoch": 1.2128133704735375,
      "grad_norm": 1.5073296034886006,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8779069781303406,
      "num_tokens": 227256729.0,
      "step": 6531
    },
    {
      "epoch": 1.2129990714948933,
      "grad_norm": 1.5179180711608649,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8665851354598999,
      "num_tokens": 227291858.0,
      "step": 6532
    },
    {
      "epoch": 1.2131847725162488,
      "grad_norm": 1.5191710171179504,
      "learning_rate": 1e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8631792068481445,
      "num_tokens": 227326911.0,
      "step": 6533
    },
    {
      "epoch": 1.2133704735376045,
      "grad_norm": 1.672972359041,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8775759935379028,
      "num_tokens": 227355063.0,
      "step": 6534
    },
    {
      "epoch": 1.21355617455896,
      "grad_norm": 1.5853260013390613,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8586868047714233,
      "num_tokens": 227390662.0,
      "step": 6535
    },
    {
      "epoch": 1.2137418755803158,
      "grad_norm": 1.4432975339584473,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8739301562309265,
      "num_tokens": 227429083.0,
      "step": 6536
    },
    {
      "epoch": 1.2139275766016713,
      "grad_norm": 1.5310925183829787,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8921833038330078,
      "num_tokens": 227457737.0,
      "step": 6537
    },
    {
      "epoch": 1.214113277623027,
      "grad_norm": 1.5902041189873803,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8636893033981323,
      "num_tokens": 227492721.0,
      "step": 6538
    },
    {
      "epoch": 1.2142989786443825,
      "grad_norm": 1.5485424995048667,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.874929666519165,
      "num_tokens": 227527093.0,
      "step": 6539
    },
    {
      "epoch": 1.2144846796657383,
      "grad_norm": 1.5435097507078877,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8655394315719604,
      "num_tokens": 227561585.0,
      "step": 6540
    },
    {
      "epoch": 1.2146703806870938,
      "grad_norm": 1.4720158627478168,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8775275945663452,
      "num_tokens": 227596921.0,
      "step": 6541
    },
    {
      "epoch": 1.2148560817084495,
      "grad_norm": 1.5826351205689149,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8787108659744263,
      "num_tokens": 227627780.0,
      "step": 6542
    },
    {
      "epoch": 1.215041782729805,
      "grad_norm": 1.510909274250614,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8912577033042908,
      "num_tokens": 227661859.0,
      "step": 6543
    },
    {
      "epoch": 1.2152274837511605,
      "grad_norm": 1.585239762455941,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8789066076278687,
      "num_tokens": 227694459.0,
      "step": 6544
    },
    {
      "epoch": 1.2154131847725163,
      "grad_norm": 1.5578769921513427,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8800078630447388,
      "num_tokens": 227726935.0,
      "step": 6545
    },
    {
      "epoch": 1.215598885793872,
      "grad_norm": 1.4235019807868419,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8783068656921387,
      "num_tokens": 227766021.0,
      "step": 6546
    },
    {
      "epoch": 1.2157845868152275,
      "grad_norm": 1.526651672134446,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.877095103263855,
      "num_tokens": 227799534.0,
      "step": 6547
    },
    {
      "epoch": 1.215970287836583,
      "grad_norm": 1.7718389369853507,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8717699646949768,
      "num_tokens": 227825920.0,
      "step": 6548
    },
    {
      "epoch": 1.2161559888579387,
      "grad_norm": 1.5446419949586168,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8839308023452759,
      "num_tokens": 227856814.0,
      "step": 6549
    },
    {
      "epoch": 1.2163416898792943,
      "grad_norm": 1.5894527107902454,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8846165537834167,
      "num_tokens": 227888568.0,
      "step": 6550
    },
    {
      "epoch": 1.21652739090065,
      "grad_norm": 1.543876382138804,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8819997310638428,
      "num_tokens": 227917733.0,
      "step": 6551
    },
    {
      "epoch": 1.2167130919220055,
      "grad_norm": 1.5227230980186077,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8745218515396118,
      "num_tokens": 227951597.0,
      "step": 6552
    },
    {
      "epoch": 1.2168987929433612,
      "grad_norm": 1.6365992201746091,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8819562196731567,
      "num_tokens": 227979955.0,
      "step": 6553
    },
    {
      "epoch": 1.2170844939647167,
      "grad_norm": 1.403029383991772,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8964288234710693,
      "num_tokens": 228016540.0,
      "step": 6554
    },
    {
      "epoch": 1.2172701949860725,
      "grad_norm": 1.3875388068641101,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8806713819503784,
      "num_tokens": 228054305.0,
      "step": 6555
    },
    {
      "epoch": 1.217455896007428,
      "grad_norm": 1.5776592641579816,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8772467970848083,
      "num_tokens": 228084852.0,
      "step": 6556
    },
    {
      "epoch": 1.2176415970287837,
      "grad_norm": 1.4027484537149113,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8864212036132812,
      "num_tokens": 228122122.0,
      "step": 6557
    },
    {
      "epoch": 1.2178272980501392,
      "grad_norm": 1.5264251617176032,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8802213668823242,
      "num_tokens": 228155203.0,
      "step": 6558
    },
    {
      "epoch": 1.218012999071495,
      "grad_norm": 1.514774831681088,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8782913088798523,
      "num_tokens": 228191882.0,
      "step": 6559
    },
    {
      "epoch": 1.2181987000928505,
      "grad_norm": 1.506614926988119,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8772455453872681,
      "num_tokens": 228226792.0,
      "step": 6560
    },
    {
      "epoch": 1.2183844011142062,
      "grad_norm": 1.4596707789707346,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8790626525878906,
      "num_tokens": 228262214.0,
      "step": 6561
    },
    {
      "epoch": 1.2185701021355617,
      "grad_norm": 1.7042956824974447,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8857178092002869,
      "num_tokens": 228288636.0,
      "step": 6562
    },
    {
      "epoch": 1.2187558031569174,
      "grad_norm": 1.483121407413737,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8731305599212646,
      "num_tokens": 228326204.0,
      "step": 6563
    },
    {
      "epoch": 1.218941504178273,
      "grad_norm": 1.3371380647213928,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8735036849975586,
      "num_tokens": 228370647.0,
      "step": 6564
    },
    {
      "epoch": 1.2191272051996287,
      "grad_norm": 1.6434170088363809,
      "learning_rate": 1e-06,
      "loss": 0.4042,
      "mean_token_accuracy": 0.8612596988677979,
      "num_tokens": 228404690.0,
      "step": 6565
    },
    {
      "epoch": 1.2193129062209842,
      "grad_norm": 1.5689593310101955,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8763365149497986,
      "num_tokens": 228437132.0,
      "step": 6566
    },
    {
      "epoch": 1.2194986072423397,
      "grad_norm": 1.3724753929302291,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8797228336334229,
      "num_tokens": 228474903.0,
      "step": 6567
    },
    {
      "epoch": 1.2196843082636954,
      "grad_norm": 1.4278749993475068,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8803554773330688,
      "num_tokens": 228513909.0,
      "step": 6568
    },
    {
      "epoch": 1.2198700092850512,
      "grad_norm": 1.6265013870128722,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8684920072555542,
      "num_tokens": 228545984.0,
      "step": 6569
    },
    {
      "epoch": 1.2200557103064067,
      "grad_norm": 1.517687889989714,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8832900524139404,
      "num_tokens": 228577955.0,
      "step": 6570
    },
    {
      "epoch": 1.2202414113277622,
      "grad_norm": 1.4798108702525927,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8822094202041626,
      "num_tokens": 228610798.0,
      "step": 6571
    },
    {
      "epoch": 1.220427112349118,
      "grad_norm": 1.4721036317748275,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8727392554283142,
      "num_tokens": 228647831.0,
      "step": 6572
    },
    {
      "epoch": 1.2206128133704734,
      "grad_norm": 1.4893751834904216,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8844227194786072,
      "num_tokens": 228681396.0,
      "step": 6573
    },
    {
      "epoch": 1.2207985143918292,
      "grad_norm": 1.5298526283958636,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8757623434066772,
      "num_tokens": 228717161.0,
      "step": 6574
    },
    {
      "epoch": 1.2209842154131847,
      "grad_norm": 1.4760203353259442,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8720211982727051,
      "num_tokens": 228752100.0,
      "step": 6575
    },
    {
      "epoch": 1.2211699164345404,
      "grad_norm": 1.4838257648223054,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8836106061935425,
      "num_tokens": 228786287.0,
      "step": 6576
    },
    {
      "epoch": 1.221355617455896,
      "grad_norm": 1.5272342185785543,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8846670389175415,
      "num_tokens": 228815049.0,
      "step": 6577
    },
    {
      "epoch": 1.2215413184772517,
      "grad_norm": 1.6361110956742044,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8687669634819031,
      "num_tokens": 228843485.0,
      "step": 6578
    },
    {
      "epoch": 1.2217270194986072,
      "grad_norm": 1.6099277958504556,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8854730129241943,
      "num_tokens": 228873376.0,
      "step": 6579
    },
    {
      "epoch": 1.221912720519963,
      "grad_norm": 1.3796997050283546,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8731443881988525,
      "num_tokens": 228916173.0,
      "step": 6580
    },
    {
      "epoch": 1.2220984215413184,
      "grad_norm": 1.4396059248005537,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8710389733314514,
      "num_tokens": 228958103.0,
      "step": 6581
    },
    {
      "epoch": 1.2222841225626742,
      "grad_norm": 1.571018100970838,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8918424844741821,
      "num_tokens": 228988270.0,
      "step": 6582
    },
    {
      "epoch": 1.2224698235840297,
      "grad_norm": 1.5900880072009418,
      "learning_rate": 1e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8551322221755981,
      "num_tokens": 229026634.0,
      "step": 6583
    },
    {
      "epoch": 1.2226555246053854,
      "grad_norm": 1.5146312417092775,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8824450969696045,
      "num_tokens": 229061096.0,
      "step": 6584
    },
    {
      "epoch": 1.222841225626741,
      "grad_norm": 1.5043911668440986,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8767572045326233,
      "num_tokens": 229096385.0,
      "step": 6585
    },
    {
      "epoch": 1.2230269266480966,
      "grad_norm": 1.7156865512239088,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8664876222610474,
      "num_tokens": 229128666.0,
      "step": 6586
    },
    {
      "epoch": 1.2232126276694522,
      "grad_norm": 1.4549625270607618,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8833001852035522,
      "num_tokens": 229162480.0,
      "step": 6587
    },
    {
      "epoch": 1.223398328690808,
      "grad_norm": 1.6772707538123772,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8666164875030518,
      "num_tokens": 229192275.0,
      "step": 6588
    },
    {
      "epoch": 1.2235840297121634,
      "grad_norm": 1.4868808872817045,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8743319511413574,
      "num_tokens": 229226374.0,
      "step": 6589
    },
    {
      "epoch": 1.223769730733519,
      "grad_norm": 1.790584414926049,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8666582703590393,
      "num_tokens": 229252046.0,
      "step": 6590
    },
    {
      "epoch": 1.2239554317548746,
      "grad_norm": 1.4099234259717293,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8788860440254211,
      "num_tokens": 229290496.0,
      "step": 6591
    },
    {
      "epoch": 1.2241411327762304,
      "grad_norm": 1.5990102398764563,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8740715980529785,
      "num_tokens": 229321181.0,
      "step": 6592
    },
    {
      "epoch": 1.224326833797586,
      "grad_norm": 1.5476409886349463,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8763124942779541,
      "num_tokens": 229356837.0,
      "step": 6593
    },
    {
      "epoch": 1.2245125348189414,
      "grad_norm": 1.617232480076042,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8762090802192688,
      "num_tokens": 229387477.0,
      "step": 6594
    },
    {
      "epoch": 1.2246982358402971,
      "grad_norm": 1.49024880078771,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8780434727668762,
      "num_tokens": 229422500.0,
      "step": 6595
    },
    {
      "epoch": 1.2248839368616526,
      "grad_norm": 1.5293019292363133,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8772609829902649,
      "num_tokens": 229453388.0,
      "step": 6596
    },
    {
      "epoch": 1.2250696378830084,
      "grad_norm": 1.6181215270665121,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8736530542373657,
      "num_tokens": 229484425.0,
      "step": 6597
    },
    {
      "epoch": 1.2252553389043639,
      "grad_norm": 1.3562529166332329,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.8954730033874512,
      "num_tokens": 229519529.0,
      "step": 6598
    },
    {
      "epoch": 1.2254410399257196,
      "grad_norm": 1.5793422805414723,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.860621452331543,
      "num_tokens": 229554566.0,
      "step": 6599
    },
    {
      "epoch": 1.2256267409470751,
      "grad_norm": 1.5972362479943114,
      "learning_rate": 1e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8575078248977661,
      "num_tokens": 229586619.0,
      "step": 6600
    },
    {
      "epoch": 1.2258124419684309,
      "grad_norm": 1.5828418119820684,
      "learning_rate": 1e-06,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8576556444168091,
      "num_tokens": 229620385.0,
      "step": 6601
    },
    {
      "epoch": 1.2259981429897864,
      "grad_norm": 1.6174308706879712,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.883419930934906,
      "num_tokens": 229650442.0,
      "step": 6602
    },
    {
      "epoch": 1.226183844011142,
      "grad_norm": 1.5981699654615467,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8835403919219971,
      "num_tokens": 229681133.0,
      "step": 6603
    },
    {
      "epoch": 1.2263695450324976,
      "grad_norm": 1.3710882087234908,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8759011626243591,
      "num_tokens": 229721610.0,
      "step": 6604
    },
    {
      "epoch": 1.2265552460538534,
      "grad_norm": 1.4208674176763396,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8810920715332031,
      "num_tokens": 229761663.0,
      "step": 6605
    },
    {
      "epoch": 1.2267409470752089,
      "grad_norm": 1.4482472901116252,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8819540143013,
      "num_tokens": 229800040.0,
      "step": 6606
    },
    {
      "epoch": 1.2269266480965646,
      "grad_norm": 1.741596773078921,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8780486583709717,
      "num_tokens": 229829521.0,
      "step": 6607
    },
    {
      "epoch": 1.22711234911792,
      "grad_norm": 1.6408009526579144,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8734840750694275,
      "num_tokens": 229860345.0,
      "step": 6608
    },
    {
      "epoch": 1.2272980501392758,
      "grad_norm": 1.431499988984523,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8722066283226013,
      "num_tokens": 229902276.0,
      "step": 6609
    },
    {
      "epoch": 1.2274837511606314,
      "grad_norm": 1.587519493305394,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8859617710113525,
      "num_tokens": 229932712.0,
      "step": 6610
    },
    {
      "epoch": 1.227669452181987,
      "grad_norm": 1.3495477999093495,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8862324953079224,
      "num_tokens": 229971749.0,
      "step": 6611
    },
    {
      "epoch": 1.2278551532033426,
      "grad_norm": 1.3030347049761475,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8747493624687195,
      "num_tokens": 230015765.0,
      "step": 6612
    },
    {
      "epoch": 1.2280408542246983,
      "grad_norm": 1.6465395479991087,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8716726303100586,
      "num_tokens": 230046351.0,
      "step": 6613
    },
    {
      "epoch": 1.2282265552460538,
      "grad_norm": 1.4167151703565106,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8706920146942139,
      "num_tokens": 230086335.0,
      "step": 6614
    },
    {
      "epoch": 1.2284122562674096,
      "grad_norm": 1.4139940527634856,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.882430911064148,
      "num_tokens": 230123907.0,
      "step": 6615
    },
    {
      "epoch": 1.228597957288765,
      "grad_norm": 1.4465114481079198,
      "learning_rate": 1e-06,
      "loss": 0.296,
      "mean_token_accuracy": 0.8962750434875488,
      "num_tokens": 230156562.0,
      "step": 6616
    },
    {
      "epoch": 1.2287836583101206,
      "grad_norm": 1.4564979190330096,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8929896354675293,
      "num_tokens": 230188681.0,
      "step": 6617
    },
    {
      "epoch": 1.2289693593314763,
      "grad_norm": 1.4119478398221343,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8754700422286987,
      "num_tokens": 230228026.0,
      "step": 6618
    },
    {
      "epoch": 1.229155060352832,
      "grad_norm": 1.452619551411527,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.886919379234314,
      "num_tokens": 230265284.0,
      "step": 6619
    },
    {
      "epoch": 1.2293407613741876,
      "grad_norm": 1.6844757047921908,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.870653510093689,
      "num_tokens": 230297614.0,
      "step": 6620
    },
    {
      "epoch": 1.229526462395543,
      "grad_norm": 1.6524103966733088,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8682315349578857,
      "num_tokens": 230331233.0,
      "step": 6621
    },
    {
      "epoch": 1.2297121634168988,
      "grad_norm": 1.476575046662182,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.8846760988235474,
      "num_tokens": 230368139.0,
      "step": 6622
    },
    {
      "epoch": 1.2298978644382543,
      "grad_norm": 1.5619753917263093,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8821887373924255,
      "num_tokens": 230399477.0,
      "step": 6623
    },
    {
      "epoch": 1.23008356545961,
      "grad_norm": 1.5321780336428246,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8858274221420288,
      "num_tokens": 230427791.0,
      "step": 6624
    },
    {
      "epoch": 1.2302692664809656,
      "grad_norm": 1.3477034259746505,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.888745903968811,
      "num_tokens": 230465929.0,
      "step": 6625
    },
    {
      "epoch": 1.2304549675023213,
      "grad_norm": 1.5024569019839735,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8776490688323975,
      "num_tokens": 230502492.0,
      "step": 6626
    },
    {
      "epoch": 1.2306406685236768,
      "grad_norm": 1.4959566046001853,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8781290054321289,
      "num_tokens": 230537037.0,
      "step": 6627
    },
    {
      "epoch": 1.2308263695450326,
      "grad_norm": 1.4322438722884812,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8737516403198242,
      "num_tokens": 230575349.0,
      "step": 6628
    },
    {
      "epoch": 1.231012070566388,
      "grad_norm": 1.5292537147294487,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8721676468849182,
      "num_tokens": 230610992.0,
      "step": 6629
    },
    {
      "epoch": 1.2311977715877438,
      "grad_norm": 1.454618111840628,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8740017414093018,
      "num_tokens": 230647439.0,
      "step": 6630
    },
    {
      "epoch": 1.2313834726090993,
      "grad_norm": 1.4203705926309909,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8785504698753357,
      "num_tokens": 230685764.0,
      "step": 6631
    },
    {
      "epoch": 1.231569173630455,
      "grad_norm": 1.4357251390225283,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8919520378112793,
      "num_tokens": 230720662.0,
      "step": 6632
    },
    {
      "epoch": 1.2317548746518105,
      "grad_norm": 1.7186295116346193,
      "learning_rate": 1e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8584447503089905,
      "num_tokens": 230754875.0,
      "step": 6633
    },
    {
      "epoch": 1.2319405756731663,
      "grad_norm": 1.6234326691811714,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8605486750602722,
      "num_tokens": 230787485.0,
      "step": 6634
    },
    {
      "epoch": 1.2321262766945218,
      "grad_norm": 1.5067472147690841,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8650400638580322,
      "num_tokens": 230826687.0,
      "step": 6635
    },
    {
      "epoch": 1.2323119777158775,
      "grad_norm": 1.5359145194035506,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8829325437545776,
      "num_tokens": 230859235.0,
      "step": 6636
    },
    {
      "epoch": 1.232497678737233,
      "grad_norm": 1.5015108541148159,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8728883266448975,
      "num_tokens": 230895370.0,
      "step": 6637
    },
    {
      "epoch": 1.2326833797585888,
      "grad_norm": 1.371594068056825,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8739169239997864,
      "num_tokens": 230933745.0,
      "step": 6638
    },
    {
      "epoch": 1.2328690807799443,
      "grad_norm": 1.655760370291388,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8808917999267578,
      "num_tokens": 230962263.0,
      "step": 6639
    },
    {
      "epoch": 1.2330547818012998,
      "grad_norm": 1.6084469390648093,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8848216533660889,
      "num_tokens": 230991090.0,
      "step": 6640
    },
    {
      "epoch": 1.2332404828226555,
      "grad_norm": 1.6690699629096526,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8964135050773621,
      "num_tokens": 231014733.0,
      "step": 6641
    },
    {
      "epoch": 1.2334261838440113,
      "grad_norm": 1.53002483578125,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8712929487228394,
      "num_tokens": 231049062.0,
      "step": 6642
    },
    {
      "epoch": 1.2336118848653668,
      "grad_norm": 1.396011878269588,
      "learning_rate": 1e-06,
      "loss": 0.3246,
      "mean_token_accuracy": 0.8849549293518066,
      "num_tokens": 231081855.0,
      "step": 6643
    },
    {
      "epoch": 1.2337975858867223,
      "grad_norm": 1.7245541447097508,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8758077025413513,
      "num_tokens": 231107242.0,
      "step": 6644
    },
    {
      "epoch": 1.233983286908078,
      "grad_norm": 1.5477909136571006,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8934083580970764,
      "num_tokens": 231138770.0,
      "step": 6645
    },
    {
      "epoch": 1.2341689879294335,
      "grad_norm": 1.5345602844674524,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8800244331359863,
      "num_tokens": 231172652.0,
      "step": 6646
    },
    {
      "epoch": 1.2343546889507893,
      "grad_norm": 1.7686112998093744,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8842077255249023,
      "num_tokens": 231195606.0,
      "step": 6647
    },
    {
      "epoch": 1.2345403899721448,
      "grad_norm": 1.8483927658257653,
      "learning_rate": 1e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8567429780960083,
      "num_tokens": 231222539.0,
      "step": 6648
    },
    {
      "epoch": 1.2347260909935005,
      "grad_norm": 1.4070433423447182,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.890029788017273,
      "num_tokens": 231259167.0,
      "step": 6649
    },
    {
      "epoch": 1.234911792014856,
      "grad_norm": 1.5551373833462034,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8594627976417542,
      "num_tokens": 231293573.0,
      "step": 6650
    },
    {
      "epoch": 1.2350974930362117,
      "grad_norm": 1.5939404196663594,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8790541887283325,
      "num_tokens": 231325816.0,
      "step": 6651
    },
    {
      "epoch": 1.2352831940575673,
      "grad_norm": 1.4099052190542132,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8853781223297119,
      "num_tokens": 231360141.0,
      "step": 6652
    },
    {
      "epoch": 1.235468895078923,
      "grad_norm": 1.6328845359291797,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8686439394950867,
      "num_tokens": 231393035.0,
      "step": 6653
    },
    {
      "epoch": 1.2356545961002785,
      "grad_norm": 1.5323299743998402,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8664060235023499,
      "num_tokens": 231428455.0,
      "step": 6654
    },
    {
      "epoch": 1.2358402971216342,
      "grad_norm": 1.5149289471200404,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8812041878700256,
      "num_tokens": 231460354.0,
      "step": 6655
    },
    {
      "epoch": 1.2360259981429897,
      "grad_norm": 1.531342060064942,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.872010350227356,
      "num_tokens": 231493548.0,
      "step": 6656
    },
    {
      "epoch": 1.2362116991643455,
      "grad_norm": 1.583088954121414,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8776371479034424,
      "num_tokens": 231523612.0,
      "step": 6657
    },
    {
      "epoch": 1.236397400185701,
      "grad_norm": 1.5033981475780007,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8781620264053345,
      "num_tokens": 231557725.0,
      "step": 6658
    },
    {
      "epoch": 1.2365831012070567,
      "grad_norm": 1.5148611082433223,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8705267906188965,
      "num_tokens": 231594287.0,
      "step": 6659
    },
    {
      "epoch": 1.2367688022284122,
      "grad_norm": 1.4508536688773344,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.875568151473999,
      "num_tokens": 231631551.0,
      "step": 6660
    },
    {
      "epoch": 1.236954503249768,
      "grad_norm": 1.4676254007316114,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8800101280212402,
      "num_tokens": 231666261.0,
      "step": 6661
    },
    {
      "epoch": 1.2371402042711235,
      "grad_norm": 1.5853135324993373,
      "learning_rate": 1e-06,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8647556304931641,
      "num_tokens": 231700547.0,
      "step": 6662
    },
    {
      "epoch": 1.237325905292479,
      "grad_norm": 1.4873254569437935,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8774459958076477,
      "num_tokens": 231737613.0,
      "step": 6663
    },
    {
      "epoch": 1.2375116063138347,
      "grad_norm": 1.6144562004495788,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8749880790710449,
      "num_tokens": 231768335.0,
      "step": 6664
    },
    {
      "epoch": 1.2376973073351905,
      "grad_norm": 1.4583052365743094,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8949671983718872,
      "num_tokens": 231799466.0,
      "step": 6665
    },
    {
      "epoch": 1.237883008356546,
      "grad_norm": 1.6508853390948517,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8696234822273254,
      "num_tokens": 231832614.0,
      "step": 6666
    },
    {
      "epoch": 1.2380687093779015,
      "grad_norm": 1.4728106914615775,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8775771856307983,
      "num_tokens": 231870348.0,
      "step": 6667
    },
    {
      "epoch": 1.2382544103992572,
      "grad_norm": 1.5062910907328138,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8853506445884705,
      "num_tokens": 231905159.0,
      "step": 6668
    },
    {
      "epoch": 1.2384401114206127,
      "grad_norm": 1.493490499125773,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8798416256904602,
      "num_tokens": 231939982.0,
      "step": 6669
    },
    {
      "epoch": 1.2386258124419685,
      "grad_norm": 1.5247131605086912,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8795467019081116,
      "num_tokens": 231974751.0,
      "step": 6670
    },
    {
      "epoch": 1.238811513463324,
      "grad_norm": 1.6674029244725312,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8785244226455688,
      "num_tokens": 232006960.0,
      "step": 6671
    },
    {
      "epoch": 1.2389972144846797,
      "grad_norm": 1.4655889581181276,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.879248857498169,
      "num_tokens": 232045264.0,
      "step": 6672
    },
    {
      "epoch": 1.2391829155060352,
      "grad_norm": 1.5090631060873338,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8749980926513672,
      "num_tokens": 232079213.0,
      "step": 6673
    },
    {
      "epoch": 1.239368616527391,
      "grad_norm": 1.5346318599016329,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8863368034362793,
      "num_tokens": 232110690.0,
      "step": 6674
    },
    {
      "epoch": 1.2395543175487465,
      "grad_norm": 1.502378907422631,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8781657218933105,
      "num_tokens": 232145589.0,
      "step": 6675
    },
    {
      "epoch": 1.2397400185701022,
      "grad_norm": 1.5076135860082092,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.875491738319397,
      "num_tokens": 232180927.0,
      "step": 6676
    },
    {
      "epoch": 1.2399257195914577,
      "grad_norm": 1.4443740933848435,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8765869140625,
      "num_tokens": 232216485.0,
      "step": 6677
    },
    {
      "epoch": 1.2401114206128134,
      "grad_norm": 1.5871476956947843,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8665195107460022,
      "num_tokens": 232249563.0,
      "step": 6678
    },
    {
      "epoch": 1.240297121634169,
      "grad_norm": 1.5914683758873223,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.871637225151062,
      "num_tokens": 232278826.0,
      "step": 6679
    },
    {
      "epoch": 1.2404828226555247,
      "grad_norm": 1.4041967798236963,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8872039914131165,
      "num_tokens": 232313754.0,
      "step": 6680
    },
    {
      "epoch": 1.2406685236768802,
      "grad_norm": 1.5580132953745782,
      "learning_rate": 1e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8651602864265442,
      "num_tokens": 232349902.0,
      "step": 6681
    },
    {
      "epoch": 1.240854224698236,
      "grad_norm": 1.6034237042214594,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8839682340621948,
      "num_tokens": 232380031.0,
      "step": 6682
    },
    {
      "epoch": 1.2410399257195914,
      "grad_norm": 1.6122651062846014,
      "learning_rate": 1e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8636013865470886,
      "num_tokens": 232414404.0,
      "step": 6683
    },
    {
      "epoch": 1.2412256267409472,
      "grad_norm": 1.5393587964421218,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8703354597091675,
      "num_tokens": 232449516.0,
      "step": 6684
    },
    {
      "epoch": 1.2414113277623027,
      "grad_norm": 1.5862271301576558,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8761830925941467,
      "num_tokens": 232482661.0,
      "step": 6685
    },
    {
      "epoch": 1.2415970287836582,
      "grad_norm": 1.5148698505231626,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8743429183959961,
      "num_tokens": 232519004.0,
      "step": 6686
    },
    {
      "epoch": 1.241782729805014,
      "grad_norm": 1.4959467232794679,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8793928623199463,
      "num_tokens": 232552725.0,
      "step": 6687
    },
    {
      "epoch": 1.2419684308263697,
      "grad_norm": 1.5918202152535688,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8561100959777832,
      "num_tokens": 232589149.0,
      "step": 6688
    },
    {
      "epoch": 1.2421541318477252,
      "grad_norm": 1.5370609930269155,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8735415935516357,
      "num_tokens": 232627965.0,
      "step": 6689
    },
    {
      "epoch": 1.2423398328690807,
      "grad_norm": 1.5702829252004824,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8809866905212402,
      "num_tokens": 232661564.0,
      "step": 6690
    },
    {
      "epoch": 1.2425255338904364,
      "grad_norm": 1.7126016055698547,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8761858940124512,
      "num_tokens": 232690110.0,
      "step": 6691
    },
    {
      "epoch": 1.2427112349117921,
      "grad_norm": 1.4395049159881297,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8806905746459961,
      "num_tokens": 232725051.0,
      "step": 6692
    },
    {
      "epoch": 1.2428969359331477,
      "grad_norm": 1.4055746717775002,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8720873594284058,
      "num_tokens": 232763116.0,
      "step": 6693
    },
    {
      "epoch": 1.2430826369545032,
      "grad_norm": 1.4538634258626653,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8736569881439209,
      "num_tokens": 232798648.0,
      "step": 6694
    },
    {
      "epoch": 1.243268337975859,
      "grad_norm": 1.5363544157404738,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8614593744277954,
      "num_tokens": 232834865.0,
      "step": 6695
    },
    {
      "epoch": 1.2434540389972144,
      "grad_norm": 1.638144853976573,
      "learning_rate": 1e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8613102436065674,
      "num_tokens": 232867844.0,
      "step": 6696
    },
    {
      "epoch": 1.2436397400185701,
      "grad_norm": 1.381921841549402,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8832525014877319,
      "num_tokens": 232907119.0,
      "step": 6697
    },
    {
      "epoch": 1.2438254410399256,
      "grad_norm": 1.4765682396499071,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8695101737976074,
      "num_tokens": 232944265.0,
      "step": 6698
    },
    {
      "epoch": 1.2440111420612814,
      "grad_norm": 1.4751918394523267,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8807861804962158,
      "num_tokens": 232978530.0,
      "step": 6699
    },
    {
      "epoch": 1.244196843082637,
      "grad_norm": 1.4888794200154918,
      "learning_rate": 1e-06,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8697289824485779,
      "num_tokens": 233015263.0,
      "step": 6700
    },
    {
      "epoch": 1.2443825441039926,
      "grad_norm": 1.568088339881681,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8842700719833374,
      "num_tokens": 233045208.0,
      "step": 6701
    },
    {
      "epoch": 1.2445682451253481,
      "grad_norm": 1.5648386923488402,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8653687834739685,
      "num_tokens": 233082052.0,
      "step": 6702
    },
    {
      "epoch": 1.2447539461467039,
      "grad_norm": 1.3377904231231124,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8820487856864929,
      "num_tokens": 233121771.0,
      "step": 6703
    },
    {
      "epoch": 1.2449396471680594,
      "grad_norm": 1.412567664696561,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8798080086708069,
      "num_tokens": 233156756.0,
      "step": 6704
    },
    {
      "epoch": 1.2451253481894151,
      "grad_norm": 1.615074685430925,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8680553436279297,
      "num_tokens": 233193376.0,
      "step": 6705
    },
    {
      "epoch": 1.2453110492107706,
      "grad_norm": 1.7923405461906572,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.868533194065094,
      "num_tokens": 233223347.0,
      "step": 6706
    },
    {
      "epoch": 1.2454967502321264,
      "grad_norm": 1.512110389439783,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8797187805175781,
      "num_tokens": 233258123.0,
      "step": 6707
    },
    {
      "epoch": 1.2456824512534819,
      "grad_norm": 1.4318553282103477,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8649015426635742,
      "num_tokens": 233299591.0,
      "step": 6708
    },
    {
      "epoch": 1.2458681522748376,
      "grad_norm": 1.4876777863118011,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.887771725654602,
      "num_tokens": 233333559.0,
      "step": 6709
    },
    {
      "epoch": 1.2460538532961931,
      "grad_norm": 1.4794936621632135,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.868317723274231,
      "num_tokens": 233371081.0,
      "step": 6710
    },
    {
      "epoch": 1.2462395543175488,
      "grad_norm": 1.6043208152807185,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8662406206130981,
      "num_tokens": 233403337.0,
      "step": 6711
    },
    {
      "epoch": 1.2464252553389044,
      "grad_norm": 1.4934593132135627,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8830163478851318,
      "num_tokens": 233436665.0,
      "step": 6712
    },
    {
      "epoch": 1.2466109563602599,
      "grad_norm": 1.512784569186309,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.873489260673523,
      "num_tokens": 233468681.0,
      "step": 6713
    },
    {
      "epoch": 1.2467966573816156,
      "grad_norm": 1.5195993483192,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8746123313903809,
      "num_tokens": 233505407.0,
      "step": 6714
    },
    {
      "epoch": 1.2469823584029713,
      "grad_norm": 1.6181078691689406,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8864821195602417,
      "num_tokens": 233537436.0,
      "step": 6715
    },
    {
      "epoch": 1.2471680594243268,
      "grad_norm": 1.528275096262093,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8781962394714355,
      "num_tokens": 233573420.0,
      "step": 6716
    },
    {
      "epoch": 1.2473537604456824,
      "grad_norm": 1.6933361088453123,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8764626383781433,
      "num_tokens": 233605822.0,
      "step": 6717
    },
    {
      "epoch": 1.247539461467038,
      "grad_norm": 1.2971466755690433,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8816319704055786,
      "num_tokens": 233650869.0,
      "step": 6718
    },
    {
      "epoch": 1.2477251624883936,
      "grad_norm": 1.441033195312957,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8715452551841736,
      "num_tokens": 233693967.0,
      "step": 6719
    },
    {
      "epoch": 1.2479108635097493,
      "grad_norm": 1.522013351636075,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.877281904220581,
      "num_tokens": 233725568.0,
      "step": 6720
    },
    {
      "epoch": 1.2480965645311048,
      "grad_norm": 1.4569905254082076,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8849024772644043,
      "num_tokens": 233762066.0,
      "step": 6721
    },
    {
      "epoch": 1.2482822655524606,
      "grad_norm": 1.4976016347227665,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8818069696426392,
      "num_tokens": 233796078.0,
      "step": 6722
    },
    {
      "epoch": 1.248467966573816,
      "grad_norm": 1.4751244151993892,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8763164281845093,
      "num_tokens": 233830815.0,
      "step": 6723
    },
    {
      "epoch": 1.2486536675951718,
      "grad_norm": 1.4471616209163376,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.87581467628479,
      "num_tokens": 233870833.0,
      "step": 6724
    },
    {
      "epoch": 1.2488393686165273,
      "grad_norm": 1.5906201322017524,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8716484308242798,
      "num_tokens": 233905655.0,
      "step": 6725
    },
    {
      "epoch": 1.249025069637883,
      "grad_norm": 1.529405211764855,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8866763114929199,
      "num_tokens": 233942112.0,
      "step": 6726
    },
    {
      "epoch": 1.2492107706592386,
      "grad_norm": 1.6539027927569077,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8843393325805664,
      "num_tokens": 233969009.0,
      "step": 6727
    },
    {
      "epoch": 1.2493964716805943,
      "grad_norm": 1.4553487855689797,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.897273063659668,
      "num_tokens": 234003933.0,
      "step": 6728
    },
    {
      "epoch": 1.2495821727019498,
      "grad_norm": 1.4565198396238974,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8714340925216675,
      "num_tokens": 234042764.0,
      "step": 6729
    },
    {
      "epoch": 1.2497678737233056,
      "grad_norm": 1.4192427284544995,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8693634271621704,
      "num_tokens": 234081307.0,
      "step": 6730
    },
    {
      "epoch": 1.249953574744661,
      "grad_norm": 1.5911076698770414,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8703048825263977,
      "num_tokens": 234118101.0,
      "step": 6731
    },
    {
      "epoch": 1.2501392757660166,
      "grad_norm": 1.5285111674968692,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.883324384689331,
      "num_tokens": 234153473.0,
      "step": 6732
    },
    {
      "epoch": 1.2503249767873723,
      "grad_norm": 1.4575236470031319,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8765040636062622,
      "num_tokens": 234193044.0,
      "step": 6733
    },
    {
      "epoch": 1.250510677808728,
      "grad_norm": 1.480611585928055,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8834747076034546,
      "num_tokens": 234228796.0,
      "step": 6734
    },
    {
      "epoch": 1.2506963788300836,
      "grad_norm": 1.5194330666582656,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8886010646820068,
      "num_tokens": 234261509.0,
      "step": 6735
    },
    {
      "epoch": 1.250882079851439,
      "grad_norm": 1.336335528128105,
      "learning_rate": 1e-06,
      "loss": 0.2753,
      "mean_token_accuracy": 0.9044710397720337,
      "num_tokens": 234296978.0,
      "step": 6736
    },
    {
      "epoch": 1.2510677808727948,
      "grad_norm": 1.447678349164314,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8820606470108032,
      "num_tokens": 234333161.0,
      "step": 6737
    },
    {
      "epoch": 1.2512534818941505,
      "grad_norm": 1.4157105805123746,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8669230937957764,
      "num_tokens": 234378850.0,
      "step": 6738
    },
    {
      "epoch": 1.251439182915506,
      "grad_norm": 1.4085701574534488,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8880791664123535,
      "num_tokens": 234416642.0,
      "step": 6739
    },
    {
      "epoch": 1.2516248839368616,
      "grad_norm": 1.3828124724512199,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8714954853057861,
      "num_tokens": 234456642.0,
      "step": 6740
    },
    {
      "epoch": 1.2518105849582173,
      "grad_norm": 1.451127653404441,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8661434650421143,
      "num_tokens": 234494119.0,
      "step": 6741
    },
    {
      "epoch": 1.251996285979573,
      "grad_norm": 1.4728637504059616,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8679748773574829,
      "num_tokens": 234531605.0,
      "step": 6742
    },
    {
      "epoch": 1.2521819870009285,
      "grad_norm": 1.5347902008029657,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8671851754188538,
      "num_tokens": 234567070.0,
      "step": 6743
    },
    {
      "epoch": 1.252367688022284,
      "grad_norm": 1.6314890395439525,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8688060641288757,
      "num_tokens": 234599214.0,
      "step": 6744
    },
    {
      "epoch": 1.2525533890436398,
      "grad_norm": 1.6321555274321997,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8714250326156616,
      "num_tokens": 234628880.0,
      "step": 6745
    },
    {
      "epoch": 1.2527390900649953,
      "grad_norm": 1.5695207060003433,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8719293475151062,
      "num_tokens": 234659452.0,
      "step": 6746
    },
    {
      "epoch": 1.252924791086351,
      "grad_norm": 1.500850320118017,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8628538250923157,
      "num_tokens": 234699146.0,
      "step": 6747
    },
    {
      "epoch": 1.2531104921077065,
      "grad_norm": 1.596888538147079,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8802340030670166,
      "num_tokens": 234732797.0,
      "step": 6748
    },
    {
      "epoch": 1.2532961931290623,
      "grad_norm": 1.5917357021870238,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8774394989013672,
      "num_tokens": 234770599.0,
      "step": 6749
    },
    {
      "epoch": 1.2534818941504178,
      "grad_norm": 1.6540806602771085,
      "learning_rate": 1e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8623470664024353,
      "num_tokens": 234800797.0,
      "step": 6750
    },
    {
      "epoch": 1.2536675951717735,
      "grad_norm": 1.6474776713535073,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8738514184951782,
      "num_tokens": 234833830.0,
      "step": 6751
    },
    {
      "epoch": 1.253853296193129,
      "grad_norm": 1.4842005302478127,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8724271059036255,
      "num_tokens": 234873069.0,
      "step": 6752
    },
    {
      "epoch": 1.2540389972144848,
      "grad_norm": 1.5695028448774129,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8737643957138062,
      "num_tokens": 234906658.0,
      "step": 6753
    },
    {
      "epoch": 1.2542246982358403,
      "grad_norm": 1.574758524717228,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8822894096374512,
      "num_tokens": 234938557.0,
      "step": 6754
    },
    {
      "epoch": 1.254410399257196,
      "grad_norm": 1.585130953021658,
      "learning_rate": 1e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8631321787834167,
      "num_tokens": 234972485.0,
      "step": 6755
    },
    {
      "epoch": 1.2545961002785515,
      "grad_norm": 1.484010937962904,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8635277152061462,
      "num_tokens": 235011446.0,
      "step": 6756
    },
    {
      "epoch": 1.2547818012999072,
      "grad_norm": 1.4395555016508335,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8815569877624512,
      "num_tokens": 235047325.0,
      "step": 6757
    },
    {
      "epoch": 1.2549675023212628,
      "grad_norm": 1.471157024967634,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8729772567749023,
      "num_tokens": 235081552.0,
      "step": 6758
    },
    {
      "epoch": 1.2551532033426183,
      "grad_norm": 1.5657226757772935,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8718760013580322,
      "num_tokens": 235114371.0,
      "step": 6759
    },
    {
      "epoch": 1.255338904363974,
      "grad_norm": 1.4749773227052718,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8835312128067017,
      "num_tokens": 235147593.0,
      "step": 6760
    },
    {
      "epoch": 1.2555246053853297,
      "grad_norm": 1.709929846110812,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8857002854347229,
      "num_tokens": 235174285.0,
      "step": 6761
    },
    {
      "epoch": 1.2557103064066852,
      "grad_norm": 1.460236276895938,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8776152729988098,
      "num_tokens": 235213427.0,
      "step": 6762
    },
    {
      "epoch": 1.2558960074280408,
      "grad_norm": 1.595619618819512,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8743727207183838,
      "num_tokens": 235244946.0,
      "step": 6763
    },
    {
      "epoch": 1.2560817084493965,
      "grad_norm": 1.5429737967764396,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8734819889068604,
      "num_tokens": 235280928.0,
      "step": 6764
    },
    {
      "epoch": 1.2562674094707522,
      "grad_norm": 1.3979701837588931,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8831201791763306,
      "num_tokens": 235321386.0,
      "step": 6765
    },
    {
      "epoch": 1.2564531104921077,
      "grad_norm": 1.3952589498842065,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8707462549209595,
      "num_tokens": 235365983.0,
      "step": 6766
    },
    {
      "epoch": 1.2566388115134632,
      "grad_norm": 1.5440478241745863,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8738020062446594,
      "num_tokens": 235401796.0,
      "step": 6767
    },
    {
      "epoch": 1.256824512534819,
      "grad_norm": 1.5627880518094397,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8700242042541504,
      "num_tokens": 235435154.0,
      "step": 6768
    },
    {
      "epoch": 1.2570102135561745,
      "grad_norm": 1.4135311998258209,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8800290822982788,
      "num_tokens": 235474404.0,
      "step": 6769
    },
    {
      "epoch": 1.2571959145775302,
      "grad_norm": 1.672019902669289,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8740520477294922,
      "num_tokens": 235503325.0,
      "step": 6770
    },
    {
      "epoch": 1.2573816155988857,
      "grad_norm": 1.5710386326674832,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8829890489578247,
      "num_tokens": 235531509.0,
      "step": 6771
    },
    {
      "epoch": 1.2575673166202415,
      "grad_norm": 2.0466511746457647,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8847463726997375,
      "num_tokens": 235566171.0,
      "step": 6772
    },
    {
      "epoch": 1.257753017641597,
      "grad_norm": 1.52060553846019,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8768224120140076,
      "num_tokens": 235598076.0,
      "step": 6773
    },
    {
      "epoch": 1.2579387186629527,
      "grad_norm": 1.4880571047670064,
      "learning_rate": 1e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.86662757396698,
      "num_tokens": 235635725.0,
      "step": 6774
    },
    {
      "epoch": 1.2581244196843082,
      "grad_norm": 1.5928331086923282,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8742338418960571,
      "num_tokens": 235668541.0,
      "step": 6775
    },
    {
      "epoch": 1.258310120705664,
      "grad_norm": 1.5905506965409064,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8665659427642822,
      "num_tokens": 235701942.0,
      "step": 6776
    },
    {
      "epoch": 1.2584958217270195,
      "grad_norm": 1.664539216967204,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8749693632125854,
      "num_tokens": 235733261.0,
      "step": 6777
    },
    {
      "epoch": 1.2586815227483752,
      "grad_norm": 1.3362526771157544,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8866208791732788,
      "num_tokens": 235770451.0,
      "step": 6778
    },
    {
      "epoch": 1.2588672237697307,
      "grad_norm": 1.5903977563323946,
      "learning_rate": 1e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8697004318237305,
      "num_tokens": 235802946.0,
      "step": 6779
    },
    {
      "epoch": 1.2590529247910864,
      "grad_norm": 1.3890355260240272,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8690484762191772,
      "num_tokens": 235844891.0,
      "step": 6780
    },
    {
      "epoch": 1.259238625812442,
      "grad_norm": 1.489072347192951,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8689553737640381,
      "num_tokens": 235883112.0,
      "step": 6781
    },
    {
      "epoch": 1.2594243268337975,
      "grad_norm": 1.7539877464691305,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8719143271446228,
      "num_tokens": 235913712.0,
      "step": 6782
    },
    {
      "epoch": 1.2596100278551532,
      "grad_norm": 1.5583068189996883,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8791695833206177,
      "num_tokens": 235947353.0,
      "step": 6783
    },
    {
      "epoch": 1.259795728876509,
      "grad_norm": 1.3867082874382208,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.876304030418396,
      "num_tokens": 235987060.0,
      "step": 6784
    },
    {
      "epoch": 1.2599814298978644,
      "grad_norm": 1.4256754492886334,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8718714714050293,
      "num_tokens": 236027123.0,
      "step": 6785
    },
    {
      "epoch": 1.26016713091922,
      "grad_norm": 1.3904577458517573,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8791891932487488,
      "num_tokens": 236065353.0,
      "step": 6786
    },
    {
      "epoch": 1.2603528319405757,
      "grad_norm": 1.4881903695368397,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8773801326751709,
      "num_tokens": 236106442.0,
      "step": 6787
    },
    {
      "epoch": 1.2605385329619314,
      "grad_norm": 1.5793302784994028,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8686660528182983,
      "num_tokens": 236141216.0,
      "step": 6788
    },
    {
      "epoch": 1.260724233983287,
      "grad_norm": 1.4868682942827445,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8756787776947021,
      "num_tokens": 236176369.0,
      "step": 6789
    },
    {
      "epoch": 1.2609099350046424,
      "grad_norm": 1.457449289198925,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8862957954406738,
      "num_tokens": 236213684.0,
      "step": 6790
    },
    {
      "epoch": 1.2610956360259982,
      "grad_norm": 1.7051475054408238,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8803183436393738,
      "num_tokens": 236249170.0,
      "step": 6791
    },
    {
      "epoch": 1.2612813370473537,
      "grad_norm": 1.4202377616449187,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8883581757545471,
      "num_tokens": 236285530.0,
      "step": 6792
    },
    {
      "epoch": 1.2614670380687094,
      "grad_norm": 1.390244032816344,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8829792737960815,
      "num_tokens": 236321982.0,
      "step": 6793
    },
    {
      "epoch": 1.261652739090065,
      "grad_norm": 1.5041701251872392,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8836085796356201,
      "num_tokens": 236356958.0,
      "step": 6794
    },
    {
      "epoch": 1.2618384401114207,
      "grad_norm": 1.5117117978830357,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8939347863197327,
      "num_tokens": 236391517.0,
      "step": 6795
    },
    {
      "epoch": 1.2620241411327762,
      "grad_norm": 1.5017064992949645,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8771172761917114,
      "num_tokens": 236424434.0,
      "step": 6796
    },
    {
      "epoch": 1.262209842154132,
      "grad_norm": 1.4025507451388886,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8852151036262512,
      "num_tokens": 236460077.0,
      "step": 6797
    },
    {
      "epoch": 1.2623955431754874,
      "grad_norm": 1.7303967686139243,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8695666790008545,
      "num_tokens": 236488640.0,
      "step": 6798
    },
    {
      "epoch": 1.2625812441968431,
      "grad_norm": 1.5962310667649815,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8741695880889893,
      "num_tokens": 236521044.0,
      "step": 6799
    },
    {
      "epoch": 1.2627669452181987,
      "grad_norm": 1.474356590074808,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8792263865470886,
      "num_tokens": 236559180.0,
      "step": 6800
    },
    {
      "epoch": 1.2629526462395544,
      "grad_norm": 1.5002817826354053,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8794465065002441,
      "num_tokens": 236593044.0,
      "step": 6801
    },
    {
      "epoch": 1.26313834726091,
      "grad_norm": 1.5438051437512883,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8759418725967407,
      "num_tokens": 236624958.0,
      "step": 6802
    },
    {
      "epoch": 1.2633240482822656,
      "grad_norm": 1.354978821143127,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8747210502624512,
      "num_tokens": 236663479.0,
      "step": 6803
    },
    {
      "epoch": 1.2635097493036211,
      "grad_norm": 1.4414345243385194,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8823065757751465,
      "num_tokens": 236697764.0,
      "step": 6804
    },
    {
      "epoch": 1.2636954503249767,
      "grad_norm": 1.5827175190818128,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8874338865280151,
      "num_tokens": 236728821.0,
      "step": 6805
    },
    {
      "epoch": 1.2638811513463324,
      "grad_norm": 1.563746469858122,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8715610504150391,
      "num_tokens": 236768418.0,
      "step": 6806
    },
    {
      "epoch": 1.2640668523676881,
      "grad_norm": 1.487272597621766,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8625555038452148,
      "num_tokens": 236805420.0,
      "step": 6807
    },
    {
      "epoch": 1.2642525533890436,
      "grad_norm": 1.426228929483828,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.884971022605896,
      "num_tokens": 236841206.0,
      "step": 6808
    },
    {
      "epoch": 1.2644382544103991,
      "grad_norm": 1.3874613879027236,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8893506526947021,
      "num_tokens": 236875555.0,
      "step": 6809
    },
    {
      "epoch": 1.2646239554317549,
      "grad_norm": 1.4447656477252446,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8803183436393738,
      "num_tokens": 236909537.0,
      "step": 6810
    },
    {
      "epoch": 1.2648096564531106,
      "grad_norm": 3.874881118462598,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8855195045471191,
      "num_tokens": 236946892.0,
      "step": 6811
    },
    {
      "epoch": 1.2649953574744661,
      "grad_norm": 1.535203441884022,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8864451050758362,
      "num_tokens": 236984300.0,
      "step": 6812
    },
    {
      "epoch": 1.2651810584958216,
      "grad_norm": 1.5522113376565982,
      "learning_rate": 1e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8586555123329163,
      "num_tokens": 237019882.0,
      "step": 6813
    },
    {
      "epoch": 1.2653667595171774,
      "grad_norm": 1.7113420243435058,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.879633903503418,
      "num_tokens": 237049787.0,
      "step": 6814
    },
    {
      "epoch": 1.265552460538533,
      "grad_norm": 1.466255235153282,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8880695104598999,
      "num_tokens": 237084583.0,
      "step": 6815
    },
    {
      "epoch": 1.2657381615598886,
      "grad_norm": 1.474337329585142,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8793497681617737,
      "num_tokens": 237122449.0,
      "step": 6816
    },
    {
      "epoch": 1.2659238625812441,
      "grad_norm": 1.524452699875859,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8763060569763184,
      "num_tokens": 237157791.0,
      "step": 6817
    },
    {
      "epoch": 1.2661095636025999,
      "grad_norm": 1.6023688512322414,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8787368535995483,
      "num_tokens": 237189013.0,
      "step": 6818
    },
    {
      "epoch": 1.2662952646239554,
      "grad_norm": 1.4553380676275567,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8830598592758179,
      "num_tokens": 237223683.0,
      "step": 6819
    },
    {
      "epoch": 1.266480965645311,
      "grad_norm": 1.4909317132826467,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8959002494812012,
      "num_tokens": 237255144.0,
      "step": 6820
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 1.4645633448409574,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8801670670509338,
      "num_tokens": 237289380.0,
      "step": 6821
    },
    {
      "epoch": 1.2668523676880223,
      "grad_norm": 1.564981542193191,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8659731149673462,
      "num_tokens": 237329528.0,
      "step": 6822
    },
    {
      "epoch": 1.2670380687093779,
      "grad_norm": 1.5707701582134443,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8645686507225037,
      "num_tokens": 237364056.0,
      "step": 6823
    },
    {
      "epoch": 1.2672237697307336,
      "grad_norm": 1.4638635649892466,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8710756897926331,
      "num_tokens": 237399755.0,
      "step": 6824
    },
    {
      "epoch": 1.267409470752089,
      "grad_norm": 1.5642604932479847,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8727055191993713,
      "num_tokens": 237433585.0,
      "step": 6825
    },
    {
      "epoch": 1.2675951717734448,
      "grad_norm": 1.5290176470815413,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8812929391860962,
      "num_tokens": 237466846.0,
      "step": 6826
    },
    {
      "epoch": 1.2677808727948003,
      "grad_norm": 1.571283513481602,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8728828430175781,
      "num_tokens": 237502105.0,
      "step": 6827
    },
    {
      "epoch": 1.2679665738161559,
      "grad_norm": 1.6059093238870183,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8642022609710693,
      "num_tokens": 237537573.0,
      "step": 6828
    },
    {
      "epoch": 1.2681522748375116,
      "grad_norm": 1.4489009771023638,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8799184560775757,
      "num_tokens": 237574955.0,
      "step": 6829
    },
    {
      "epoch": 1.2683379758588673,
      "grad_norm": 1.6353431063185007,
      "learning_rate": 1e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8619715571403503,
      "num_tokens": 237610221.0,
      "step": 6830
    },
    {
      "epoch": 1.2685236768802228,
      "grad_norm": 1.5959732797524042,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.879935622215271,
      "num_tokens": 237643241.0,
      "step": 6831
    },
    {
      "epoch": 1.2687093779015783,
      "grad_norm": 1.4950181582809483,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8811080455780029,
      "num_tokens": 237677670.0,
      "step": 6832
    },
    {
      "epoch": 1.268895078922934,
      "grad_norm": 1.5530247598136206,
      "learning_rate": 1e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8605595827102661,
      "num_tokens": 237712160.0,
      "step": 6833
    },
    {
      "epoch": 1.2690807799442898,
      "grad_norm": 1.4635734575437382,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8863844275474548,
      "num_tokens": 237745061.0,
      "step": 6834
    },
    {
      "epoch": 1.2692664809656453,
      "grad_norm": 1.533248710418875,
      "learning_rate": 1e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.8788473606109619,
      "num_tokens": 237779157.0,
      "step": 6835
    },
    {
      "epoch": 1.2694521819870008,
      "grad_norm": 1.4411109229086214,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.864566445350647,
      "num_tokens": 237816789.0,
      "step": 6836
    },
    {
      "epoch": 1.2696378830083566,
      "grad_norm": 1.5195838405271267,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8781477212905884,
      "num_tokens": 237849694.0,
      "step": 6837
    },
    {
      "epoch": 1.2698235840297123,
      "grad_norm": 1.5281510659154736,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8758910298347473,
      "num_tokens": 237884229.0,
      "step": 6838
    },
    {
      "epoch": 1.2700092850510678,
      "grad_norm": 1.4664860274709073,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8793776631355286,
      "num_tokens": 237924058.0,
      "step": 6839
    },
    {
      "epoch": 1.2701949860724233,
      "grad_norm": 1.4153768126979438,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8756921291351318,
      "num_tokens": 237963258.0,
      "step": 6840
    },
    {
      "epoch": 1.270380687093779,
      "grad_norm": 1.5429996962808994,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8814418911933899,
      "num_tokens": 237996978.0,
      "step": 6841
    },
    {
      "epoch": 1.2705663881151346,
      "grad_norm": 1.409423372875953,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8710033297538757,
      "num_tokens": 238037893.0,
      "step": 6842
    },
    {
      "epoch": 1.2707520891364903,
      "grad_norm": 1.6408188284455238,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8702473640441895,
      "num_tokens": 238069446.0,
      "step": 6843
    },
    {
      "epoch": 1.2709377901578458,
      "grad_norm": 1.650376391237746,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8611935377120972,
      "num_tokens": 238101156.0,
      "step": 6844
    },
    {
      "epoch": 1.2711234911792015,
      "grad_norm": 1.6038969482717533,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8828455209732056,
      "num_tokens": 238130168.0,
      "step": 6845
    },
    {
      "epoch": 1.271309192200557,
      "grad_norm": 1.4695665421079585,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8783642053604126,
      "num_tokens": 238168046.0,
      "step": 6846
    },
    {
      "epoch": 1.2714948932219128,
      "grad_norm": 1.4334015373109512,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8779589533805847,
      "num_tokens": 238205901.0,
      "step": 6847
    },
    {
      "epoch": 1.2716805942432683,
      "grad_norm": 1.3681335389973994,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8795677423477173,
      "num_tokens": 238251188.0,
      "step": 6848
    },
    {
      "epoch": 1.271866295264624,
      "grad_norm": 1.5620700449541918,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8670035600662231,
      "num_tokens": 238281239.0,
      "step": 6849
    },
    {
      "epoch": 1.2720519962859795,
      "grad_norm": 1.42428744416367,
      "learning_rate": 1e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8639127612113953,
      "num_tokens": 238322664.0,
      "step": 6850
    },
    {
      "epoch": 1.2722376973073353,
      "grad_norm": 1.4128691137535712,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8741530179977417,
      "num_tokens": 238362285.0,
      "step": 6851
    },
    {
      "epoch": 1.2724233983286908,
      "grad_norm": 1.3529720128485916,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8757572174072266,
      "num_tokens": 238401079.0,
      "step": 6852
    },
    {
      "epoch": 1.2726090993500465,
      "grad_norm": 1.3390851588557666,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8722785115242004,
      "num_tokens": 238446291.0,
      "step": 6853
    },
    {
      "epoch": 1.272794800371402,
      "grad_norm": 1.5276241570011693,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8693113923072815,
      "num_tokens": 238481249.0,
      "step": 6854
    },
    {
      "epoch": 1.2729805013927575,
      "grad_norm": 1.5134355142818383,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8872725963592529,
      "num_tokens": 238516844.0,
      "step": 6855
    },
    {
      "epoch": 1.2731662024141133,
      "grad_norm": 1.577310350919349,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8721198439598083,
      "num_tokens": 238551746.0,
      "step": 6856
    },
    {
      "epoch": 1.273351903435469,
      "grad_norm": 1.6002931298529264,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.867135226726532,
      "num_tokens": 238586054.0,
      "step": 6857
    },
    {
      "epoch": 1.2735376044568245,
      "grad_norm": 1.4318298551031399,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.872398853302002,
      "num_tokens": 238625651.0,
      "step": 6858
    },
    {
      "epoch": 1.27372330547818,
      "grad_norm": 1.387913714847521,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8993328809738159,
      "num_tokens": 238659760.0,
      "step": 6859
    },
    {
      "epoch": 1.2739090064995358,
      "grad_norm": 1.5032749190604064,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8748419880867004,
      "num_tokens": 238692676.0,
      "step": 6860
    },
    {
      "epoch": 1.2740947075208915,
      "grad_norm": 1.5226293105482558,
      "learning_rate": 1e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.874168336391449,
      "num_tokens": 238725625.0,
      "step": 6861
    },
    {
      "epoch": 1.274280408542247,
      "grad_norm": 1.636238658416987,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8753352165222168,
      "num_tokens": 238754245.0,
      "step": 6862
    },
    {
      "epoch": 1.2744661095636025,
      "grad_norm": 1.6074864086165082,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8761969804763794,
      "num_tokens": 238787498.0,
      "step": 6863
    },
    {
      "epoch": 1.2746518105849582,
      "grad_norm": 1.4873386163650073,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8722694516181946,
      "num_tokens": 238827907.0,
      "step": 6864
    },
    {
      "epoch": 1.2748375116063138,
      "grad_norm": 1.4647431538368492,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.883643627166748,
      "num_tokens": 238862378.0,
      "step": 6865
    },
    {
      "epoch": 1.2750232126276695,
      "grad_norm": 1.4522040941054866,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8748027682304382,
      "num_tokens": 238901465.0,
      "step": 6866
    },
    {
      "epoch": 1.275208913649025,
      "grad_norm": 1.650598718228631,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.87946617603302,
      "num_tokens": 238929635.0,
      "step": 6867
    },
    {
      "epoch": 1.2753946146703807,
      "grad_norm": 1.6547648384889482,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8752291202545166,
      "num_tokens": 238962507.0,
      "step": 6868
    },
    {
      "epoch": 1.2755803156917362,
      "grad_norm": 1.4647702228324821,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8884572982788086,
      "num_tokens": 238998295.0,
      "step": 6869
    },
    {
      "epoch": 1.275766016713092,
      "grad_norm": 1.5110614704459808,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8642877340316772,
      "num_tokens": 239034942.0,
      "step": 6870
    },
    {
      "epoch": 1.2759517177344475,
      "grad_norm": 1.5867499923784443,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.859136164188385,
      "num_tokens": 239070191.0,
      "step": 6871
    },
    {
      "epoch": 1.2761374187558032,
      "grad_norm": 1.424707218845227,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8853439688682556,
      "num_tokens": 239104325.0,
      "step": 6872
    },
    {
      "epoch": 1.2763231197771587,
      "grad_norm": 1.5453828927217603,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8868574500083923,
      "num_tokens": 239139255.0,
      "step": 6873
    },
    {
      "epoch": 1.2765088207985145,
      "grad_norm": 1.5033440230099162,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8682442903518677,
      "num_tokens": 239177156.0,
      "step": 6874
    },
    {
      "epoch": 1.27669452181987,
      "grad_norm": 1.4242542184160838,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8857917785644531,
      "num_tokens": 239216675.0,
      "step": 6875
    },
    {
      "epoch": 1.2768802228412257,
      "grad_norm": 1.5837271281870393,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8805043697357178,
      "num_tokens": 239248540.0,
      "step": 6876
    },
    {
      "epoch": 1.2770659238625812,
      "grad_norm": 1.5912921816209347,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8750997185707092,
      "num_tokens": 239276886.0,
      "step": 6877
    },
    {
      "epoch": 1.2772516248839367,
      "grad_norm": 1.5451651155792498,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8765674829483032,
      "num_tokens": 239308210.0,
      "step": 6878
    },
    {
      "epoch": 1.2774373259052925,
      "grad_norm": 1.6400849719494666,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.867461085319519,
      "num_tokens": 239341947.0,
      "step": 6879
    },
    {
      "epoch": 1.2776230269266482,
      "grad_norm": 1.4882032294232888,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8769584894180298,
      "num_tokens": 239378128.0,
      "step": 6880
    },
    {
      "epoch": 1.2778087279480037,
      "grad_norm": 1.6130414236252728,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8844766020774841,
      "num_tokens": 239408629.0,
      "step": 6881
    },
    {
      "epoch": 1.2779944289693592,
      "grad_norm": 1.4862248277894106,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8640962839126587,
      "num_tokens": 239445930.0,
      "step": 6882
    },
    {
      "epoch": 1.278180129990715,
      "grad_norm": 1.6475062773944875,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8730424642562866,
      "num_tokens": 239482335.0,
      "step": 6883
    },
    {
      "epoch": 1.2783658310120707,
      "grad_norm": 1.4929345791859994,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8771489858627319,
      "num_tokens": 239519820.0,
      "step": 6884
    },
    {
      "epoch": 1.2785515320334262,
      "grad_norm": 1.4859430251971006,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8652936220169067,
      "num_tokens": 239559760.0,
      "step": 6885
    },
    {
      "epoch": 1.2787372330547817,
      "grad_norm": 1.427558998838203,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.88019859790802,
      "num_tokens": 239597018.0,
      "step": 6886
    },
    {
      "epoch": 1.2789229340761374,
      "grad_norm": 1.4899084068005393,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8791938424110413,
      "num_tokens": 239633765.0,
      "step": 6887
    },
    {
      "epoch": 1.2791086350974932,
      "grad_norm": 1.6255426782335294,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8592207431793213,
      "num_tokens": 239669274.0,
      "step": 6888
    },
    {
      "epoch": 1.2792943361188487,
      "grad_norm": 1.477649453857887,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8777607083320618,
      "num_tokens": 239707675.0,
      "step": 6889
    },
    {
      "epoch": 1.2794800371402042,
      "grad_norm": 1.5236033640829052,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8772476315498352,
      "num_tokens": 239746870.0,
      "step": 6890
    },
    {
      "epoch": 1.27966573816156,
      "grad_norm": 1.75421940984978,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8799775838851929,
      "num_tokens": 239771909.0,
      "step": 6891
    },
    {
      "epoch": 1.2798514391829154,
      "grad_norm": 1.4891816243052505,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8911013603210449,
      "num_tokens": 239800793.0,
      "step": 6892
    },
    {
      "epoch": 1.2800371402042712,
      "grad_norm": 1.4238804810216041,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8824417591094971,
      "num_tokens": 239836006.0,
      "step": 6893
    },
    {
      "epoch": 1.2802228412256267,
      "grad_norm": 1.5767579943817724,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8883801698684692,
      "num_tokens": 239864668.0,
      "step": 6894
    },
    {
      "epoch": 1.2804085422469824,
      "grad_norm": 1.6812759017074832,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.866624116897583,
      "num_tokens": 239893445.0,
      "step": 6895
    },
    {
      "epoch": 1.280594243268338,
      "grad_norm": 1.4669270065471636,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8711584210395813,
      "num_tokens": 239929806.0,
      "step": 6896
    },
    {
      "epoch": 1.2807799442896937,
      "grad_norm": 1.6871370950357916,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8794376850128174,
      "num_tokens": 239960535.0,
      "step": 6897
    },
    {
      "epoch": 1.2809656453110492,
      "grad_norm": 1.5209249292841667,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8760225772857666,
      "num_tokens": 239993028.0,
      "step": 6898
    },
    {
      "epoch": 1.281151346332405,
      "grad_norm": 1.505606254278572,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8657104969024658,
      "num_tokens": 240035412.0,
      "step": 6899
    },
    {
      "epoch": 1.2813370473537604,
      "grad_norm": 1.553175597672042,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8602227568626404,
      "num_tokens": 240072698.0,
      "step": 6900
    },
    {
      "epoch": 1.281522748375116,
      "grad_norm": 1.3393896026913699,
      "learning_rate": 1e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8685622215270996,
      "num_tokens": 240120309.0,
      "step": 6901
    },
    {
      "epoch": 1.2817084493964717,
      "grad_norm": 1.5641181442757566,
      "learning_rate": 1e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8613215684890747,
      "num_tokens": 240158204.0,
      "step": 6902
    },
    {
      "epoch": 1.2818941504178274,
      "grad_norm": 1.393185094060457,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8879480361938477,
      "num_tokens": 240194177.0,
      "step": 6903
    },
    {
      "epoch": 1.282079851439183,
      "grad_norm": 1.4542628329367666,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.879859209060669,
      "num_tokens": 240231608.0,
      "step": 6904
    },
    {
      "epoch": 1.2822655524605384,
      "grad_norm": 1.5010301311159333,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8853899240493774,
      "num_tokens": 240266360.0,
      "step": 6905
    },
    {
      "epoch": 1.2824512534818941,
      "grad_norm": 1.5526987721800476,
      "learning_rate": 1e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8667764663696289,
      "num_tokens": 240301466.0,
      "step": 6906
    },
    {
      "epoch": 1.2826369545032499,
      "grad_norm": 1.4413837676330508,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8815736770629883,
      "num_tokens": 240340616.0,
      "step": 6907
    },
    {
      "epoch": 1.2828226555246054,
      "grad_norm": 1.7657646384770462,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8802061080932617,
      "num_tokens": 240363141.0,
      "step": 6908
    },
    {
      "epoch": 1.283008356545961,
      "grad_norm": 1.5731703476523469,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8824641108512878,
      "num_tokens": 240394672.0,
      "step": 6909
    },
    {
      "epoch": 1.2831940575673166,
      "grad_norm": 1.5925659313530223,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8657963275909424,
      "num_tokens": 240428981.0,
      "step": 6910
    },
    {
      "epoch": 1.2833797585886724,
      "grad_norm": 1.4014057915957097,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8643988966941833,
      "num_tokens": 240469421.0,
      "step": 6911
    },
    {
      "epoch": 1.2835654596100279,
      "grad_norm": 1.4605315191846426,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8852756023406982,
      "num_tokens": 240501309.0,
      "step": 6912
    },
    {
      "epoch": 1.2837511606313834,
      "grad_norm": 1.4666850872531527,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8888587951660156,
      "num_tokens": 240538132.0,
      "step": 6913
    },
    {
      "epoch": 1.2839368616527391,
      "grad_norm": 1.5358446129287848,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.868811845779419,
      "num_tokens": 240574441.0,
      "step": 6914
    },
    {
      "epoch": 1.2841225626740946,
      "grad_norm": 1.657101699201701,
      "learning_rate": 1e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8687227964401245,
      "num_tokens": 240605914.0,
      "step": 6915
    },
    {
      "epoch": 1.2843082636954504,
      "grad_norm": 1.5114671446319596,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8810293674468994,
      "num_tokens": 240637991.0,
      "step": 6916
    },
    {
      "epoch": 1.2844939647168059,
      "grad_norm": 1.3893020108662761,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.87868332862854,
      "num_tokens": 240673964.0,
      "step": 6917
    },
    {
      "epoch": 1.2846796657381616,
      "grad_norm": 1.4833632274067698,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8715095520019531,
      "num_tokens": 240713246.0,
      "step": 6918
    },
    {
      "epoch": 1.2848653667595171,
      "grad_norm": 1.321975083567243,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8904283046722412,
      "num_tokens": 240754480.0,
      "step": 6919
    },
    {
      "epoch": 1.2850510677808729,
      "grad_norm": 1.6292577965136759,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8797577023506165,
      "num_tokens": 240781188.0,
      "step": 6920
    },
    {
      "epoch": 1.2852367688022284,
      "grad_norm": 1.4901320499252668,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8694167137145996,
      "num_tokens": 240814367.0,
      "step": 6921
    },
    {
      "epoch": 1.285422469823584,
      "grad_norm": 1.400956856543474,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8873895406723022,
      "num_tokens": 240848175.0,
      "step": 6922
    },
    {
      "epoch": 1.2856081708449396,
      "grad_norm": 1.5628941406916408,
      "learning_rate": 1e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8700054883956909,
      "num_tokens": 240879235.0,
      "step": 6923
    },
    {
      "epoch": 1.2857938718662953,
      "grad_norm": 1.584463027957126,
      "learning_rate": 1e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8571302890777588,
      "num_tokens": 240914312.0,
      "step": 6924
    },
    {
      "epoch": 1.2859795728876509,
      "grad_norm": 1.5185082679744337,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8735126256942749,
      "num_tokens": 240947140.0,
      "step": 6925
    },
    {
      "epoch": 1.2861652739090066,
      "grad_norm": 1.4828910668519701,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8844851851463318,
      "num_tokens": 240981705.0,
      "step": 6926
    },
    {
      "epoch": 1.286350974930362,
      "grad_norm": 1.5056415507641625,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8681379556655884,
      "num_tokens": 241016548.0,
      "step": 6927
    },
    {
      "epoch": 1.2865366759517176,
      "grad_norm": 1.4096245560075915,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8782594203948975,
      "num_tokens": 241053852.0,
      "step": 6928
    },
    {
      "epoch": 1.2867223769730733,
      "grad_norm": 1.5247150263807479,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8761809468269348,
      "num_tokens": 241087302.0,
      "step": 6929
    },
    {
      "epoch": 1.286908077994429,
      "grad_norm": 1.504130495272317,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8759201765060425,
      "num_tokens": 241120539.0,
      "step": 6930
    },
    {
      "epoch": 1.2870937790157846,
      "grad_norm": 1.4319610356438948,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8855803608894348,
      "num_tokens": 241155079.0,
      "step": 6931
    },
    {
      "epoch": 1.28727948003714,
      "grad_norm": 1.390698694482093,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8849248886108398,
      "num_tokens": 241197056.0,
      "step": 6932
    },
    {
      "epoch": 1.2874651810584958,
      "grad_norm": 1.4980682910367822,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.876494288444519,
      "num_tokens": 241233224.0,
      "step": 6933
    },
    {
      "epoch": 1.2876508820798516,
      "grad_norm": 1.4709185779861689,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8702096939086914,
      "num_tokens": 241269548.0,
      "step": 6934
    },
    {
      "epoch": 1.287836583101207,
      "grad_norm": 1.438616586453991,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8894214034080505,
      "num_tokens": 241303170.0,
      "step": 6935
    },
    {
      "epoch": 1.2880222841225626,
      "grad_norm": 1.6243419803320474,
      "learning_rate": 1e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.86446213722229,
      "num_tokens": 241335986.0,
      "step": 6936
    },
    {
      "epoch": 1.2882079851439183,
      "grad_norm": 2.2812922666699165,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8794088363647461,
      "num_tokens": 241373374.0,
      "step": 6937
    },
    {
      "epoch": 1.2883936861652738,
      "grad_norm": 1.8148562744868948,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8743166923522949,
      "num_tokens": 241403916.0,
      "step": 6938
    },
    {
      "epoch": 1.2885793871866296,
      "grad_norm": 1.5958062036784644,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8657700419425964,
      "num_tokens": 241437802.0,
      "step": 6939
    },
    {
      "epoch": 1.288765088207985,
      "grad_norm": 1.4822279213856133,
      "learning_rate": 1e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8561831116676331,
      "num_tokens": 241479501.0,
      "step": 6940
    },
    {
      "epoch": 1.2889507892293408,
      "grad_norm": 1.4503165313192925,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8725869655609131,
      "num_tokens": 241516832.0,
      "step": 6941
    },
    {
      "epoch": 1.2891364902506963,
      "grad_norm": 1.4832245678704326,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8823468089103699,
      "num_tokens": 241553636.0,
      "step": 6942
    },
    {
      "epoch": 1.289322191272052,
      "grad_norm": 1.4434345184059623,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8737250566482544,
      "num_tokens": 241591185.0,
      "step": 6943
    },
    {
      "epoch": 1.2895078922934076,
      "grad_norm": 1.6105252530695504,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8771383762359619,
      "num_tokens": 241624960.0,
      "step": 6944
    },
    {
      "epoch": 1.2896935933147633,
      "grad_norm": 1.5542202433185264,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8847329616546631,
      "num_tokens": 241656330.0,
      "step": 6945
    },
    {
      "epoch": 1.2898792943361188,
      "grad_norm": 1.6530120967666373,
      "learning_rate": 1e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8553977608680725,
      "num_tokens": 241687397.0,
      "step": 6946
    },
    {
      "epoch": 1.2900649953574745,
      "grad_norm": 1.4127606562096704,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8663965463638306,
      "num_tokens": 241729231.0,
      "step": 6947
    },
    {
      "epoch": 1.29025069637883,
      "grad_norm": 1.5456846579607366,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8852521181106567,
      "num_tokens": 241759612.0,
      "step": 6948
    },
    {
      "epoch": 1.2904363974001858,
      "grad_norm": 1.467724597516012,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8917505145072937,
      "num_tokens": 241794255.0,
      "step": 6949
    },
    {
      "epoch": 1.2906220984215413,
      "grad_norm": 1.5126790233845113,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8814820051193237,
      "num_tokens": 241824331.0,
      "step": 6950
    },
    {
      "epoch": 1.2908077994428968,
      "grad_norm": 1.4896040163585254,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8731944561004639,
      "num_tokens": 241861278.0,
      "step": 6951
    },
    {
      "epoch": 1.2909935004642525,
      "grad_norm": 1.3995077936159264,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8744811415672302,
      "num_tokens": 241899831.0,
      "step": 6952
    },
    {
      "epoch": 1.2911792014856083,
      "grad_norm": 1.419559073897857,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8713111877441406,
      "num_tokens": 241938853.0,
      "step": 6953
    },
    {
      "epoch": 1.2913649025069638,
      "grad_norm": 1.6213978419082171,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8813151717185974,
      "num_tokens": 241967873.0,
      "step": 6954
    },
    {
      "epoch": 1.2915506035283193,
      "grad_norm": 1.569844074907719,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8886722326278687,
      "num_tokens": 242003493.0,
      "step": 6955
    },
    {
      "epoch": 1.291736304549675,
      "grad_norm": 1.4205368209384723,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8771227598190308,
      "num_tokens": 242039352.0,
      "step": 6956
    },
    {
      "epoch": 1.2919220055710308,
      "grad_norm": 1.6272805886638335,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8718322515487671,
      "num_tokens": 242068680.0,
      "step": 6957
    },
    {
      "epoch": 1.2921077065923863,
      "grad_norm": 1.4380453503487398,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8661298155784607,
      "num_tokens": 242109244.0,
      "step": 6958
    },
    {
      "epoch": 1.2922934076137418,
      "grad_norm": 1.572102468830696,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8729208707809448,
      "num_tokens": 242138812.0,
      "step": 6959
    },
    {
      "epoch": 1.2924791086350975,
      "grad_norm": 1.5502855461983303,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8798018097877502,
      "num_tokens": 242170450.0,
      "step": 6960
    },
    {
      "epoch": 1.292664809656453,
      "grad_norm": 1.5186741327394175,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8843353986740112,
      "num_tokens": 242202721.0,
      "step": 6961
    },
    {
      "epoch": 1.2928505106778088,
      "grad_norm": 1.4468959475143346,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8707456588745117,
      "num_tokens": 242243545.0,
      "step": 6962
    },
    {
      "epoch": 1.2930362116991643,
      "grad_norm": 1.5307043070062465,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8875699043273926,
      "num_tokens": 242274534.0,
      "step": 6963
    },
    {
      "epoch": 1.29322191272052,
      "grad_norm": 1.5490179804633357,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8792969584465027,
      "num_tokens": 242308301.0,
      "step": 6964
    },
    {
      "epoch": 1.2934076137418755,
      "grad_norm": 1.5040538513567117,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8713952302932739,
      "num_tokens": 242346117.0,
      "step": 6965
    },
    {
      "epoch": 1.2935933147632313,
      "grad_norm": 1.7291456988858285,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8638094663619995,
      "num_tokens": 242374764.0,
      "step": 6966
    },
    {
      "epoch": 1.2937790157845868,
      "grad_norm": 1.425647734549748,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8778718709945679,
      "num_tokens": 242410664.0,
      "step": 6967
    },
    {
      "epoch": 1.2939647168059425,
      "grad_norm": 1.4628041434535473,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8652418851852417,
      "num_tokens": 242447057.0,
      "step": 6968
    },
    {
      "epoch": 1.294150417827298,
      "grad_norm": 1.3263966389229145,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8875926733016968,
      "num_tokens": 242481062.0,
      "step": 6969
    },
    {
      "epoch": 1.2943361188486537,
      "grad_norm": 1.4376426749834275,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8735288977622986,
      "num_tokens": 242520644.0,
      "step": 6970
    },
    {
      "epoch": 1.2945218198700092,
      "grad_norm": 1.4642032864886183,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8922391533851624,
      "num_tokens": 242554456.0,
      "step": 6971
    },
    {
      "epoch": 1.294707520891365,
      "grad_norm": 1.4898807531861658,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8744823932647705,
      "num_tokens": 242589379.0,
      "step": 6972
    },
    {
      "epoch": 1.2948932219127205,
      "grad_norm": 1.564423677452804,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8840464949607849,
      "num_tokens": 242618699.0,
      "step": 6973
    },
    {
      "epoch": 1.295078922934076,
      "grad_norm": 1.6799079747451893,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8736286163330078,
      "num_tokens": 242645506.0,
      "step": 6974
    },
    {
      "epoch": 1.2952646239554317,
      "grad_norm": 1.4273611669994006,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8812592029571533,
      "num_tokens": 242681740.0,
      "step": 6975
    },
    {
      "epoch": 1.2954503249767875,
      "grad_norm": 1.4705967353137792,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8782385587692261,
      "num_tokens": 242716912.0,
      "step": 6976
    },
    {
      "epoch": 1.295636025998143,
      "grad_norm": 1.581258482862635,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.875146746635437,
      "num_tokens": 242750851.0,
      "step": 6977
    },
    {
      "epoch": 1.2958217270194985,
      "grad_norm": 1.5394803852786991,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8638532757759094,
      "num_tokens": 242786515.0,
      "step": 6978
    },
    {
      "epoch": 1.2960074280408542,
      "grad_norm": 1.7022223729307384,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.881793737411499,
      "num_tokens": 242813262.0,
      "step": 6979
    },
    {
      "epoch": 1.29619312906221,
      "grad_norm": 1.543097012045537,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8724033832550049,
      "num_tokens": 242846189.0,
      "step": 6980
    },
    {
      "epoch": 1.2963788300835655,
      "grad_norm": 1.6691584815392446,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8734560012817383,
      "num_tokens": 242876832.0,
      "step": 6981
    },
    {
      "epoch": 1.296564531104921,
      "grad_norm": 1.3604486605298522,
      "learning_rate": 1e-06,
      "loss": 0.2887,
      "mean_token_accuracy": 0.8968892693519592,
      "num_tokens": 242911693.0,
      "step": 6982
    },
    {
      "epoch": 1.2967502321262767,
      "grad_norm": 1.5014117294882623,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8773757219314575,
      "num_tokens": 242949113.0,
      "step": 6983
    },
    {
      "epoch": 1.2969359331476324,
      "grad_norm": 1.6666452510657133,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8834086656570435,
      "num_tokens": 242976330.0,
      "step": 6984
    },
    {
      "epoch": 1.297121634168988,
      "grad_norm": 1.5415686464862803,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8669753074645996,
      "num_tokens": 243013190.0,
      "step": 6985
    },
    {
      "epoch": 1.2973073351903435,
      "grad_norm": 1.5918024058669937,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8729456067085266,
      "num_tokens": 243045445.0,
      "step": 6986
    },
    {
      "epoch": 1.2974930362116992,
      "grad_norm": 1.4895485084080737,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8662750124931335,
      "num_tokens": 243083099.0,
      "step": 6987
    },
    {
      "epoch": 1.2976787372330547,
      "grad_norm": 1.4928462768554283,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8820333480834961,
      "num_tokens": 243115593.0,
      "step": 6988
    },
    {
      "epoch": 1.2978644382544104,
      "grad_norm": 1.5237470523535788,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8790020942687988,
      "num_tokens": 243151169.0,
      "step": 6989
    },
    {
      "epoch": 1.298050139275766,
      "grad_norm": 1.3275720766355579,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.8991873264312744,
      "num_tokens": 243187968.0,
      "step": 6990
    },
    {
      "epoch": 1.2982358402971217,
      "grad_norm": 1.8259034164607362,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8820726871490479,
      "num_tokens": 243214476.0,
      "step": 6991
    },
    {
      "epoch": 1.2984215413184772,
      "grad_norm": 1.58072262849414,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8821275234222412,
      "num_tokens": 243246119.0,
      "step": 6992
    },
    {
      "epoch": 1.298607242339833,
      "grad_norm": 1.662820332640818,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8812215328216553,
      "num_tokens": 243275847.0,
      "step": 6993
    },
    {
      "epoch": 1.2987929433611884,
      "grad_norm": 1.4681833205791792,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8889479637145996,
      "num_tokens": 243312125.0,
      "step": 6994
    },
    {
      "epoch": 1.2989786443825442,
      "grad_norm": 1.4484523384402306,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8816410899162292,
      "num_tokens": 243349288.0,
      "step": 6995
    },
    {
      "epoch": 1.2991643454038997,
      "grad_norm": 1.5158700237069462,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.86552894115448,
      "num_tokens": 243386088.0,
      "step": 6996
    },
    {
      "epoch": 1.2993500464252552,
      "grad_norm": 1.467279286454959,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8698515295982361,
      "num_tokens": 243424923.0,
      "step": 6997
    },
    {
      "epoch": 1.299535747446611,
      "grad_norm": 1.6261370296743867,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8660474419593811,
      "num_tokens": 243454589.0,
      "step": 6998
    },
    {
      "epoch": 1.2997214484679667,
      "grad_norm": 1.5687994796928744,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8830559849739075,
      "num_tokens": 243487951.0,
      "step": 6999
    },
    {
      "epoch": 1.2999071494893222,
      "grad_norm": 1.3256585249370227,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8729221820831299,
      "num_tokens": 243533056.0,
      "step": 7000
    },
    {
      "epoch": 1.3000928505106777,
      "grad_norm": 1.6127233373505723,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8778185844421387,
      "num_tokens": 243568127.0,
      "step": 7001
    },
    {
      "epoch": 1.3002785515320334,
      "grad_norm": 1.3930694863385475,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8884550333023071,
      "num_tokens": 243604612.0,
      "step": 7002
    },
    {
      "epoch": 1.3004642525533892,
      "grad_norm": 1.4259483519392249,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8763297200202942,
      "num_tokens": 243643598.0,
      "step": 7003
    },
    {
      "epoch": 1.3006499535747447,
      "grad_norm": 1.4164165393821457,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8771375417709351,
      "num_tokens": 243683264.0,
      "step": 7004
    },
    {
      "epoch": 1.3008356545961002,
      "grad_norm": 1.4600198086306613,
      "learning_rate": 1e-06,
      "loss": 0.3826,
      "mean_token_accuracy": 0.865121066570282,
      "num_tokens": 243721240.0,
      "step": 7005
    },
    {
      "epoch": 1.301021355617456,
      "grad_norm": 1.675690501020399,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.874442994594574,
      "num_tokens": 243750881.0,
      "step": 7006
    },
    {
      "epoch": 1.3012070566388116,
      "grad_norm": 1.4699375170776614,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8860481977462769,
      "num_tokens": 243784885.0,
      "step": 7007
    },
    {
      "epoch": 1.3013927576601672,
      "grad_norm": 1.48130641784867,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8756069540977478,
      "num_tokens": 243821790.0,
      "step": 7008
    },
    {
      "epoch": 1.3015784586815227,
      "grad_norm": 1.4545065249096394,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8806214928627014,
      "num_tokens": 243857085.0,
      "step": 7009
    },
    {
      "epoch": 1.3017641597028784,
      "grad_norm": 1.5802608879903237,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.874496579170227,
      "num_tokens": 243887654.0,
      "step": 7010
    },
    {
      "epoch": 1.301949860724234,
      "grad_norm": 1.591768259610262,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8685764074325562,
      "num_tokens": 243922873.0,
      "step": 7011
    },
    {
      "epoch": 1.3021355617455896,
      "grad_norm": 1.3453784882362292,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8799164295196533,
      "num_tokens": 243964466.0,
      "step": 7012
    },
    {
      "epoch": 1.3023212627669452,
      "grad_norm": 1.4517477433253143,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8858587741851807,
      "num_tokens": 243999203.0,
      "step": 7013
    },
    {
      "epoch": 1.3025069637883009,
      "grad_norm": 1.6488719315101499,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.876943051815033,
      "num_tokens": 244029897.0,
      "step": 7014
    },
    {
      "epoch": 1.3026926648096564,
      "grad_norm": 1.6023306144525473,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8727691769599915,
      "num_tokens": 244063863.0,
      "step": 7015
    },
    {
      "epoch": 1.3028783658310121,
      "grad_norm": 1.616296361206982,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8719967603683472,
      "num_tokens": 244097247.0,
      "step": 7016
    },
    {
      "epoch": 1.3030640668523676,
      "grad_norm": 1.470477626376945,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8835208415985107,
      "num_tokens": 244131750.0,
      "step": 7017
    },
    {
      "epoch": 1.3032497678737234,
      "grad_norm": 1.6117423231626524,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8755065202713013,
      "num_tokens": 244162882.0,
      "step": 7018
    },
    {
      "epoch": 1.3034354688950789,
      "grad_norm": 1.4205999542176353,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8803764581680298,
      "num_tokens": 244200441.0,
      "step": 7019
    },
    {
      "epoch": 1.3036211699164346,
      "grad_norm": 1.5951275285046593,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8653386235237122,
      "num_tokens": 244236345.0,
      "step": 7020
    },
    {
      "epoch": 1.3038068709377901,
      "grad_norm": 1.5185428629002873,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8784350752830505,
      "num_tokens": 244272960.0,
      "step": 7021
    },
    {
      "epoch": 1.3039925719591459,
      "grad_norm": 1.5350440692603187,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8899188041687012,
      "num_tokens": 244307241.0,
      "step": 7022
    },
    {
      "epoch": 1.3041782729805014,
      "grad_norm": 1.3887559305913724,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8851218223571777,
      "num_tokens": 244344977.0,
      "step": 7023
    },
    {
      "epoch": 1.3043639740018569,
      "grad_norm": 1.4866564900892663,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8807588219642639,
      "num_tokens": 244379103.0,
      "step": 7024
    },
    {
      "epoch": 1.3045496750232126,
      "grad_norm": 1.4699883671779512,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8768252730369568,
      "num_tokens": 244417773.0,
      "step": 7025
    },
    {
      "epoch": 1.3047353760445684,
      "grad_norm": 1.5097459103722635,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8714255094528198,
      "num_tokens": 244451684.0,
      "step": 7026
    },
    {
      "epoch": 1.3049210770659239,
      "grad_norm": 1.5101339383369126,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8760405778884888,
      "num_tokens": 244482427.0,
      "step": 7027
    },
    {
      "epoch": 1.3051067780872794,
      "grad_norm": 1.6435712512528042,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.8995238542556763,
      "num_tokens": 244512678.0,
      "step": 7028
    },
    {
      "epoch": 1.305292479108635,
      "grad_norm": 1.5606617226501986,
      "learning_rate": 1e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8608955144882202,
      "num_tokens": 244550369.0,
      "step": 7029
    },
    {
      "epoch": 1.3054781801299908,
      "grad_norm": 1.5382728621013713,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8707771897315979,
      "num_tokens": 244589493.0,
      "step": 7030
    },
    {
      "epoch": 1.3056638811513464,
      "grad_norm": 1.4551857563286303,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8759164214134216,
      "num_tokens": 244624786.0,
      "step": 7031
    },
    {
      "epoch": 1.3058495821727019,
      "grad_norm": 1.5435081761906517,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8698586225509644,
      "num_tokens": 244657808.0,
      "step": 7032
    },
    {
      "epoch": 1.3060352831940576,
      "grad_norm": 1.550891701627682,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8711976408958435,
      "num_tokens": 244689066.0,
      "step": 7033
    },
    {
      "epoch": 1.306220984215413,
      "grad_norm": 1.3914988368541406,
      "learning_rate": 1e-06,
      "loss": 0.277,
      "mean_token_accuracy": 0.9021258354187012,
      "num_tokens": 244723260.0,
      "step": 7034
    },
    {
      "epoch": 1.3064066852367688,
      "grad_norm": 1.5473255181408925,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8830069303512573,
      "num_tokens": 244754870.0,
      "step": 7035
    },
    {
      "epoch": 1.3065923862581243,
      "grad_norm": 1.5729943748291901,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8752506971359253,
      "num_tokens": 244789779.0,
      "step": 7036
    },
    {
      "epoch": 1.30677808727948,
      "grad_norm": 1.6811718728085612,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8760135769844055,
      "num_tokens": 244817701.0,
      "step": 7037
    },
    {
      "epoch": 1.3069637883008356,
      "grad_norm": 1.6094044803591925,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8785253763198853,
      "num_tokens": 244847545.0,
      "step": 7038
    },
    {
      "epoch": 1.3071494893221913,
      "grad_norm": 1.5969113366718035,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8839967250823975,
      "num_tokens": 244877313.0,
      "step": 7039
    },
    {
      "epoch": 1.3073351903435468,
      "grad_norm": 1.6001017654703122,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.873023271560669,
      "num_tokens": 244906488.0,
      "step": 7040
    },
    {
      "epoch": 1.3075208913649026,
      "grad_norm": 1.6684998615883617,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8754341006278992,
      "num_tokens": 244934422.0,
      "step": 7041
    },
    {
      "epoch": 1.307706592386258,
      "grad_norm": 1.4153556089568973,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8603560328483582,
      "num_tokens": 244977446.0,
      "step": 7042
    },
    {
      "epoch": 1.3078922934076138,
      "grad_norm": 1.618899380272539,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8798034191131592,
      "num_tokens": 245004974.0,
      "step": 7043
    },
    {
      "epoch": 1.3080779944289693,
      "grad_norm": 1.504398698835633,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8687570095062256,
      "num_tokens": 245039997.0,
      "step": 7044
    },
    {
      "epoch": 1.308263695450325,
      "grad_norm": 1.4263191906460273,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.881333589553833,
      "num_tokens": 245076072.0,
      "step": 7045
    },
    {
      "epoch": 1.3084493964716806,
      "grad_norm": 1.4525770768327075,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8791569471359253,
      "num_tokens": 245111462.0,
      "step": 7046
    },
    {
      "epoch": 1.308635097493036,
      "grad_norm": 1.4246818120310116,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8862179517745972,
      "num_tokens": 245146067.0,
      "step": 7047
    },
    {
      "epoch": 1.3088207985143918,
      "grad_norm": 1.5418389145194829,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8703646659851074,
      "num_tokens": 245183143.0,
      "step": 7048
    },
    {
      "epoch": 1.3090064995357475,
      "grad_norm": 1.649397306042619,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8791391253471375,
      "num_tokens": 245216003.0,
      "step": 7049
    },
    {
      "epoch": 1.309192200557103,
      "grad_norm": 1.4824113272611148,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8714708685874939,
      "num_tokens": 245254278.0,
      "step": 7050
    },
    {
      "epoch": 1.3093779015784586,
      "grad_norm": 1.7161784636893846,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8793090581893921,
      "num_tokens": 245279314.0,
      "step": 7051
    },
    {
      "epoch": 1.3095636025998143,
      "grad_norm": 1.5487990857402874,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8846840858459473,
      "num_tokens": 245313533.0,
      "step": 7052
    },
    {
      "epoch": 1.30974930362117,
      "grad_norm": 1.625666215036356,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8624605536460876,
      "num_tokens": 245345331.0,
      "step": 7053
    },
    {
      "epoch": 1.3099350046425255,
      "grad_norm": 1.4864273930168244,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8798840045928955,
      "num_tokens": 245381033.0,
      "step": 7054
    },
    {
      "epoch": 1.310120705663881,
      "grad_norm": 1.5491935348256636,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8700045347213745,
      "num_tokens": 245414511.0,
      "step": 7055
    },
    {
      "epoch": 1.3103064066852368,
      "grad_norm": 1.525482860020795,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8744297027587891,
      "num_tokens": 245449674.0,
      "step": 7056
    },
    {
      "epoch": 1.3104921077065925,
      "grad_norm": 1.5190691525618858,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.873278796672821,
      "num_tokens": 245483367.0,
      "step": 7057
    },
    {
      "epoch": 1.310677808727948,
      "grad_norm": 1.5256734765592108,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8770503401756287,
      "num_tokens": 245517125.0,
      "step": 7058
    },
    {
      "epoch": 1.3108635097493035,
      "grad_norm": 1.4910057673205763,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8764683604240417,
      "num_tokens": 245556541.0,
      "step": 7059
    },
    {
      "epoch": 1.3110492107706593,
      "grad_norm": 1.4600332516254193,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8782154321670532,
      "num_tokens": 245593214.0,
      "step": 7060
    },
    {
      "epoch": 1.3112349117920148,
      "grad_norm": 1.5391591462302434,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8763119578361511,
      "num_tokens": 245625136.0,
      "step": 7061
    },
    {
      "epoch": 1.3114206128133705,
      "grad_norm": 1.4168475307868185,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8735637068748474,
      "num_tokens": 245667775.0,
      "step": 7062
    },
    {
      "epoch": 1.311606313834726,
      "grad_norm": 1.4306997835277036,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8793555498123169,
      "num_tokens": 245707489.0,
      "step": 7063
    },
    {
      "epoch": 1.3117920148560818,
      "grad_norm": 1.620782056866007,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.865077018737793,
      "num_tokens": 245738454.0,
      "step": 7064
    },
    {
      "epoch": 1.3119777158774373,
      "grad_norm": 1.4027044063395817,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8852721452713013,
      "num_tokens": 245776331.0,
      "step": 7065
    },
    {
      "epoch": 1.312163416898793,
      "grad_norm": 1.5077821711983095,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8821849822998047,
      "num_tokens": 245812426.0,
      "step": 7066
    },
    {
      "epoch": 1.3123491179201485,
      "grad_norm": 1.4717573763881868,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8834967017173767,
      "num_tokens": 245846129.0,
      "step": 7067
    },
    {
      "epoch": 1.3125348189415043,
      "grad_norm": 1.7150916962143365,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8675328493118286,
      "num_tokens": 245874673.0,
      "step": 7068
    },
    {
      "epoch": 1.3127205199628598,
      "grad_norm": 1.532067042956527,
      "learning_rate": 1e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.85783851146698,
      "num_tokens": 245916455.0,
      "step": 7069
    },
    {
      "epoch": 1.3129062209842153,
      "grad_norm": 1.4358843423347671,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8744063973426819,
      "num_tokens": 245954376.0,
      "step": 7070
    },
    {
      "epoch": 1.313091922005571,
      "grad_norm": 1.443054019659926,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8842004537582397,
      "num_tokens": 245989578.0,
      "step": 7071
    },
    {
      "epoch": 1.3132776230269267,
      "grad_norm": 1.5359322318599329,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8705583810806274,
      "num_tokens": 246029766.0,
      "step": 7072
    },
    {
      "epoch": 1.3134633240482823,
      "grad_norm": 1.5406770951384858,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8797693252563477,
      "num_tokens": 246063327.0,
      "step": 7073
    },
    {
      "epoch": 1.3136490250696378,
      "grad_norm": 1.3392657438898852,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8888358473777771,
      "num_tokens": 246106114.0,
      "step": 7074
    },
    {
      "epoch": 1.3138347260909935,
      "grad_norm": 1.5038875640183096,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8816953301429749,
      "num_tokens": 246138476.0,
      "step": 7075
    },
    {
      "epoch": 1.3140204271123492,
      "grad_norm": 1.4842771108436716,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8736655116081238,
      "num_tokens": 246177578.0,
      "step": 7076
    },
    {
      "epoch": 1.3142061281337047,
      "grad_norm": 1.3538189271522014,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8813754320144653,
      "num_tokens": 246219021.0,
      "step": 7077
    },
    {
      "epoch": 1.3143918291550603,
      "grad_norm": 1.6632736778590183,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8651646375656128,
      "num_tokens": 246253020.0,
      "step": 7078
    },
    {
      "epoch": 1.314577530176416,
      "grad_norm": 1.6788646244250305,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8751000761985779,
      "num_tokens": 246285123.0,
      "step": 7079
    },
    {
      "epoch": 1.3147632311977717,
      "grad_norm": 1.5618384586875327,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8662259578704834,
      "num_tokens": 246317010.0,
      "step": 7080
    },
    {
      "epoch": 1.3149489322191272,
      "grad_norm": 1.3294584501849107,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8922785520553589,
      "num_tokens": 246356619.0,
      "step": 7081
    },
    {
      "epoch": 1.3151346332404827,
      "grad_norm": 1.6651880183938823,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8715879917144775,
      "num_tokens": 246387954.0,
      "step": 7082
    },
    {
      "epoch": 1.3153203342618385,
      "grad_norm": 1.6009299380473652,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8785624504089355,
      "num_tokens": 246422681.0,
      "step": 7083
    },
    {
      "epoch": 1.315506035283194,
      "grad_norm": 1.519448804683932,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.87740159034729,
      "num_tokens": 246456035.0,
      "step": 7084
    },
    {
      "epoch": 1.3156917363045497,
      "grad_norm": 1.5458603254110153,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8886967897415161,
      "num_tokens": 246485547.0,
      "step": 7085
    },
    {
      "epoch": 1.3158774373259052,
      "grad_norm": 1.5892312469724783,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8659398555755615,
      "num_tokens": 246518961.0,
      "step": 7086
    },
    {
      "epoch": 1.316063138347261,
      "grad_norm": 1.3590850569320023,
      "learning_rate": 1e-06,
      "loss": 0.286,
      "mean_token_accuracy": 0.8981878161430359,
      "num_tokens": 246556716.0,
      "step": 7087
    },
    {
      "epoch": 1.3162488393686165,
      "grad_norm": 1.4303257620735024,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8887852430343628,
      "num_tokens": 246592636.0,
      "step": 7088
    },
    {
      "epoch": 1.3164345403899722,
      "grad_norm": 1.498456812491287,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.863845944404602,
      "num_tokens": 246634480.0,
      "step": 7089
    },
    {
      "epoch": 1.3166202414113277,
      "grad_norm": 1.4540844763639558,
      "learning_rate": 1e-06,
      "loss": 0.2983,
      "mean_token_accuracy": 0.893649160861969,
      "num_tokens": 246667182.0,
      "step": 7090
    },
    {
      "epoch": 1.3168059424326835,
      "grad_norm": 1.5132674962551271,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8868396282196045,
      "num_tokens": 246701170.0,
      "step": 7091
    },
    {
      "epoch": 1.316991643454039,
      "grad_norm": 1.6267786022666022,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.872125506401062,
      "num_tokens": 246732263.0,
      "step": 7092
    },
    {
      "epoch": 1.3171773444753947,
      "grad_norm": 1.455360039411815,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8782728910446167,
      "num_tokens": 246768336.0,
      "step": 7093
    },
    {
      "epoch": 1.3173630454967502,
      "grad_norm": 1.4774232227674202,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8852074146270752,
      "num_tokens": 246802449.0,
      "step": 7094
    },
    {
      "epoch": 1.317548746518106,
      "grad_norm": 1.3845621905159549,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.88661128282547,
      "num_tokens": 246838976.0,
      "step": 7095
    },
    {
      "epoch": 1.3177344475394615,
      "grad_norm": 1.6137092504372816,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8643928170204163,
      "num_tokens": 246871447.0,
      "step": 7096
    },
    {
      "epoch": 1.317920148560817,
      "grad_norm": 1.560162325252161,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8750895857810974,
      "num_tokens": 246903237.0,
      "step": 7097
    },
    {
      "epoch": 1.3181058495821727,
      "grad_norm": 1.5506159141099198,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8745653629302979,
      "num_tokens": 246934791.0,
      "step": 7098
    },
    {
      "epoch": 1.3182915506035284,
      "grad_norm": 1.4251274867709067,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8834782242774963,
      "num_tokens": 246970781.0,
      "step": 7099
    },
    {
      "epoch": 1.318477251624884,
      "grad_norm": 1.413632796953167,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8768851161003113,
      "num_tokens": 247003511.0,
      "step": 7100
    },
    {
      "epoch": 1.3186629526462395,
      "grad_norm": 1.4114577507540855,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8747526407241821,
      "num_tokens": 247041346.0,
      "step": 7101
    },
    {
      "epoch": 1.3188486536675952,
      "grad_norm": 1.4497218334777398,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8827161192893982,
      "num_tokens": 247078177.0,
      "step": 7102
    },
    {
      "epoch": 1.319034354688951,
      "grad_norm": 1.5876027299766473,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8672488927841187,
      "num_tokens": 247113729.0,
      "step": 7103
    },
    {
      "epoch": 1.3192200557103064,
      "grad_norm": 1.5226758670336944,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8757011294364929,
      "num_tokens": 247149729.0,
      "step": 7104
    },
    {
      "epoch": 1.319405756731662,
      "grad_norm": 1.5364055324493255,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8807390332221985,
      "num_tokens": 247178743.0,
      "step": 7105
    },
    {
      "epoch": 1.3195914577530177,
      "grad_norm": 1.5027756489333564,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8734480142593384,
      "num_tokens": 247212077.0,
      "step": 7106
    },
    {
      "epoch": 1.3197771587743732,
      "grad_norm": 1.4893200510309088,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8793449997901917,
      "num_tokens": 247249652.0,
      "step": 7107
    },
    {
      "epoch": 1.319962859795729,
      "grad_norm": 1.4400594282166834,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8919771909713745,
      "num_tokens": 247283062.0,
      "step": 7108
    },
    {
      "epoch": 1.3201485608170844,
      "grad_norm": 1.5662100658373015,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8675270080566406,
      "num_tokens": 247319625.0,
      "step": 7109
    },
    {
      "epoch": 1.3203342618384402,
      "grad_norm": 1.5284263510061271,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8697051405906677,
      "num_tokens": 247357603.0,
      "step": 7110
    },
    {
      "epoch": 1.3205199628597957,
      "grad_norm": 1.458461155261111,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.882596492767334,
      "num_tokens": 247392106.0,
      "step": 7111
    },
    {
      "epoch": 1.3207056638811514,
      "grad_norm": 1.6107189187022377,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8741575479507446,
      "num_tokens": 247422155.0,
      "step": 7112
    },
    {
      "epoch": 1.320891364902507,
      "grad_norm": 1.4710924388129705,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8839123249053955,
      "num_tokens": 247458602.0,
      "step": 7113
    },
    {
      "epoch": 1.3210770659238626,
      "grad_norm": 1.441084619260668,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8775441646575928,
      "num_tokens": 247497141.0,
      "step": 7114
    },
    {
      "epoch": 1.3212627669452182,
      "grad_norm": 1.5028064854948462,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.868202269077301,
      "num_tokens": 247535630.0,
      "step": 7115
    },
    {
      "epoch": 1.321448467966574,
      "grad_norm": 1.4579971015617161,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.8923879265785217,
      "num_tokens": 247567504.0,
      "step": 7116
    },
    {
      "epoch": 1.3216341689879294,
      "grad_norm": 1.3433746210511206,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8797763586044312,
      "num_tokens": 247608749.0,
      "step": 7117
    },
    {
      "epoch": 1.3218198700092851,
      "grad_norm": 1.4174025673131956,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8716888427734375,
      "num_tokens": 247651614.0,
      "step": 7118
    },
    {
      "epoch": 1.3220055710306406,
      "grad_norm": 1.442778797871895,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.885430097579956,
      "num_tokens": 247685311.0,
      "step": 7119
    },
    {
      "epoch": 1.3221912720519962,
      "grad_norm": 1.3715943753380266,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.864970326423645,
      "num_tokens": 247730484.0,
      "step": 7120
    },
    {
      "epoch": 1.322376973073352,
      "grad_norm": 1.4745539189205261,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8768197298049927,
      "num_tokens": 247767248.0,
      "step": 7121
    },
    {
      "epoch": 1.3225626740947076,
      "grad_norm": 1.566504201418158,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8732319474220276,
      "num_tokens": 247800530.0,
      "step": 7122
    },
    {
      "epoch": 1.3227483751160631,
      "grad_norm": 1.629278659310225,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8760890960693359,
      "num_tokens": 247831522.0,
      "step": 7123
    },
    {
      "epoch": 1.3229340761374186,
      "grad_norm": 1.511952080773475,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8834449052810669,
      "num_tokens": 247864182.0,
      "step": 7124
    },
    {
      "epoch": 1.3231197771587744,
      "grad_norm": 1.4469292563054952,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8795070648193359,
      "num_tokens": 247898608.0,
      "step": 7125
    },
    {
      "epoch": 1.3233054781801301,
      "grad_norm": 1.341748476276432,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8795796632766724,
      "num_tokens": 247941079.0,
      "step": 7126
    },
    {
      "epoch": 1.3234911792014856,
      "grad_norm": 1.53135982367593,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8728100657463074,
      "num_tokens": 247974736.0,
      "step": 7127
    },
    {
      "epoch": 1.3236768802228411,
      "grad_norm": 1.664553633241436,
      "learning_rate": 1e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8562581539154053,
      "num_tokens": 248006091.0,
      "step": 7128
    },
    {
      "epoch": 1.3238625812441969,
      "grad_norm": 1.4326447368590587,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8750447034835815,
      "num_tokens": 248043268.0,
      "step": 7129
    },
    {
      "epoch": 1.3240482822655524,
      "grad_norm": 1.684338847803399,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8678377866744995,
      "num_tokens": 248070807.0,
      "step": 7130
    },
    {
      "epoch": 1.3242339832869081,
      "grad_norm": 1.4740484123431337,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8731374740600586,
      "num_tokens": 248106228.0,
      "step": 7131
    },
    {
      "epoch": 1.3244196843082636,
      "grad_norm": 1.560727301233091,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8792112469673157,
      "num_tokens": 248139750.0,
      "step": 7132
    },
    {
      "epoch": 1.3246053853296194,
      "grad_norm": 1.538360232832459,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8822920322418213,
      "num_tokens": 248172750.0,
      "step": 7133
    },
    {
      "epoch": 1.3247910863509749,
      "grad_norm": 1.4537671042107063,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8668888807296753,
      "num_tokens": 248216597.0,
      "step": 7134
    },
    {
      "epoch": 1.3249767873723306,
      "grad_norm": 1.4177145146004575,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8863493800163269,
      "num_tokens": 248251601.0,
      "step": 7135
    },
    {
      "epoch": 1.325162488393686,
      "grad_norm": 1.502084380090442,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8678661584854126,
      "num_tokens": 248289944.0,
      "step": 7136
    },
    {
      "epoch": 1.3253481894150418,
      "grad_norm": 1.5511843567243568,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8688209056854248,
      "num_tokens": 248325131.0,
      "step": 7137
    },
    {
      "epoch": 1.3255338904363974,
      "grad_norm": 1.6484706267399958,
      "learning_rate": 1e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8547639846801758,
      "num_tokens": 248360235.0,
      "step": 7138
    },
    {
      "epoch": 1.325719591457753,
      "grad_norm": 1.4739247701772966,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8841931819915771,
      "num_tokens": 248395122.0,
      "step": 7139
    },
    {
      "epoch": 1.3259052924791086,
      "grad_norm": 1.298172154457448,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8723046183586121,
      "num_tokens": 248440912.0,
      "step": 7140
    },
    {
      "epoch": 1.3260909935004643,
      "grad_norm": 1.6358300969432586,
      "learning_rate": 1e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8522845506668091,
      "num_tokens": 248475731.0,
      "step": 7141
    },
    {
      "epoch": 1.3262766945218198,
      "grad_norm": 1.5227796655459316,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8789113759994507,
      "num_tokens": 248512247.0,
      "step": 7142
    },
    {
      "epoch": 1.3264623955431754,
      "grad_norm": 1.6726288769754816,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8722859621047974,
      "num_tokens": 248541627.0,
      "step": 7143
    },
    {
      "epoch": 1.326648096564531,
      "grad_norm": 1.5572790696178946,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8756787776947021,
      "num_tokens": 248576666.0,
      "step": 7144
    },
    {
      "epoch": 1.3268337975858868,
      "grad_norm": 1.5234785142082607,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.863358736038208,
      "num_tokens": 248612472.0,
      "step": 7145
    },
    {
      "epoch": 1.3270194986072423,
      "grad_norm": 1.5176890176957842,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.872972846031189,
      "num_tokens": 248650276.0,
      "step": 7146
    },
    {
      "epoch": 1.3272051996285978,
      "grad_norm": 1.5832728883353886,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8667463064193726,
      "num_tokens": 248686760.0,
      "step": 7147
    },
    {
      "epoch": 1.3273909006499536,
      "grad_norm": 1.4650333364636907,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8706667423248291,
      "num_tokens": 248722316.0,
      "step": 7148
    },
    {
      "epoch": 1.3275766016713093,
      "grad_norm": 1.5656377925131937,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8776117563247681,
      "num_tokens": 248750856.0,
      "step": 7149
    },
    {
      "epoch": 1.3277623026926648,
      "grad_norm": 1.5587086626260342,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8735880851745605,
      "num_tokens": 248782590.0,
      "step": 7150
    },
    {
      "epoch": 1.3279480037140203,
      "grad_norm": 1.526577658381632,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8711926937103271,
      "num_tokens": 248815302.0,
      "step": 7151
    },
    {
      "epoch": 1.328133704735376,
      "grad_norm": 1.6740122267626276,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8776239156723022,
      "num_tokens": 248843215.0,
      "step": 7152
    },
    {
      "epoch": 1.3283194057567318,
      "grad_norm": 1.87619389222882,
      "learning_rate": 1e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8550788164138794,
      "num_tokens": 248870755.0,
      "step": 7153
    },
    {
      "epoch": 1.3285051067780873,
      "grad_norm": 1.4245406658928945,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8815643787384033,
      "num_tokens": 248907766.0,
      "step": 7154
    },
    {
      "epoch": 1.3286908077994428,
      "grad_norm": 1.4949083915341834,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8795003890991211,
      "num_tokens": 248940860.0,
      "step": 7155
    },
    {
      "epoch": 1.3288765088207986,
      "grad_norm": 1.5676780445708867,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8775415420532227,
      "num_tokens": 248972798.0,
      "step": 7156
    },
    {
      "epoch": 1.329062209842154,
      "grad_norm": 1.501962460673896,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8723717927932739,
      "num_tokens": 249009006.0,
      "step": 7157
    },
    {
      "epoch": 1.3292479108635098,
      "grad_norm": 1.425561539488962,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8709544539451599,
      "num_tokens": 249046237.0,
      "step": 7158
    },
    {
      "epoch": 1.3294336118848653,
      "grad_norm": 1.5329559101071635,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8768861889839172,
      "num_tokens": 249082112.0,
      "step": 7159
    },
    {
      "epoch": 1.329619312906221,
      "grad_norm": 1.5161619308979106,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8782212734222412,
      "num_tokens": 249112962.0,
      "step": 7160
    },
    {
      "epoch": 1.3298050139275766,
      "grad_norm": 1.4357463739148606,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8788100481033325,
      "num_tokens": 249152329.0,
      "step": 7161
    },
    {
      "epoch": 1.3299907149489323,
      "grad_norm": 1.5440529138294294,
      "learning_rate": 1e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.853805422782898,
      "num_tokens": 249189915.0,
      "step": 7162
    },
    {
      "epoch": 1.3301764159702878,
      "grad_norm": 1.3689497646967665,
      "learning_rate": 1e-06,
      "loss": 0.3086,
      "mean_token_accuracy": 0.8900232911109924,
      "num_tokens": 249225993.0,
      "step": 7163
    },
    {
      "epoch": 1.3303621169916435,
      "grad_norm": 1.3556418088148359,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8837231397628784,
      "num_tokens": 249264878.0,
      "step": 7164
    },
    {
      "epoch": 1.330547818012999,
      "grad_norm": 1.598887642935284,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8663411140441895,
      "num_tokens": 249294934.0,
      "step": 7165
    },
    {
      "epoch": 1.3307335190343546,
      "grad_norm": 1.4648821938870018,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8752405047416687,
      "num_tokens": 249331118.0,
      "step": 7166
    },
    {
      "epoch": 1.3309192200557103,
      "grad_norm": 1.443715332935541,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8776851892471313,
      "num_tokens": 249368956.0,
      "step": 7167
    },
    {
      "epoch": 1.331104921077066,
      "grad_norm": 1.3448792247890868,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8948386907577515,
      "num_tokens": 249409863.0,
      "step": 7168
    },
    {
      "epoch": 1.3312906220984215,
      "grad_norm": 1.545837588799568,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8713512420654297,
      "num_tokens": 249443476.0,
      "step": 7169
    },
    {
      "epoch": 1.331476323119777,
      "grad_norm": 1.4178238718249765,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8759089112281799,
      "num_tokens": 249481126.0,
      "step": 7170
    },
    {
      "epoch": 1.3316620241411328,
      "grad_norm": 1.5465657947945273,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.862089991569519,
      "num_tokens": 249517602.0,
      "step": 7171
    },
    {
      "epoch": 1.3318477251624885,
      "grad_norm": 1.4433300313887458,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8906752467155457,
      "num_tokens": 249554465.0,
      "step": 7172
    },
    {
      "epoch": 1.332033426183844,
      "grad_norm": 1.6197124527260687,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8708357214927673,
      "num_tokens": 249585826.0,
      "step": 7173
    },
    {
      "epoch": 1.3322191272051995,
      "grad_norm": 1.6258495457676359,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8686107397079468,
      "num_tokens": 249615633.0,
      "step": 7174
    },
    {
      "epoch": 1.3324048282265553,
      "grad_norm": 1.4101370347158018,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8888239860534668,
      "num_tokens": 249652433.0,
      "step": 7175
    },
    {
      "epoch": 1.332590529247911,
      "grad_norm": 1.9196877960843666,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8780652284622192,
      "num_tokens": 249677051.0,
      "step": 7176
    },
    {
      "epoch": 1.3327762302692665,
      "grad_norm": 1.7071559492105115,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8744543790817261,
      "num_tokens": 249703958.0,
      "step": 7177
    },
    {
      "epoch": 1.332961931290622,
      "grad_norm": 1.5235449261673109,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8680377006530762,
      "num_tokens": 249736936.0,
      "step": 7178
    },
    {
      "epoch": 1.3331476323119777,
      "grad_norm": 1.3714105916874169,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8887453675270081,
      "num_tokens": 249778169.0,
      "step": 7179
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.4140098564711607,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.877784252166748,
      "num_tokens": 249816248.0,
      "step": 7180
    },
    {
      "epoch": 1.333519034354689,
      "grad_norm": 1.4260227490798179,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8776594996452332,
      "num_tokens": 249854159.0,
      "step": 7181
    },
    {
      "epoch": 1.3337047353760445,
      "grad_norm": 1.432584919394482,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8779610991477966,
      "num_tokens": 249894465.0,
      "step": 7182
    },
    {
      "epoch": 1.3338904363974002,
      "grad_norm": 1.6563857791876528,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8660005331039429,
      "num_tokens": 249923160.0,
      "step": 7183
    },
    {
      "epoch": 1.3340761374187557,
      "grad_norm": 1.4670218761816303,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8808348178863525,
      "num_tokens": 249958702.0,
      "step": 7184
    },
    {
      "epoch": 1.3342618384401115,
      "grad_norm": 1.577840885343097,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8829386830329895,
      "num_tokens": 249990129.0,
      "step": 7185
    },
    {
      "epoch": 1.334447539461467,
      "grad_norm": 1.471002630037832,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.887040376663208,
      "num_tokens": 250023863.0,
      "step": 7186
    },
    {
      "epoch": 1.3346332404828227,
      "grad_norm": 1.5395269501861433,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8788707256317139,
      "num_tokens": 250057641.0,
      "step": 7187
    },
    {
      "epoch": 1.3348189415041782,
      "grad_norm": 1.5776516429578518,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8626712560653687,
      "num_tokens": 250093480.0,
      "step": 7188
    },
    {
      "epoch": 1.335004642525534,
      "grad_norm": 1.5973341188021795,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8668371438980103,
      "num_tokens": 250127418.0,
      "step": 7189
    },
    {
      "epoch": 1.3351903435468895,
      "grad_norm": 1.4830207940558708,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8748063445091248,
      "num_tokens": 250162527.0,
      "step": 7190
    },
    {
      "epoch": 1.3353760445682452,
      "grad_norm": 1.5248490555083551,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8827769756317139,
      "num_tokens": 250197670.0,
      "step": 7191
    },
    {
      "epoch": 1.3355617455896007,
      "grad_norm": 1.2984779641000428,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.8982919454574585,
      "num_tokens": 250235218.0,
      "step": 7192
    },
    {
      "epoch": 1.3357474466109562,
      "grad_norm": 1.5659127611461912,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8702219724655151,
      "num_tokens": 250268902.0,
      "step": 7193
    },
    {
      "epoch": 1.335933147632312,
      "grad_norm": 1.6330214797907214,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8697688579559326,
      "num_tokens": 250301233.0,
      "step": 7194
    },
    {
      "epoch": 1.3361188486536677,
      "grad_norm": 1.6285295233299308,
      "learning_rate": 1e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8606299757957458,
      "num_tokens": 250333887.0,
      "step": 7195
    },
    {
      "epoch": 1.3363045496750232,
      "grad_norm": 1.4962964010491386,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8820346593856812,
      "num_tokens": 250366686.0,
      "step": 7196
    },
    {
      "epoch": 1.3364902506963787,
      "grad_norm": 1.7581899631413274,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8754942417144775,
      "num_tokens": 250394136.0,
      "step": 7197
    },
    {
      "epoch": 1.3366759517177345,
      "grad_norm": 1.6904588527084456,
      "learning_rate": 1e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8676405549049377,
      "num_tokens": 250424226.0,
      "step": 7198
    },
    {
      "epoch": 1.3368616527390902,
      "grad_norm": 1.6011743656701511,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8777561187744141,
      "num_tokens": 250456763.0,
      "step": 7199
    },
    {
      "epoch": 1.3370473537604457,
      "grad_norm": 1.5586477398564142,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8729333281517029,
      "num_tokens": 250492057.0,
      "step": 7200
    },
    {
      "epoch": 1.3372330547818012,
      "grad_norm": 1.455021138952764,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.884771466255188,
      "num_tokens": 250525520.0,
      "step": 7201
    },
    {
      "epoch": 1.337418755803157,
      "grad_norm": 1.654529848476636,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8625748157501221,
      "num_tokens": 250555532.0,
      "step": 7202
    },
    {
      "epoch": 1.3376044568245125,
      "grad_norm": 1.6091158093981306,
      "learning_rate": 1e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8625658750534058,
      "num_tokens": 250590251.0,
      "step": 7203
    },
    {
      "epoch": 1.3377901578458682,
      "grad_norm": 1.5113684980476831,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.875525951385498,
      "num_tokens": 250620972.0,
      "step": 7204
    },
    {
      "epoch": 1.3379758588672237,
      "grad_norm": 1.4729975981959855,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8636122941970825,
      "num_tokens": 250655751.0,
      "step": 7205
    },
    {
      "epoch": 1.3381615598885794,
      "grad_norm": 1.5617198515227078,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8787050247192383,
      "num_tokens": 250688855.0,
      "step": 7206
    },
    {
      "epoch": 1.338347260909935,
      "grad_norm": 1.528273415147543,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8852972984313965,
      "num_tokens": 250719572.0,
      "step": 7207
    },
    {
      "epoch": 1.3385329619312907,
      "grad_norm": 1.7003026325548143,
      "learning_rate": 1e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8653035163879395,
      "num_tokens": 250751166.0,
      "step": 7208
    },
    {
      "epoch": 1.3387186629526462,
      "grad_norm": 1.5349632777021465,
      "learning_rate": 1e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8601412773132324,
      "num_tokens": 250788838.0,
      "step": 7209
    },
    {
      "epoch": 1.338904363974002,
      "grad_norm": 1.427489119957522,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8686631917953491,
      "num_tokens": 250830517.0,
      "step": 7210
    },
    {
      "epoch": 1.3390900649953574,
      "grad_norm": 1.5534293512577628,
      "learning_rate": 1e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8606492280960083,
      "num_tokens": 250868421.0,
      "step": 7211
    },
    {
      "epoch": 1.3392757660167132,
      "grad_norm": 1.7294770433718867,
      "learning_rate": 1e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8563995361328125,
      "num_tokens": 250903212.0,
      "step": 7212
    },
    {
      "epoch": 1.3394614670380687,
      "grad_norm": 1.4293058142248343,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.886259138584137,
      "num_tokens": 250937777.0,
      "step": 7213
    },
    {
      "epoch": 1.3396471680594244,
      "grad_norm": 1.756935333384008,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8831318020820618,
      "num_tokens": 250962124.0,
      "step": 7214
    },
    {
      "epoch": 1.33983286908078,
      "grad_norm": 1.729451875410237,
      "learning_rate": 1e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8601155281066895,
      "num_tokens": 250991774.0,
      "step": 7215
    },
    {
      "epoch": 1.3400185701021354,
      "grad_norm": 1.6797784491212566,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8821409940719604,
      "num_tokens": 251019494.0,
      "step": 7216
    },
    {
      "epoch": 1.3402042711234912,
      "grad_norm": 1.561151531746365,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8801276683807373,
      "num_tokens": 251050735.0,
      "step": 7217
    },
    {
      "epoch": 1.340389972144847,
      "grad_norm": 1.5080777358837472,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8670026063919067,
      "num_tokens": 251088065.0,
      "step": 7218
    },
    {
      "epoch": 1.3405756731662024,
      "grad_norm": 1.6484494032101087,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8732126951217651,
      "num_tokens": 251119287.0,
      "step": 7219
    },
    {
      "epoch": 1.340761374187558,
      "grad_norm": 1.6710601635014128,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8727405667304993,
      "num_tokens": 251150471.0,
      "step": 7220
    },
    {
      "epoch": 1.3409470752089137,
      "grad_norm": 1.545844962324973,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8966500759124756,
      "num_tokens": 251180840.0,
      "step": 7221
    },
    {
      "epoch": 1.3411327762302694,
      "grad_norm": 1.5121171479253637,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8803567886352539,
      "num_tokens": 251217042.0,
      "step": 7222
    },
    {
      "epoch": 1.341318477251625,
      "grad_norm": 1.7641579842614317,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8737165927886963,
      "num_tokens": 251245531.0,
      "step": 7223
    },
    {
      "epoch": 1.3415041782729804,
      "grad_norm": 1.6300500809154495,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8893053531646729,
      "num_tokens": 251280089.0,
      "step": 7224
    },
    {
      "epoch": 1.3416898792943361,
      "grad_norm": 1.505913957284069,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8724770545959473,
      "num_tokens": 251317375.0,
      "step": 7225
    },
    {
      "epoch": 1.3418755803156919,
      "grad_norm": 1.6124041877916029,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8748821020126343,
      "num_tokens": 251357117.0,
      "step": 7226
    },
    {
      "epoch": 1.3420612813370474,
      "grad_norm": 1.4623583522315116,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8818222284317017,
      "num_tokens": 251395481.0,
      "step": 7227
    },
    {
      "epoch": 1.342246982358403,
      "grad_norm": 1.3876679168650314,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8901267647743225,
      "num_tokens": 251433729.0,
      "step": 7228
    },
    {
      "epoch": 1.3424326833797586,
      "grad_norm": 1.4705074456564489,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8782652020454407,
      "num_tokens": 251469408.0,
      "step": 7229
    },
    {
      "epoch": 1.3426183844011141,
      "grad_norm": 1.4411347549318616,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.872515082359314,
      "num_tokens": 251505383.0,
      "step": 7230
    },
    {
      "epoch": 1.3428040854224699,
      "grad_norm": 1.4177085262242026,
      "learning_rate": 1e-06,
      "loss": 0.3254,
      "mean_token_accuracy": 0.8857120275497437,
      "num_tokens": 251545426.0,
      "step": 7231
    },
    {
      "epoch": 1.3429897864438254,
      "grad_norm": 1.452590525721628,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8801712989807129,
      "num_tokens": 251584692.0,
      "step": 7232
    },
    {
      "epoch": 1.3431754874651811,
      "grad_norm": 1.3417444305805692,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8840789794921875,
      "num_tokens": 251627521.0,
      "step": 7233
    },
    {
      "epoch": 1.3433611884865366,
      "grad_norm": 1.433599787692476,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8905690908432007,
      "num_tokens": 251659545.0,
      "step": 7234
    },
    {
      "epoch": 1.3435468895078924,
      "grad_norm": 1.5259845590784917,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8717806339263916,
      "num_tokens": 251692975.0,
      "step": 7235
    },
    {
      "epoch": 1.3437325905292479,
      "grad_norm": 1.550184029237242,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8781419992446899,
      "num_tokens": 251728408.0,
      "step": 7236
    },
    {
      "epoch": 1.3439182915506036,
      "grad_norm": 1.4730494115449673,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8749788999557495,
      "num_tokens": 251765552.0,
      "step": 7237
    },
    {
      "epoch": 1.3441039925719591,
      "grad_norm": 1.5003457564398406,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8774312734603882,
      "num_tokens": 251800196.0,
      "step": 7238
    },
    {
      "epoch": 1.3442896935933146,
      "grad_norm": 1.4870581914874472,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8821832537651062,
      "num_tokens": 251836055.0,
      "step": 7239
    },
    {
      "epoch": 1.3444753946146704,
      "grad_norm": 1.4737243055143703,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8783199191093445,
      "num_tokens": 251872455.0,
      "step": 7240
    },
    {
      "epoch": 1.344661095636026,
      "grad_norm": 1.5517512962730338,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8743137121200562,
      "num_tokens": 251907112.0,
      "step": 7241
    },
    {
      "epoch": 1.3448467966573816,
      "grad_norm": 1.5852236702679694,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8629029989242554,
      "num_tokens": 251938454.0,
      "step": 7242
    },
    {
      "epoch": 1.3450324976787371,
      "grad_norm": 1.546796166682161,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8760755062103271,
      "num_tokens": 251975555.0,
      "step": 7243
    },
    {
      "epoch": 1.3452181987000928,
      "grad_norm": 1.5921885184788203,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8749498128890991,
      "num_tokens": 252012253.0,
      "step": 7244
    },
    {
      "epoch": 1.3454038997214486,
      "grad_norm": 1.4984528029145692,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8848516941070557,
      "num_tokens": 252046260.0,
      "step": 7245
    },
    {
      "epoch": 1.345589600742804,
      "grad_norm": 1.3656426791860226,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8808127045631409,
      "num_tokens": 252087750.0,
      "step": 7246
    },
    {
      "epoch": 1.3457753017641596,
      "grad_norm": 1.4898511317886458,
      "learning_rate": 1e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8642936944961548,
      "num_tokens": 252128257.0,
      "step": 7247
    },
    {
      "epoch": 1.3459610027855153,
      "grad_norm": 1.564350377969322,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8854616284370422,
      "num_tokens": 252157789.0,
      "step": 7248
    },
    {
      "epoch": 1.346146703806871,
      "grad_norm": 1.7147098717376292,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8692052364349365,
      "num_tokens": 252187746.0,
      "step": 7249
    },
    {
      "epoch": 1.3463324048282266,
      "grad_norm": 1.4419756230668073,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8800626397132874,
      "num_tokens": 252224326.0,
      "step": 7250
    },
    {
      "epoch": 1.346518105849582,
      "grad_norm": 1.5092225606389014,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8798065185546875,
      "num_tokens": 252263630.0,
      "step": 7251
    },
    {
      "epoch": 1.3467038068709378,
      "grad_norm": 1.5625612864694518,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8854379653930664,
      "num_tokens": 252297381.0,
      "step": 7252
    },
    {
      "epoch": 1.3468895078922933,
      "grad_norm": 1.610143464756094,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8774912357330322,
      "num_tokens": 252330447.0,
      "step": 7253
    },
    {
      "epoch": 1.347075208913649,
      "grad_norm": 1.450205466610541,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.868455708026886,
      "num_tokens": 252369941.0,
      "step": 7254
    },
    {
      "epoch": 1.3472609099350046,
      "grad_norm": 1.3881811215960516,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8765432834625244,
      "num_tokens": 252408821.0,
      "step": 7255
    },
    {
      "epoch": 1.3474466109563603,
      "grad_norm": 1.4367972021890043,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8798366785049438,
      "num_tokens": 252444078.0,
      "step": 7256
    },
    {
      "epoch": 1.3476323119777158,
      "grad_norm": 1.4776757252621462,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8906785249710083,
      "num_tokens": 252477732.0,
      "step": 7257
    },
    {
      "epoch": 1.3478180129990716,
      "grad_norm": 1.4872975410997482,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.881438136100769,
      "num_tokens": 252512825.0,
      "step": 7258
    },
    {
      "epoch": 1.348003714020427,
      "grad_norm": 1.453722548767704,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8735072016716003,
      "num_tokens": 252549808.0,
      "step": 7259
    },
    {
      "epoch": 1.3481894150417828,
      "grad_norm": 1.5899829278363569,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8708622455596924,
      "num_tokens": 252580282.0,
      "step": 7260
    },
    {
      "epoch": 1.3483751160631383,
      "grad_norm": 1.4096969111559032,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8681544065475464,
      "num_tokens": 252621731.0,
      "step": 7261
    },
    {
      "epoch": 1.348560817084494,
      "grad_norm": 1.4186059326645764,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8808354139328003,
      "num_tokens": 252659066.0,
      "step": 7262
    },
    {
      "epoch": 1.3487465181058496,
      "grad_norm": 1.4441327600335296,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8659927845001221,
      "num_tokens": 252695930.0,
      "step": 7263
    },
    {
      "epoch": 1.3489322191272053,
      "grad_norm": 1.389012277533331,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8938851356506348,
      "num_tokens": 252730326.0,
      "step": 7264
    },
    {
      "epoch": 1.3491179201485608,
      "grad_norm": 1.4285280549577473,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8681203722953796,
      "num_tokens": 252770777.0,
      "step": 7265
    },
    {
      "epoch": 1.3493036211699163,
      "grad_norm": 1.572880388757456,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8787973523139954,
      "num_tokens": 252804050.0,
      "step": 7266
    },
    {
      "epoch": 1.349489322191272,
      "grad_norm": 1.4702030822118262,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8738016486167908,
      "num_tokens": 252840539.0,
      "step": 7267
    },
    {
      "epoch": 1.3496750232126278,
      "grad_norm": 1.442492028775709,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8764918446540833,
      "num_tokens": 252878596.0,
      "step": 7268
    },
    {
      "epoch": 1.3498607242339833,
      "grad_norm": 1.6333017143614503,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8696221113204956,
      "num_tokens": 252908345.0,
      "step": 7269
    },
    {
      "epoch": 1.3500464252553388,
      "grad_norm": 1.6688736095463361,
      "learning_rate": 1e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8489871621131897,
      "num_tokens": 252942268.0,
      "step": 7270
    },
    {
      "epoch": 1.3502321262766945,
      "grad_norm": 1.3916645974697228,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8841694593429565,
      "num_tokens": 252981771.0,
      "step": 7271
    },
    {
      "epoch": 1.3504178272980503,
      "grad_norm": 1.6872004877073974,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8778842687606812,
      "num_tokens": 253012373.0,
      "step": 7272
    },
    {
      "epoch": 1.3506035283194058,
      "grad_norm": 1.4255336141304613,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8680991530418396,
      "num_tokens": 253051061.0,
      "step": 7273
    },
    {
      "epoch": 1.3507892293407613,
      "grad_norm": 1.5491557255924489,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8844568133354187,
      "num_tokens": 253081190.0,
      "step": 7274
    },
    {
      "epoch": 1.350974930362117,
      "grad_norm": 1.522434137758405,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8694378137588501,
      "num_tokens": 253120570.0,
      "step": 7275
    },
    {
      "epoch": 1.3511606313834725,
      "grad_norm": 1.5494207936412603,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8788096904754639,
      "num_tokens": 253154123.0,
      "step": 7276
    },
    {
      "epoch": 1.3513463324048283,
      "grad_norm": 1.5150497725341727,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8693057894706726,
      "num_tokens": 253189430.0,
      "step": 7277
    },
    {
      "epoch": 1.3515320334261838,
      "grad_norm": 1.388781631814282,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8856384754180908,
      "num_tokens": 253226298.0,
      "step": 7278
    },
    {
      "epoch": 1.3517177344475395,
      "grad_norm": 1.4570587495511307,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.875821590423584,
      "num_tokens": 253262753.0,
      "step": 7279
    },
    {
      "epoch": 1.351903435468895,
      "grad_norm": 1.442141696104905,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.895401120185852,
      "num_tokens": 253296499.0,
      "step": 7280
    },
    {
      "epoch": 1.3520891364902508,
      "grad_norm": 1.52983513026328,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8794710040092468,
      "num_tokens": 253326254.0,
      "step": 7281
    },
    {
      "epoch": 1.3522748375116063,
      "grad_norm": 1.5778615535923193,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8726451992988586,
      "num_tokens": 253358143.0,
      "step": 7282
    },
    {
      "epoch": 1.352460538532962,
      "grad_norm": 1.5104731808726772,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8825994729995728,
      "num_tokens": 253389226.0,
      "step": 7283
    },
    {
      "epoch": 1.3526462395543175,
      "grad_norm": 1.44658672959394,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8712023496627808,
      "num_tokens": 253427775.0,
      "step": 7284
    },
    {
      "epoch": 1.3528319405756732,
      "grad_norm": 1.5190896841843577,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8780579566955566,
      "num_tokens": 253461171.0,
      "step": 7285
    },
    {
      "epoch": 1.3530176415970288,
      "grad_norm": 1.5628879756309042,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8556578755378723,
      "num_tokens": 253499917.0,
      "step": 7286
    },
    {
      "epoch": 1.3532033426183845,
      "grad_norm": 1.574039904478101,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.878656268119812,
      "num_tokens": 253532690.0,
      "step": 7287
    },
    {
      "epoch": 1.35338904363974,
      "grad_norm": 1.5151052763865838,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8656332492828369,
      "num_tokens": 253568101.0,
      "step": 7288
    },
    {
      "epoch": 1.3535747446610955,
      "grad_norm": 1.3792653019075,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8882894515991211,
      "num_tokens": 253606787.0,
      "step": 7289
    },
    {
      "epoch": 1.3537604456824512,
      "grad_norm": 1.44172426179328,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8864409923553467,
      "num_tokens": 253646326.0,
      "step": 7290
    },
    {
      "epoch": 1.353946146703807,
      "grad_norm": 1.3933319238281685,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.880377471446991,
      "num_tokens": 253690074.0,
      "step": 7291
    },
    {
      "epoch": 1.3541318477251625,
      "grad_norm": 1.5038373892606656,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.886666476726532,
      "num_tokens": 253721997.0,
      "step": 7292
    },
    {
      "epoch": 1.354317548746518,
      "grad_norm": 1.469267859212044,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8775997161865234,
      "num_tokens": 253756685.0,
      "step": 7293
    },
    {
      "epoch": 1.3545032497678737,
      "grad_norm": 1.528059999588206,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8885911107063293,
      "num_tokens": 253789747.0,
      "step": 7294
    },
    {
      "epoch": 1.3546889507892295,
      "grad_norm": 1.560940989664221,
      "learning_rate": 1e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8680363297462463,
      "num_tokens": 253823600.0,
      "step": 7295
    },
    {
      "epoch": 1.354874651810585,
      "grad_norm": 1.5665091278486247,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8764021992683411,
      "num_tokens": 253855730.0,
      "step": 7296
    },
    {
      "epoch": 1.3550603528319405,
      "grad_norm": 1.5292210172652794,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8673614263534546,
      "num_tokens": 253891335.0,
      "step": 7297
    },
    {
      "epoch": 1.3552460538532962,
      "grad_norm": 1.4115659393998463,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8759281635284424,
      "num_tokens": 253926739.0,
      "step": 7298
    },
    {
      "epoch": 1.3554317548746517,
      "grad_norm": 1.5207347924974481,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8698609471321106,
      "num_tokens": 253959469.0,
      "step": 7299
    },
    {
      "epoch": 1.3556174558960075,
      "grad_norm": 1.466667853184908,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8778494596481323,
      "num_tokens": 253997877.0,
      "step": 7300
    },
    {
      "epoch": 1.355803156917363,
      "grad_norm": 1.5026359080970324,
      "learning_rate": 1e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8652336001396179,
      "num_tokens": 254033670.0,
      "step": 7301
    },
    {
      "epoch": 1.3559888579387187,
      "grad_norm": 1.4860375769577063,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8715620040893555,
      "num_tokens": 254069602.0,
      "step": 7302
    },
    {
      "epoch": 1.3561745589600742,
      "grad_norm": 1.4868422160221355,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8837812542915344,
      "num_tokens": 254102225.0,
      "step": 7303
    },
    {
      "epoch": 1.35636025998143,
      "grad_norm": 1.6776236332764765,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8742177486419678,
      "num_tokens": 254128956.0,
      "step": 7304
    },
    {
      "epoch": 1.3565459610027855,
      "grad_norm": 1.4401305270268576,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8792752027511597,
      "num_tokens": 254166453.0,
      "step": 7305
    },
    {
      "epoch": 1.3567316620241412,
      "grad_norm": 1.3578857115665783,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8810099959373474,
      "num_tokens": 254205406.0,
      "step": 7306
    },
    {
      "epoch": 1.3569173630454967,
      "grad_norm": 1.4979218484984471,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8823936581611633,
      "num_tokens": 254238648.0,
      "step": 7307
    },
    {
      "epoch": 1.3571030640668524,
      "grad_norm": 1.683937215446479,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8736823201179504,
      "num_tokens": 254268302.0,
      "step": 7308
    },
    {
      "epoch": 1.357288765088208,
      "grad_norm": 1.692783015396827,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8731405735015869,
      "num_tokens": 254299697.0,
      "step": 7309
    },
    {
      "epoch": 1.3574744661095637,
      "grad_norm": 1.4868773079981217,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8838486671447754,
      "num_tokens": 254335117.0,
      "step": 7310
    },
    {
      "epoch": 1.3576601671309192,
      "grad_norm": 1.6413859563266326,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8675302863121033,
      "num_tokens": 254369381.0,
      "step": 7311
    },
    {
      "epoch": 1.3578458681522747,
      "grad_norm": 1.4649147635236548,
      "learning_rate": 1e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8569396734237671,
      "num_tokens": 254410918.0,
      "step": 7312
    },
    {
      "epoch": 1.3580315691736304,
      "grad_norm": 1.4364763646592678,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8803688287734985,
      "num_tokens": 254450111.0,
      "step": 7313
    },
    {
      "epoch": 1.3582172701949862,
      "grad_norm": 1.672300113722124,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8692030310630798,
      "num_tokens": 254482787.0,
      "step": 7314
    },
    {
      "epoch": 1.3584029712163417,
      "grad_norm": 1.666184167606896,
      "learning_rate": 1e-06,
      "loss": 0.4011,
      "mean_token_accuracy": 0.8680462837219238,
      "num_tokens": 254513296.0,
      "step": 7315
    },
    {
      "epoch": 1.3585886722376972,
      "grad_norm": 1.626385731858575,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8674923777580261,
      "num_tokens": 254548836.0,
      "step": 7316
    },
    {
      "epoch": 1.358774373259053,
      "grad_norm": 1.4301319861302086,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.88310706615448,
      "num_tokens": 254586533.0,
      "step": 7317
    },
    {
      "epoch": 1.3589600742804087,
      "grad_norm": 1.452906708608014,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8784651160240173,
      "num_tokens": 254623841.0,
      "step": 7318
    },
    {
      "epoch": 1.3591457753017642,
      "grad_norm": 1.4946198384929201,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8689401745796204,
      "num_tokens": 254660948.0,
      "step": 7319
    },
    {
      "epoch": 1.3593314763231197,
      "grad_norm": 1.3921704973239852,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8812214136123657,
      "num_tokens": 254700376.0,
      "step": 7320
    },
    {
      "epoch": 1.3595171773444754,
      "grad_norm": 1.6440420876326007,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8627073168754578,
      "num_tokens": 254729425.0,
      "step": 7321
    },
    {
      "epoch": 1.3597028783658311,
      "grad_norm": 1.5474433977096207,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8829070329666138,
      "num_tokens": 254761804.0,
      "step": 7322
    },
    {
      "epoch": 1.3598885793871867,
      "grad_norm": 1.4738202375409333,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8890565037727356,
      "num_tokens": 254792684.0,
      "step": 7323
    },
    {
      "epoch": 1.3600742804085422,
      "grad_norm": 1.591959214984266,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.856264054775238,
      "num_tokens": 254827295.0,
      "step": 7324
    },
    {
      "epoch": 1.360259981429898,
      "grad_norm": 1.440664133084707,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8764781951904297,
      "num_tokens": 254863420.0,
      "step": 7325
    },
    {
      "epoch": 1.3604456824512534,
      "grad_norm": 1.3421851586909426,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8884338140487671,
      "num_tokens": 254903452.0,
      "step": 7326
    },
    {
      "epoch": 1.3606313834726091,
      "grad_norm": 1.5393139125343676,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8772585391998291,
      "num_tokens": 254937426.0,
      "step": 7327
    },
    {
      "epoch": 1.3608170844939647,
      "grad_norm": 1.456740164107856,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8791303038597107,
      "num_tokens": 254975052.0,
      "step": 7328
    },
    {
      "epoch": 1.3610027855153204,
      "grad_norm": 1.455866731555751,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8876791596412659,
      "num_tokens": 255007831.0,
      "step": 7329
    },
    {
      "epoch": 1.361188486536676,
      "grad_norm": 1.4057165050160945,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8854339122772217,
      "num_tokens": 255041855.0,
      "step": 7330
    },
    {
      "epoch": 1.3613741875580316,
      "grad_norm": 1.557988896940539,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8827986717224121,
      "num_tokens": 255076273.0,
      "step": 7331
    },
    {
      "epoch": 1.3615598885793871,
      "grad_norm": 1.4476072886900826,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8846129179000854,
      "num_tokens": 255112827.0,
      "step": 7332
    },
    {
      "epoch": 1.3617455896007429,
      "grad_norm": 1.5686184048603045,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8751192092895508,
      "num_tokens": 255147552.0,
      "step": 7333
    },
    {
      "epoch": 1.3619312906220984,
      "grad_norm": 1.4709633992479993,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8764239549636841,
      "num_tokens": 255185330.0,
      "step": 7334
    },
    {
      "epoch": 1.362116991643454,
      "grad_norm": 1.4156785613170915,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8896089196205139,
      "num_tokens": 255219847.0,
      "step": 7335
    },
    {
      "epoch": 1.3623026926648096,
      "grad_norm": 1.6236963807720053,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8722556829452515,
      "num_tokens": 255248801.0,
      "step": 7336
    },
    {
      "epoch": 1.3624883936861654,
      "grad_norm": 1.5860771414685069,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8834316730499268,
      "num_tokens": 255280131.0,
      "step": 7337
    },
    {
      "epoch": 1.3626740947075209,
      "grad_norm": 1.5652163519234423,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8815479278564453,
      "num_tokens": 255309373.0,
      "step": 7338
    },
    {
      "epoch": 1.3628597957288764,
      "grad_norm": 1.4410818186344803,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8825508952140808,
      "num_tokens": 255345086.0,
      "step": 7339
    },
    {
      "epoch": 1.3630454967502321,
      "grad_norm": 1.608018401776407,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8755391240119934,
      "num_tokens": 255375756.0,
      "step": 7340
    },
    {
      "epoch": 1.3632311977715879,
      "grad_norm": 1.6247749002002356,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8804492950439453,
      "num_tokens": 255405307.0,
      "step": 7341
    },
    {
      "epoch": 1.3634168987929434,
      "grad_norm": 1.5492332286379829,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8759167194366455,
      "num_tokens": 255438493.0,
      "step": 7342
    },
    {
      "epoch": 1.3636025998142989,
      "grad_norm": 1.5539362187037182,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8911904692649841,
      "num_tokens": 255466910.0,
      "step": 7343
    },
    {
      "epoch": 1.3637883008356546,
      "grad_norm": 1.602960396408011,
      "learning_rate": 1e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.8710353374481201,
      "num_tokens": 255500134.0,
      "step": 7344
    },
    {
      "epoch": 1.3639740018570103,
      "grad_norm": 1.4106714933910385,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8820052742958069,
      "num_tokens": 255537865.0,
      "step": 7345
    },
    {
      "epoch": 1.3641597028783659,
      "grad_norm": 1.4409933101532282,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8811627626419067,
      "num_tokens": 255575072.0,
      "step": 7346
    },
    {
      "epoch": 1.3643454038997214,
      "grad_norm": 1.4170921177850027,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8859843015670776,
      "num_tokens": 255610955.0,
      "step": 7347
    },
    {
      "epoch": 1.364531104921077,
      "grad_norm": 1.5379110959517188,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8762552738189697,
      "num_tokens": 255645061.0,
      "step": 7348
    },
    {
      "epoch": 1.3647168059424326,
      "grad_norm": 1.5233739805978854,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.8788527250289917,
      "num_tokens": 255678038.0,
      "step": 7349
    },
    {
      "epoch": 1.3649025069637883,
      "grad_norm": 1.341243862070498,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8926968574523926,
      "num_tokens": 255717929.0,
      "step": 7350
    },
    {
      "epoch": 1.3650882079851439,
      "grad_norm": 1.4015008623892622,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8717970848083496,
      "num_tokens": 255756375.0,
      "step": 7351
    },
    {
      "epoch": 1.3652739090064996,
      "grad_norm": 1.44905302492339,
      "learning_rate": 1e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8677893877029419,
      "num_tokens": 255797354.0,
      "step": 7352
    },
    {
      "epoch": 1.365459610027855,
      "grad_norm": 1.576579341700432,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8703492879867554,
      "num_tokens": 255828248.0,
      "step": 7353
    },
    {
      "epoch": 1.3656453110492108,
      "grad_norm": 1.5478370754884396,
      "learning_rate": 1e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8569624423980713,
      "num_tokens": 255862871.0,
      "step": 7354
    },
    {
      "epoch": 1.3658310120705663,
      "grad_norm": 1.6386236387763053,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8706402778625488,
      "num_tokens": 255894861.0,
      "step": 7355
    },
    {
      "epoch": 1.366016713091922,
      "grad_norm": 1.3571530863122425,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8768991827964783,
      "num_tokens": 255936522.0,
      "step": 7356
    },
    {
      "epoch": 1.3662024141132776,
      "grad_norm": 1.5614196862832397,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8637343645095825,
      "num_tokens": 255975762.0,
      "step": 7357
    },
    {
      "epoch": 1.3663881151346333,
      "grad_norm": 1.5043873557092842,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8829809427261353,
      "num_tokens": 256012582.0,
      "step": 7358
    },
    {
      "epoch": 1.3665738161559888,
      "grad_norm": 1.5571655275574856,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8856236934661865,
      "num_tokens": 256047670.0,
      "step": 7359
    },
    {
      "epoch": 1.3667595171773446,
      "grad_norm": 1.57846092649077,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8778887987136841,
      "num_tokens": 256080942.0,
      "step": 7360
    },
    {
      "epoch": 1.3669452181987,
      "grad_norm": 1.6016642167077861,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8796550035476685,
      "num_tokens": 256115629.0,
      "step": 7361
    },
    {
      "epoch": 1.3671309192200556,
      "grad_norm": 1.512222466913337,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8685250282287598,
      "num_tokens": 256152228.0,
      "step": 7362
    },
    {
      "epoch": 1.3673166202414113,
      "grad_norm": 1.6197782761170925,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.869396448135376,
      "num_tokens": 256185890.0,
      "step": 7363
    },
    {
      "epoch": 1.367502321262767,
      "grad_norm": 1.47266972422169,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8787633776664734,
      "num_tokens": 256221356.0,
      "step": 7364
    },
    {
      "epoch": 1.3676880222841226,
      "grad_norm": 1.5213863292627663,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8664860725402832,
      "num_tokens": 256257832.0,
      "step": 7365
    },
    {
      "epoch": 1.367873723305478,
      "grad_norm": 1.5215340311246528,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8725626468658447,
      "num_tokens": 256295327.0,
      "step": 7366
    },
    {
      "epoch": 1.3680594243268338,
      "grad_norm": 1.519329568284796,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8766962289810181,
      "num_tokens": 256332736.0,
      "step": 7367
    },
    {
      "epoch": 1.3682451253481895,
      "grad_norm": 1.610897933020123,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8881200551986694,
      "num_tokens": 256360604.0,
      "step": 7368
    },
    {
      "epoch": 1.368430826369545,
      "grad_norm": 1.3491479731771925,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8651521801948547,
      "num_tokens": 256407340.0,
      "step": 7369
    },
    {
      "epoch": 1.3686165273909006,
      "grad_norm": 1.4921824863367894,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8775501251220703,
      "num_tokens": 256439640.0,
      "step": 7370
    },
    {
      "epoch": 1.3688022284122563,
      "grad_norm": 1.4894996896245472,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8744772672653198,
      "num_tokens": 256473641.0,
      "step": 7371
    },
    {
      "epoch": 1.3689879294336118,
      "grad_norm": 1.5541286765239117,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8772169947624207,
      "num_tokens": 256505185.0,
      "step": 7372
    },
    {
      "epoch": 1.3691736304549675,
      "grad_norm": 1.7386255102945596,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8779041767120361,
      "num_tokens": 256537929.0,
      "step": 7373
    },
    {
      "epoch": 1.369359331476323,
      "grad_norm": 1.6186180191902817,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8910745978355408,
      "num_tokens": 256563895.0,
      "step": 7374
    },
    {
      "epoch": 1.3695450324976788,
      "grad_norm": 1.4887500947053751,
      "learning_rate": 1e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8600683212280273,
      "num_tokens": 256602587.0,
      "step": 7375
    },
    {
      "epoch": 1.3697307335190343,
      "grad_norm": 1.5833620490260205,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8747086524963379,
      "num_tokens": 256635403.0,
      "step": 7376
    },
    {
      "epoch": 1.36991643454039,
      "grad_norm": 1.4378704109344833,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.869737982749939,
      "num_tokens": 256677963.0,
      "step": 7377
    },
    {
      "epoch": 1.3701021355617455,
      "grad_norm": 1.4513529179643425,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8746367692947388,
      "num_tokens": 256716022.0,
      "step": 7378
    },
    {
      "epoch": 1.3702878365831013,
      "grad_norm": 1.5287735180381068,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8719836473464966,
      "num_tokens": 256749623.0,
      "step": 7379
    },
    {
      "epoch": 1.3704735376044568,
      "grad_norm": 1.4133183586067988,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8662430047988892,
      "num_tokens": 256790161.0,
      "step": 7380
    },
    {
      "epoch": 1.3706592386258125,
      "grad_norm": 1.5268988473410405,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.875457763671875,
      "num_tokens": 256826612.0,
      "step": 7381
    },
    {
      "epoch": 1.370844939647168,
      "grad_norm": 1.5007040743320605,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8936446905136108,
      "num_tokens": 256858049.0,
      "step": 7382
    },
    {
      "epoch": 1.3710306406685238,
      "grad_norm": 1.5748462741895093,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.866463303565979,
      "num_tokens": 256892970.0,
      "step": 7383
    },
    {
      "epoch": 1.3712163416898793,
      "grad_norm": 1.4861277167773708,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8683112263679504,
      "num_tokens": 256924240.0,
      "step": 7384
    },
    {
      "epoch": 1.3714020427112348,
      "grad_norm": 1.4224447627419183,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8746411800384521,
      "num_tokens": 256960567.0,
      "step": 7385
    },
    {
      "epoch": 1.3715877437325905,
      "grad_norm": 1.5215875227186153,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.867240309715271,
      "num_tokens": 256999447.0,
      "step": 7386
    },
    {
      "epoch": 1.3717734447539462,
      "grad_norm": 1.5221535547962473,
      "learning_rate": 1e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8587198853492737,
      "num_tokens": 257036166.0,
      "step": 7387
    },
    {
      "epoch": 1.3719591457753018,
      "grad_norm": 1.421067288091922,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8845230340957642,
      "num_tokens": 257071832.0,
      "step": 7388
    },
    {
      "epoch": 1.3721448467966573,
      "grad_norm": 1.759761792456008,
      "learning_rate": 1e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8547286987304688,
      "num_tokens": 257101480.0,
      "step": 7389
    },
    {
      "epoch": 1.372330547818013,
      "grad_norm": 1.643076835324714,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8727083206176758,
      "num_tokens": 257133660.0,
      "step": 7390
    },
    {
      "epoch": 1.3725162488393687,
      "grad_norm": 1.439362608999435,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8751978874206543,
      "num_tokens": 257171256.0,
      "step": 7391
    },
    {
      "epoch": 1.3727019498607242,
      "grad_norm": 1.455926565275916,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8776078224182129,
      "num_tokens": 257205266.0,
      "step": 7392
    },
    {
      "epoch": 1.3728876508820798,
      "grad_norm": 1.45883799375033,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8847228288650513,
      "num_tokens": 257241818.0,
      "step": 7393
    },
    {
      "epoch": 1.3730733519034355,
      "grad_norm": 1.5412796552709604,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8815369606018066,
      "num_tokens": 257275742.0,
      "step": 7394
    },
    {
      "epoch": 1.3732590529247912,
      "grad_norm": 1.4527600345492033,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8847582340240479,
      "num_tokens": 257311189.0,
      "step": 7395
    },
    {
      "epoch": 1.3734447539461467,
      "grad_norm": 1.438861158895079,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8747925162315369,
      "num_tokens": 257346228.0,
      "step": 7396
    },
    {
      "epoch": 1.3736304549675022,
      "grad_norm": 1.4989720079845377,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.866387665271759,
      "num_tokens": 257380498.0,
      "step": 7397
    },
    {
      "epoch": 1.373816155988858,
      "grad_norm": 1.5007445907152277,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8896455764770508,
      "num_tokens": 257412877.0,
      "step": 7398
    },
    {
      "epoch": 1.3740018570102135,
      "grad_norm": 1.5305794419904515,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8780386447906494,
      "num_tokens": 257443759.0,
      "step": 7399
    },
    {
      "epoch": 1.3741875580315692,
      "grad_norm": 1.4058003676512671,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8917343616485596,
      "num_tokens": 257482252.0,
      "step": 7400
    },
    {
      "epoch": 1.3743732590529247,
      "grad_norm": 1.5387951171769194,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8734458684921265,
      "num_tokens": 257519855.0,
      "step": 7401
    },
    {
      "epoch": 1.3745589600742805,
      "grad_norm": 1.5674357050997505,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8750960826873779,
      "num_tokens": 257553456.0,
      "step": 7402
    },
    {
      "epoch": 1.374744661095636,
      "grad_norm": 1.3631935519303715,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.887920081615448,
      "num_tokens": 257591657.0,
      "step": 7403
    },
    {
      "epoch": 1.3749303621169917,
      "grad_norm": 1.442866845399256,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8806530237197876,
      "num_tokens": 257625945.0,
      "step": 7404
    },
    {
      "epoch": 1.3751160631383472,
      "grad_norm": 1.4153483929916608,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8878755569458008,
      "num_tokens": 257661236.0,
      "step": 7405
    },
    {
      "epoch": 1.375301764159703,
      "grad_norm": 1.4171308542242078,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8784186840057373,
      "num_tokens": 257697977.0,
      "step": 7406
    },
    {
      "epoch": 1.3754874651810585,
      "grad_norm": 1.6317101507850253,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8765439391136169,
      "num_tokens": 257732189.0,
      "step": 7407
    },
    {
      "epoch": 1.375673166202414,
      "grad_norm": 1.6077267943308748,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.865019679069519,
      "num_tokens": 257766588.0,
      "step": 7408
    },
    {
      "epoch": 1.3758588672237697,
      "grad_norm": 1.4233622220353683,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.8952783346176147,
      "num_tokens": 257801590.0,
      "step": 7409
    },
    {
      "epoch": 1.3760445682451254,
      "grad_norm": 1.4388283519123652,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8743001818656921,
      "num_tokens": 257839900.0,
      "step": 7410
    },
    {
      "epoch": 1.376230269266481,
      "grad_norm": 1.5078908148179995,
      "learning_rate": 1e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8584122657775879,
      "num_tokens": 257879213.0,
      "step": 7411
    },
    {
      "epoch": 1.3764159702878365,
      "grad_norm": 1.4552146519501403,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8824817538261414,
      "num_tokens": 257912212.0,
      "step": 7412
    },
    {
      "epoch": 1.3766016713091922,
      "grad_norm": 1.5935063491394086,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8777226209640503,
      "num_tokens": 257943834.0,
      "step": 7413
    },
    {
      "epoch": 1.376787372330548,
      "grad_norm": 1.4756648735354927,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8730359077453613,
      "num_tokens": 257981078.0,
      "step": 7414
    },
    {
      "epoch": 1.3769730733519034,
      "grad_norm": 1.6902404543248952,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8700695037841797,
      "num_tokens": 258010694.0,
      "step": 7415
    },
    {
      "epoch": 1.377158774373259,
      "grad_norm": 1.4811302461534515,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8880976438522339,
      "num_tokens": 258043483.0,
      "step": 7416
    },
    {
      "epoch": 1.3773444753946147,
      "grad_norm": 1.4133056507413078,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8875914812088013,
      "num_tokens": 258083312.0,
      "step": 7417
    },
    {
      "epoch": 1.3775301764159704,
      "grad_norm": 1.4176055780257064,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8856109380722046,
      "num_tokens": 258121405.0,
      "step": 7418
    },
    {
      "epoch": 1.377715877437326,
      "grad_norm": 1.6352728802798546,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8759382963180542,
      "num_tokens": 258149896.0,
      "step": 7419
    },
    {
      "epoch": 1.3779015784586814,
      "grad_norm": 1.4525472520133058,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8737541437149048,
      "num_tokens": 258185171.0,
      "step": 7420
    },
    {
      "epoch": 1.3780872794800372,
      "grad_norm": 1.6502758598448335,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8749944567680359,
      "num_tokens": 258214562.0,
      "step": 7421
    },
    {
      "epoch": 1.3782729805013927,
      "grad_norm": 1.422186596300351,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8859816789627075,
      "num_tokens": 258250277.0,
      "step": 7422
    },
    {
      "epoch": 1.3784586815227484,
      "grad_norm": 1.5312580722862443,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8840768337249756,
      "num_tokens": 258281841.0,
      "step": 7423
    },
    {
      "epoch": 1.378644382544104,
      "grad_norm": 1.4412176101164107,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8715671300888062,
      "num_tokens": 258320371.0,
      "step": 7424
    },
    {
      "epoch": 1.3788300835654597,
      "grad_norm": 1.336481043030588,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.886277973651886,
      "num_tokens": 258358342.0,
      "step": 7425
    },
    {
      "epoch": 1.3790157845868152,
      "grad_norm": 1.4927141887650424,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8686641454696655,
      "num_tokens": 258395730.0,
      "step": 7426
    },
    {
      "epoch": 1.379201485608171,
      "grad_norm": 1.578621487450194,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8760634660720825,
      "num_tokens": 258424837.0,
      "step": 7427
    },
    {
      "epoch": 1.3793871866295264,
      "grad_norm": 1.6117931618339205,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8734526634216309,
      "num_tokens": 258460429.0,
      "step": 7428
    },
    {
      "epoch": 1.3795728876508822,
      "grad_norm": 1.512126033564224,
      "learning_rate": 1e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.865760326385498,
      "num_tokens": 258497121.0,
      "step": 7429
    },
    {
      "epoch": 1.3797585886722377,
      "grad_norm": 1.570139218209804,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.880010724067688,
      "num_tokens": 258527341.0,
      "step": 7430
    },
    {
      "epoch": 1.3799442896935934,
      "grad_norm": 1.2895780102839725,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8717049956321716,
      "num_tokens": 258569011.0,
      "step": 7431
    },
    {
      "epoch": 1.380129990714949,
      "grad_norm": 1.7879629331462776,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8689113855361938,
      "num_tokens": 258598054.0,
      "step": 7432
    },
    {
      "epoch": 1.3803156917363046,
      "grad_norm": 1.4483401296003606,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8819401860237122,
      "num_tokens": 258632187.0,
      "step": 7433
    },
    {
      "epoch": 1.3805013927576602,
      "grad_norm": 1.5507939017984762,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8714910745620728,
      "num_tokens": 258667117.0,
      "step": 7434
    },
    {
      "epoch": 1.3806870937790157,
      "grad_norm": 1.46563010985859,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8771263360977173,
      "num_tokens": 258702540.0,
      "step": 7435
    },
    {
      "epoch": 1.3808727948003714,
      "grad_norm": 1.5948832436920752,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8826225399971008,
      "num_tokens": 258731076.0,
      "step": 7436
    },
    {
      "epoch": 1.3810584958217271,
      "grad_norm": 1.5780716328721331,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8784522414207458,
      "num_tokens": 258763140.0,
      "step": 7437
    },
    {
      "epoch": 1.3812441968430826,
      "grad_norm": 1.4501146815641313,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8742324113845825,
      "num_tokens": 258803522.0,
      "step": 7438
    },
    {
      "epoch": 1.3814298978644381,
      "grad_norm": 1.4741614309068514,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8851487636566162,
      "num_tokens": 258838362.0,
      "step": 7439
    },
    {
      "epoch": 1.3816155988857939,
      "grad_norm": 1.5194139646656406,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8876076340675354,
      "num_tokens": 258870727.0,
      "step": 7440
    },
    {
      "epoch": 1.3818012999071496,
      "grad_norm": 1.4124529315226237,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.895600438117981,
      "num_tokens": 258905329.0,
      "step": 7441
    },
    {
      "epoch": 1.3819870009285051,
      "grad_norm": 1.742023438378879,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8583448529243469,
      "num_tokens": 258940892.0,
      "step": 7442
    },
    {
      "epoch": 1.3821727019498606,
      "grad_norm": 1.6963321276985701,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8685933351516724,
      "num_tokens": 258970508.0,
      "step": 7443
    },
    {
      "epoch": 1.3823584029712164,
      "grad_norm": 1.5858769274274775,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8764262199401855,
      "num_tokens": 259003745.0,
      "step": 7444
    },
    {
      "epoch": 1.3825441039925719,
      "grad_norm": 1.4455017547989664,
      "learning_rate": 1e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8588746786117554,
      "num_tokens": 259043200.0,
      "step": 7445
    },
    {
      "epoch": 1.3827298050139276,
      "grad_norm": 1.4732616623689676,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8743473291397095,
      "num_tokens": 259076836.0,
      "step": 7446
    },
    {
      "epoch": 1.3829155060352831,
      "grad_norm": 1.4316768341057957,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8846678733825684,
      "num_tokens": 259110998.0,
      "step": 7447
    },
    {
      "epoch": 1.3831012070566389,
      "grad_norm": 1.558251067533884,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8854329586029053,
      "num_tokens": 259144593.0,
      "step": 7448
    },
    {
      "epoch": 1.3832869080779944,
      "grad_norm": 1.4826405913579905,
      "learning_rate": 1e-06,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8762176036834717,
      "num_tokens": 259182962.0,
      "step": 7449
    },
    {
      "epoch": 1.38347260909935,
      "grad_norm": 1.6444024533725494,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8669333457946777,
      "num_tokens": 259214188.0,
      "step": 7450
    },
    {
      "epoch": 1.3836583101207056,
      "grad_norm": 1.6014320128273223,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8717520236968994,
      "num_tokens": 259243592.0,
      "step": 7451
    },
    {
      "epoch": 1.3838440111420613,
      "grad_norm": 1.5622047110865598,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.881964921951294,
      "num_tokens": 259274839.0,
      "step": 7452
    },
    {
      "epoch": 1.3840297121634169,
      "grad_norm": 1.6455940356589442,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8726884126663208,
      "num_tokens": 259305728.0,
      "step": 7453
    },
    {
      "epoch": 1.3842154131847726,
      "grad_norm": 1.5420871015173452,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8728159666061401,
      "num_tokens": 259338545.0,
      "step": 7454
    },
    {
      "epoch": 1.384401114206128,
      "grad_norm": 1.4985662149212446,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.890953540802002,
      "num_tokens": 259370311.0,
      "step": 7455
    },
    {
      "epoch": 1.3845868152274838,
      "grad_norm": 1.435228623309064,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8871496915817261,
      "num_tokens": 259405345.0,
      "step": 7456
    },
    {
      "epoch": 1.3847725162488393,
      "grad_norm": 1.681586248633673,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.8726305961608887,
      "num_tokens": 259435347.0,
      "step": 7457
    },
    {
      "epoch": 1.3849582172701949,
      "grad_norm": 1.6220860296988926,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.868573784828186,
      "num_tokens": 259467664.0,
      "step": 7458
    },
    {
      "epoch": 1.3851439182915506,
      "grad_norm": 1.4982961599281244,
      "learning_rate": 1e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8613630533218384,
      "num_tokens": 259504589.0,
      "step": 7459
    },
    {
      "epoch": 1.3853296193129063,
      "grad_norm": 1.4197283689894071,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8691869378089905,
      "num_tokens": 259546466.0,
      "step": 7460
    },
    {
      "epoch": 1.3855153203342618,
      "grad_norm": 1.5006533105495186,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8765066862106323,
      "num_tokens": 259578076.0,
      "step": 7461
    },
    {
      "epoch": 1.3857010213556173,
      "grad_norm": 1.495979418918524,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.876253604888916,
      "num_tokens": 259612480.0,
      "step": 7462
    },
    {
      "epoch": 1.385886722376973,
      "grad_norm": 1.578815218506182,
      "learning_rate": 1e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8591156005859375,
      "num_tokens": 259646349.0,
      "step": 7463
    },
    {
      "epoch": 1.3860724233983288,
      "grad_norm": 1.537902088286721,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8755313158035278,
      "num_tokens": 259678282.0,
      "step": 7464
    },
    {
      "epoch": 1.3862581244196843,
      "grad_norm": 1.4800045867530887,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8769601583480835,
      "num_tokens": 259713165.0,
      "step": 7465
    },
    {
      "epoch": 1.3864438254410398,
      "grad_norm": 1.5150268943033995,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8797075748443604,
      "num_tokens": 259748342.0,
      "step": 7466
    },
    {
      "epoch": 1.3866295264623956,
      "grad_norm": 1.3889520584410089,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8845436573028564,
      "num_tokens": 259785355.0,
      "step": 7467
    },
    {
      "epoch": 1.386815227483751,
      "grad_norm": 1.445447134156474,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8824888467788696,
      "num_tokens": 259822104.0,
      "step": 7468
    },
    {
      "epoch": 1.3870009285051068,
      "grad_norm": 1.4869915838563332,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.8772526383399963,
      "num_tokens": 259857145.0,
      "step": 7469
    },
    {
      "epoch": 1.3871866295264623,
      "grad_norm": 1.4873903756685725,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8931930065155029,
      "num_tokens": 259890725.0,
      "step": 7470
    },
    {
      "epoch": 1.387372330547818,
      "grad_norm": 1.6592886212279256,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8697673082351685,
      "num_tokens": 259925316.0,
      "step": 7471
    },
    {
      "epoch": 1.3875580315691736,
      "grad_norm": 1.4911894313067204,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.87254399061203,
      "num_tokens": 259961586.0,
      "step": 7472
    },
    {
      "epoch": 1.3877437325905293,
      "grad_norm": 1.4756993764450999,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8731905221939087,
      "num_tokens": 259997651.0,
      "step": 7473
    },
    {
      "epoch": 1.3879294336118848,
      "grad_norm": 1.4812993494404063,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8794437646865845,
      "num_tokens": 260032416.0,
      "step": 7474
    },
    {
      "epoch": 1.3881151346332405,
      "grad_norm": 1.549256986373203,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8880849480628967,
      "num_tokens": 260065442.0,
      "step": 7475
    },
    {
      "epoch": 1.388300835654596,
      "grad_norm": 1.6127661903185329,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8777294754981995,
      "num_tokens": 260097944.0,
      "step": 7476
    },
    {
      "epoch": 1.3884865366759518,
      "grad_norm": 1.4566468756380333,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8690001964569092,
      "num_tokens": 260137900.0,
      "step": 7477
    },
    {
      "epoch": 1.3886722376973073,
      "grad_norm": 1.394065207354998,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8754671216011047,
      "num_tokens": 260178488.0,
      "step": 7478
    },
    {
      "epoch": 1.388857938718663,
      "grad_norm": 1.533541724939808,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8679925203323364,
      "num_tokens": 260214370.0,
      "step": 7479
    },
    {
      "epoch": 1.3890436397400185,
      "grad_norm": 1.4907796214954336,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.887154757976532,
      "num_tokens": 260247812.0,
      "step": 7480
    },
    {
      "epoch": 1.389229340761374,
      "grad_norm": 1.3975272940227719,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8887219429016113,
      "num_tokens": 260282516.0,
      "step": 7481
    },
    {
      "epoch": 1.3894150417827298,
      "grad_norm": 1.3898766242114373,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8754648566246033,
      "num_tokens": 260327461.0,
      "step": 7482
    },
    {
      "epoch": 1.3896007428040855,
      "grad_norm": 1.478246418993376,
      "learning_rate": 1e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8588806390762329,
      "num_tokens": 260364688.0,
      "step": 7483
    },
    {
      "epoch": 1.389786443825441,
      "grad_norm": 1.6163631272483259,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8701379299163818,
      "num_tokens": 260395681.0,
      "step": 7484
    },
    {
      "epoch": 1.3899721448467965,
      "grad_norm": 1.5495623302515757,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8725879788398743,
      "num_tokens": 260429482.0,
      "step": 7485
    },
    {
      "epoch": 1.3901578458681523,
      "grad_norm": 1.651804635239436,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8749277591705322,
      "num_tokens": 260462563.0,
      "step": 7486
    },
    {
      "epoch": 1.390343546889508,
      "grad_norm": 1.6169565520466402,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8821040987968445,
      "num_tokens": 260492785.0,
      "step": 7487
    },
    {
      "epoch": 1.3905292479108635,
      "grad_norm": 1.591110915454897,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8774941563606262,
      "num_tokens": 260526361.0,
      "step": 7488
    },
    {
      "epoch": 1.390714948932219,
      "grad_norm": 1.517383622183404,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.88274085521698,
      "num_tokens": 260561162.0,
      "step": 7489
    },
    {
      "epoch": 1.3909006499535748,
      "grad_norm": 1.5782302009126645,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8696677684783936,
      "num_tokens": 260599121.0,
      "step": 7490
    },
    {
      "epoch": 1.3910863509749305,
      "grad_norm": 1.4826149202614824,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8804914355278015,
      "num_tokens": 260632030.0,
      "step": 7491
    },
    {
      "epoch": 1.391272051996286,
      "grad_norm": 1.5843721801282067,
      "learning_rate": 1e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8620110750198364,
      "num_tokens": 260667985.0,
      "step": 7492
    },
    {
      "epoch": 1.3914577530176415,
      "grad_norm": 1.3880442514719964,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8804357051849365,
      "num_tokens": 260707251.0,
      "step": 7493
    },
    {
      "epoch": 1.3916434540389973,
      "grad_norm": 1.4041704105198392,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8850033283233643,
      "num_tokens": 260747341.0,
      "step": 7494
    },
    {
      "epoch": 1.3918291550603528,
      "grad_norm": 1.5037275302252298,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8675695061683655,
      "num_tokens": 260785576.0,
      "step": 7495
    },
    {
      "epoch": 1.3920148560817085,
      "grad_norm": 1.5201771286542634,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8884854316711426,
      "num_tokens": 260814812.0,
      "step": 7496
    },
    {
      "epoch": 1.392200557103064,
      "grad_norm": 1.4524388220618731,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8763903379440308,
      "num_tokens": 260850062.0,
      "step": 7497
    },
    {
      "epoch": 1.3923862581244197,
      "grad_norm": 1.5368224524730636,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8719052672386169,
      "num_tokens": 260883385.0,
      "step": 7498
    },
    {
      "epoch": 1.3925719591457753,
      "grad_norm": 1.357986456136153,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8894277811050415,
      "num_tokens": 260921057.0,
      "step": 7499
    },
    {
      "epoch": 1.392757660167131,
      "grad_norm": 1.5581882324629721,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8736207485198975,
      "num_tokens": 260955602.0,
      "step": 7500
    },
    {
      "epoch": 1.3929433611884865,
      "grad_norm": 1.455998371160628,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8706303834915161,
      "num_tokens": 260993245.0,
      "step": 7501
    },
    {
      "epoch": 1.3931290622098422,
      "grad_norm": 1.4391029756827083,
      "learning_rate": 1e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.8736830949783325,
      "num_tokens": 261030945.0,
      "step": 7502
    },
    {
      "epoch": 1.3933147632311977,
      "grad_norm": 1.3597015634429814,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8733985424041748,
      "num_tokens": 261071736.0,
      "step": 7503
    },
    {
      "epoch": 1.3935004642525533,
      "grad_norm": 1.442513231044011,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8779608011245728,
      "num_tokens": 261108273.0,
      "step": 7504
    },
    {
      "epoch": 1.393686165273909,
      "grad_norm": 1.3732653643048245,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8826210498809814,
      "num_tokens": 261147524.0,
      "step": 7505
    },
    {
      "epoch": 1.3938718662952647,
      "grad_norm": 1.5232943388134892,
      "learning_rate": 1e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8570538759231567,
      "num_tokens": 261184691.0,
      "step": 7506
    },
    {
      "epoch": 1.3940575673166202,
      "grad_norm": 1.5727283858779608,
      "learning_rate": 1e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.8770771622657776,
      "num_tokens": 261213055.0,
      "step": 7507
    },
    {
      "epoch": 1.3942432683379757,
      "grad_norm": 1.4255103395751645,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8815364241600037,
      "num_tokens": 261247259.0,
      "step": 7508
    },
    {
      "epoch": 1.3944289693593315,
      "grad_norm": 1.6776608714145664,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8735103607177734,
      "num_tokens": 261279359.0,
      "step": 7509
    },
    {
      "epoch": 1.3946146703806872,
      "grad_norm": 1.4896359455020962,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8817506432533264,
      "num_tokens": 261317945.0,
      "step": 7510
    },
    {
      "epoch": 1.3948003714020427,
      "grad_norm": 1.4266967745764951,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.873156726360321,
      "num_tokens": 261355601.0,
      "step": 7511
    },
    {
      "epoch": 1.3949860724233982,
      "grad_norm": 1.4574674352030894,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8753283023834229,
      "num_tokens": 261394411.0,
      "step": 7512
    },
    {
      "epoch": 1.395171773444754,
      "grad_norm": 1.6839902252423669,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8768619298934937,
      "num_tokens": 261420378.0,
      "step": 7513
    },
    {
      "epoch": 1.3953574744661097,
      "grad_norm": 1.5899036778106264,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8617979288101196,
      "num_tokens": 261454609.0,
      "step": 7514
    },
    {
      "epoch": 1.3955431754874652,
      "grad_norm": 1.2920309012180022,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8882446885108948,
      "num_tokens": 261496958.0,
      "step": 7515
    },
    {
      "epoch": 1.3957288765088207,
      "grad_norm": 1.392304140276965,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8750218152999878,
      "num_tokens": 261539947.0,
      "step": 7516
    },
    {
      "epoch": 1.3959145775301764,
      "grad_norm": 1.5133204331426875,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8857175707817078,
      "num_tokens": 261570478.0,
      "step": 7517
    },
    {
      "epoch": 1.396100278551532,
      "grad_norm": 1.5245694582967277,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8802956342697144,
      "num_tokens": 261605304.0,
      "step": 7518
    },
    {
      "epoch": 1.3962859795728877,
      "grad_norm": 1.6718988686540346,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8783851861953735,
      "num_tokens": 261635804.0,
      "step": 7519
    },
    {
      "epoch": 1.3964716805942432,
      "grad_norm": 1.4401591627586605,
      "learning_rate": 1e-06,
      "loss": 0.4152,
      "mean_token_accuracy": 0.8591598868370056,
      "num_tokens": 261674307.0,
      "step": 7520
    },
    {
      "epoch": 1.396657381615599,
      "grad_norm": 1.411731446612128,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8962030410766602,
      "num_tokens": 261707662.0,
      "step": 7521
    },
    {
      "epoch": 1.3968430826369544,
      "grad_norm": 1.4034242823434102,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8793293833732605,
      "num_tokens": 261742261.0,
      "step": 7522
    },
    {
      "epoch": 1.3970287836583102,
      "grad_norm": 1.4692476832619827,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8681968450546265,
      "num_tokens": 261780285.0,
      "step": 7523
    },
    {
      "epoch": 1.3972144846796657,
      "grad_norm": 1.5787344731164314,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8867785930633545,
      "num_tokens": 261810808.0,
      "step": 7524
    },
    {
      "epoch": 1.3974001857010214,
      "grad_norm": 1.4202137616449173,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8926267027854919,
      "num_tokens": 261844913.0,
      "step": 7525
    },
    {
      "epoch": 1.397585886722377,
      "grad_norm": 1.5261331491646062,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8760607242584229,
      "num_tokens": 261879498.0,
      "step": 7526
    },
    {
      "epoch": 1.3977715877437327,
      "grad_norm": 1.389458285986729,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8761206269264221,
      "num_tokens": 261916032.0,
      "step": 7527
    },
    {
      "epoch": 1.3979572887650882,
      "grad_norm": 1.461725536117452,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8813900351524353,
      "num_tokens": 261952101.0,
      "step": 7528
    },
    {
      "epoch": 1.398142989786444,
      "grad_norm": 1.5499047208071284,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8611450791358948,
      "num_tokens": 261984693.0,
      "step": 7529
    },
    {
      "epoch": 1.3983286908077994,
      "grad_norm": 1.5638895495476646,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8849766254425049,
      "num_tokens": 262014897.0,
      "step": 7530
    },
    {
      "epoch": 1.398514391829155,
      "grad_norm": 1.5147753860990205,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.873302698135376,
      "num_tokens": 262050756.0,
      "step": 7531
    },
    {
      "epoch": 1.3987000928505107,
      "grad_norm": 1.3854098456840303,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8887405395507812,
      "num_tokens": 262088254.0,
      "step": 7532
    },
    {
      "epoch": 1.3988857938718664,
      "grad_norm": 1.5373739328782121,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8645849227905273,
      "num_tokens": 262123637.0,
      "step": 7533
    },
    {
      "epoch": 1.399071494893222,
      "grad_norm": 1.606960030611624,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8624122142791748,
      "num_tokens": 262158639.0,
      "step": 7534
    },
    {
      "epoch": 1.3992571959145774,
      "grad_norm": 1.527849777031828,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8846303224563599,
      "num_tokens": 262189904.0,
      "step": 7535
    },
    {
      "epoch": 1.3994428969359332,
      "grad_norm": 1.6687404664568237,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.88275545835495,
      "num_tokens": 262214612.0,
      "step": 7536
    },
    {
      "epoch": 1.399628597957289,
      "grad_norm": 1.525191027476251,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8720414042472839,
      "num_tokens": 262248476.0,
      "step": 7537
    },
    {
      "epoch": 1.3998142989786444,
      "grad_norm": 1.5141153492612973,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.891686201095581,
      "num_tokens": 262282834.0,
      "step": 7538
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.516854810068948,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8769251108169556,
      "num_tokens": 262315599.0,
      "step": 7539
    },
    {
      "epoch": 1.4001857010213556,
      "grad_norm": 1.4130858106376114,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8885470628738403,
      "num_tokens": 262351746.0,
      "step": 7540
    },
    {
      "epoch": 1.4003714020427112,
      "grad_norm": 1.4134953605069234,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8768514394760132,
      "num_tokens": 262388121.0,
      "step": 7541
    },
    {
      "epoch": 1.4005571030640669,
      "grad_norm": 1.5386587529787896,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8810722827911377,
      "num_tokens": 262419613.0,
      "step": 7542
    },
    {
      "epoch": 1.4007428040854224,
      "grad_norm": 1.5936462648506056,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8764438033103943,
      "num_tokens": 262453837.0,
      "step": 7543
    },
    {
      "epoch": 1.4009285051067781,
      "grad_norm": 1.5239100717650973,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8664833307266235,
      "num_tokens": 262490025.0,
      "step": 7544
    },
    {
      "epoch": 1.4011142061281336,
      "grad_norm": 1.5462735967712513,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8818644285202026,
      "num_tokens": 262522907.0,
      "step": 7545
    },
    {
      "epoch": 1.4012999071494894,
      "grad_norm": 1.3854800294888796,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8854176998138428,
      "num_tokens": 262562656.0,
      "step": 7546
    },
    {
      "epoch": 1.4014856081708449,
      "grad_norm": 1.7461974351825837,
      "learning_rate": 1e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8707685470581055,
      "num_tokens": 262591486.0,
      "step": 7547
    },
    {
      "epoch": 1.4016713091922006,
      "grad_norm": 1.56394629095258,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8858519792556763,
      "num_tokens": 262620386.0,
      "step": 7548
    },
    {
      "epoch": 1.4018570102135561,
      "grad_norm": 1.621178871059312,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8668771982192993,
      "num_tokens": 262654769.0,
      "step": 7549
    },
    {
      "epoch": 1.4020427112349119,
      "grad_norm": 1.4847571770825052,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8654506802558899,
      "num_tokens": 262690898.0,
      "step": 7550
    },
    {
      "epoch": 1.4022284122562674,
      "grad_norm": 1.6080605736919673,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8851069211959839,
      "num_tokens": 262722102.0,
      "step": 7551
    },
    {
      "epoch": 1.402414113277623,
      "grad_norm": 1.383304550292262,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8734743595123291,
      "num_tokens": 262764103.0,
      "step": 7552
    },
    {
      "epoch": 1.4025998142989786,
      "grad_norm": 1.562120806984453,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8755955696105957,
      "num_tokens": 262795984.0,
      "step": 7553
    },
    {
      "epoch": 1.4027855153203341,
      "grad_norm": 1.3257904207967333,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8864400386810303,
      "num_tokens": 262835476.0,
      "step": 7554
    },
    {
      "epoch": 1.4029712163416899,
      "grad_norm": 1.3882264101039778,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8822228312492371,
      "num_tokens": 262875856.0,
      "step": 7555
    },
    {
      "epoch": 1.4031569173630456,
      "grad_norm": 1.4912984129144669,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.863579273223877,
      "num_tokens": 262912761.0,
      "step": 7556
    },
    {
      "epoch": 1.403342618384401,
      "grad_norm": 1.5323272507884917,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8652011156082153,
      "num_tokens": 262951169.0,
      "step": 7557
    },
    {
      "epoch": 1.4035283194057566,
      "grad_norm": 1.480998160394074,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8816277980804443,
      "num_tokens": 262992353.0,
      "step": 7558
    },
    {
      "epoch": 1.4037140204271124,
      "grad_norm": 1.48229978925666,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8647184371948242,
      "num_tokens": 263029633.0,
      "step": 7559
    },
    {
      "epoch": 1.403899721448468,
      "grad_norm": 1.4673395745588047,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8889744281768799,
      "num_tokens": 263066408.0,
      "step": 7560
    },
    {
      "epoch": 1.4040854224698236,
      "grad_norm": 1.526389445830271,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8857536315917969,
      "num_tokens": 263096398.0,
      "step": 7561
    },
    {
      "epoch": 1.404271123491179,
      "grad_norm": 1.5926949680011555,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8886157274246216,
      "num_tokens": 263127744.0,
      "step": 7562
    },
    {
      "epoch": 1.4044568245125348,
      "grad_norm": 1.4730875020653462,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8876193165779114,
      "num_tokens": 263162220.0,
      "step": 7563
    },
    {
      "epoch": 1.4046425255338906,
      "grad_norm": 1.524210468936022,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8859496116638184,
      "num_tokens": 263195385.0,
      "step": 7564
    },
    {
      "epoch": 1.404828226555246,
      "grad_norm": 1.5196545384251863,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8764237761497498,
      "num_tokens": 263229427.0,
      "step": 7565
    },
    {
      "epoch": 1.4050139275766016,
      "grad_norm": 1.4864309327975966,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8745156526565552,
      "num_tokens": 263265533.0,
      "step": 7566
    },
    {
      "epoch": 1.4051996285979573,
      "grad_norm": 1.2670506555213679,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8862709999084473,
      "num_tokens": 263312343.0,
      "step": 7567
    },
    {
      "epoch": 1.4053853296193128,
      "grad_norm": 1.6741802135863912,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.871299147605896,
      "num_tokens": 263343892.0,
      "step": 7568
    },
    {
      "epoch": 1.4055710306406686,
      "grad_norm": 1.4849716423123147,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8901110887527466,
      "num_tokens": 263376448.0,
      "step": 7569
    },
    {
      "epoch": 1.405756731662024,
      "grad_norm": 1.5713250317682779,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8730491399765015,
      "num_tokens": 263410927.0,
      "step": 7570
    },
    {
      "epoch": 1.4059424326833798,
      "grad_norm": 1.5915819004961638,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8742992877960205,
      "num_tokens": 263445790.0,
      "step": 7571
    },
    {
      "epoch": 1.4061281337047353,
      "grad_norm": 1.5835742638982198,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8744171857833862,
      "num_tokens": 263475880.0,
      "step": 7572
    },
    {
      "epoch": 1.406313834726091,
      "grad_norm": 1.6698646135752981,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8794568777084351,
      "num_tokens": 263504951.0,
      "step": 7573
    },
    {
      "epoch": 1.4064995357474466,
      "grad_norm": 1.615782430405423,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8767428398132324,
      "num_tokens": 263536624.0,
      "step": 7574
    },
    {
      "epoch": 1.4066852367688023,
      "grad_norm": 1.3480013140326457,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8808242678642273,
      "num_tokens": 263575699.0,
      "step": 7575
    },
    {
      "epoch": 1.4068709377901578,
      "grad_norm": 1.616080255802051,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8727318644523621,
      "num_tokens": 263607730.0,
      "step": 7576
    },
    {
      "epoch": 1.4070566388115133,
      "grad_norm": 1.4552582017460398,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8807715177536011,
      "num_tokens": 263641803.0,
      "step": 7577
    },
    {
      "epoch": 1.407242339832869,
      "grad_norm": 1.556362272576152,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8777104616165161,
      "num_tokens": 263676513.0,
      "step": 7578
    },
    {
      "epoch": 1.4074280408542248,
      "grad_norm": 1.5198798573162178,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8797829151153564,
      "num_tokens": 263710830.0,
      "step": 7579
    },
    {
      "epoch": 1.4076137418755803,
      "grad_norm": 1.3467148505620659,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8874682784080505,
      "num_tokens": 263750505.0,
      "step": 7580
    },
    {
      "epoch": 1.4077994428969358,
      "grad_norm": 1.5589807761979049,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8721261620521545,
      "num_tokens": 263782188.0,
      "step": 7581
    },
    {
      "epoch": 1.4079851439182915,
      "grad_norm": 1.4798737408558715,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.8713949918746948,
      "num_tokens": 263822442.0,
      "step": 7582
    },
    {
      "epoch": 1.4081708449396473,
      "grad_norm": 1.7016785447987273,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8733897805213928,
      "num_tokens": 263851863.0,
      "step": 7583
    },
    {
      "epoch": 1.4083565459610028,
      "grad_norm": 1.452998986619967,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8732413053512573,
      "num_tokens": 263891196.0,
      "step": 7584
    },
    {
      "epoch": 1.4085422469823583,
      "grad_norm": 1.577466549644904,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8721120357513428,
      "num_tokens": 263923907.0,
      "step": 7585
    },
    {
      "epoch": 1.408727948003714,
      "grad_norm": 1.5191889940139804,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8909050226211548,
      "num_tokens": 263957391.0,
      "step": 7586
    },
    {
      "epoch": 1.4089136490250698,
      "grad_norm": 1.634592688371208,
      "learning_rate": 1e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8521797060966492,
      "num_tokens": 263992032.0,
      "step": 7587
    },
    {
      "epoch": 1.4090993500464253,
      "grad_norm": 1.4104884922700969,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8800053596496582,
      "num_tokens": 264030887.0,
      "step": 7588
    },
    {
      "epoch": 1.4092850510677808,
      "grad_norm": 1.5499277384786927,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8790217638015747,
      "num_tokens": 264063789.0,
      "step": 7589
    },
    {
      "epoch": 1.4094707520891365,
      "grad_norm": 1.4739791320508346,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.873732328414917,
      "num_tokens": 264099751.0,
      "step": 7590
    },
    {
      "epoch": 1.409656453110492,
      "grad_norm": 1.334011414351811,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8770747780799866,
      "num_tokens": 264142321.0,
      "step": 7591
    },
    {
      "epoch": 1.4098421541318478,
      "grad_norm": 1.551387237793353,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8795377016067505,
      "num_tokens": 264173707.0,
      "step": 7592
    },
    {
      "epoch": 1.4100278551532033,
      "grad_norm": 1.5664874492054346,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8617541790008545,
      "num_tokens": 264211050.0,
      "step": 7593
    },
    {
      "epoch": 1.410213556174559,
      "grad_norm": 1.7003866629335234,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8754135966300964,
      "num_tokens": 264239670.0,
      "step": 7594
    },
    {
      "epoch": 1.4103992571959145,
      "grad_norm": 1.3951195935840663,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8819364905357361,
      "num_tokens": 264277116.0,
      "step": 7595
    },
    {
      "epoch": 1.4105849582172703,
      "grad_norm": 1.5200166105366917,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.880768358707428,
      "num_tokens": 264311334.0,
      "step": 7596
    },
    {
      "epoch": 1.4107706592386258,
      "grad_norm": 1.367471290164795,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8831923007965088,
      "num_tokens": 264352503.0,
      "step": 7597
    },
    {
      "epoch": 1.4109563602599815,
      "grad_norm": 1.4477695459505677,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8744500875473022,
      "num_tokens": 264388489.0,
      "step": 7598
    },
    {
      "epoch": 1.411142061281337,
      "grad_norm": 1.6277497744837806,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.862889289855957,
      "num_tokens": 264417747.0,
      "step": 7599
    },
    {
      "epoch": 1.4113277623026927,
      "grad_norm": 1.549378134270274,
      "learning_rate": 1e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8599226474761963,
      "num_tokens": 264454241.0,
      "step": 7600
    },
    {
      "epoch": 1.4115134633240483,
      "grad_norm": 1.4585984407317008,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8702452182769775,
      "num_tokens": 264492533.0,
      "step": 7601
    },
    {
      "epoch": 1.411699164345404,
      "grad_norm": 1.6428944484406254,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8835645318031311,
      "num_tokens": 264520555.0,
      "step": 7602
    },
    {
      "epoch": 1.4118848653667595,
      "grad_norm": 1.4844839150087403,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8776201009750366,
      "num_tokens": 264555089.0,
      "step": 7603
    },
    {
      "epoch": 1.412070566388115,
      "grad_norm": 1.5195791907911147,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8765184283256531,
      "num_tokens": 264589131.0,
      "step": 7604
    },
    {
      "epoch": 1.4122562674094707,
      "grad_norm": 1.6137008541669984,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.879313051700592,
      "num_tokens": 264620189.0,
      "step": 7605
    },
    {
      "epoch": 1.4124419684308265,
      "grad_norm": 1.476292517557004,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8769923448562622,
      "num_tokens": 264655145.0,
      "step": 7606
    },
    {
      "epoch": 1.412627669452182,
      "grad_norm": 1.6323490726618397,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.867188572883606,
      "num_tokens": 264687556.0,
      "step": 7607
    },
    {
      "epoch": 1.4128133704735375,
      "grad_norm": 1.3895859109234732,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8931794762611389,
      "num_tokens": 264723523.0,
      "step": 7608
    },
    {
      "epoch": 1.4129990714948932,
      "grad_norm": 1.591411203294152,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8610291481018066,
      "num_tokens": 264755249.0,
      "step": 7609
    },
    {
      "epoch": 1.413184772516249,
      "grad_norm": 1.383047263224287,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8704023361206055,
      "num_tokens": 264798059.0,
      "step": 7610
    },
    {
      "epoch": 1.4133704735376045,
      "grad_norm": 1.503333971796377,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8792295455932617,
      "num_tokens": 264833185.0,
      "step": 7611
    },
    {
      "epoch": 1.41355617455896,
      "grad_norm": 1.5711241996284409,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8729925155639648,
      "num_tokens": 264868788.0,
      "step": 7612
    },
    {
      "epoch": 1.4137418755803157,
      "grad_norm": 1.3405954690608466,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8795786499977112,
      "num_tokens": 264910665.0,
      "step": 7613
    },
    {
      "epoch": 1.4139275766016712,
      "grad_norm": 1.4710611355284082,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8833435773849487,
      "num_tokens": 264946304.0,
      "step": 7614
    },
    {
      "epoch": 1.414113277623027,
      "grad_norm": 1.533477439845634,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8743534088134766,
      "num_tokens": 264977788.0,
      "step": 7615
    },
    {
      "epoch": 1.4142989786443825,
      "grad_norm": 1.3998325360150683,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8744074106216431,
      "num_tokens": 265017044.0,
      "step": 7616
    },
    {
      "epoch": 1.4144846796657382,
      "grad_norm": 1.6063257034217795,
      "learning_rate": 1e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8579821586608887,
      "num_tokens": 265049239.0,
      "step": 7617
    },
    {
      "epoch": 1.4146703806870937,
      "grad_norm": 1.6210283852598473,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8728299140930176,
      "num_tokens": 265079999.0,
      "step": 7618
    },
    {
      "epoch": 1.4148560817084495,
      "grad_norm": 1.5230392122656997,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8793356418609619,
      "num_tokens": 265114346.0,
      "step": 7619
    },
    {
      "epoch": 1.415041782729805,
      "grad_norm": 1.4849422949916562,
      "learning_rate": 1e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8623734712600708,
      "num_tokens": 265150851.0,
      "step": 7620
    },
    {
      "epoch": 1.4152274837511607,
      "grad_norm": 1.5350267073766675,
      "learning_rate": 1e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8675761222839355,
      "num_tokens": 265184380.0,
      "step": 7621
    },
    {
      "epoch": 1.4154131847725162,
      "grad_norm": 1.5016125774042572,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8843769431114197,
      "num_tokens": 265214231.0,
      "step": 7622
    },
    {
      "epoch": 1.415598885793872,
      "grad_norm": 1.6341336429833033,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.8803685307502747,
      "num_tokens": 265243018.0,
      "step": 7623
    },
    {
      "epoch": 1.4157845868152275,
      "grad_norm": 1.4684889368445289,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8766788244247437,
      "num_tokens": 265282064.0,
      "step": 7624
    },
    {
      "epoch": 1.4159702878365832,
      "grad_norm": 1.4411151796408979,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.873674750328064,
      "num_tokens": 265321960.0,
      "step": 7625
    },
    {
      "epoch": 1.4161559888579387,
      "grad_norm": 1.5056485467508876,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.873948872089386,
      "num_tokens": 265354678.0,
      "step": 7626
    },
    {
      "epoch": 1.4163416898792942,
      "grad_norm": 1.451776703142929,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8781079053878784,
      "num_tokens": 265389988.0,
      "step": 7627
    },
    {
      "epoch": 1.41652739090065,
      "grad_norm": 1.6648925124152492,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8744629621505737,
      "num_tokens": 265420333.0,
      "step": 7628
    },
    {
      "epoch": 1.4167130919220057,
      "grad_norm": 1.579726668497791,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8790914416313171,
      "num_tokens": 265452327.0,
      "step": 7629
    },
    {
      "epoch": 1.4168987929433612,
      "grad_norm": 1.683305459077829,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8684224486351013,
      "num_tokens": 265483399.0,
      "step": 7630
    },
    {
      "epoch": 1.4170844939647167,
      "grad_norm": 1.493649282549906,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8906638026237488,
      "num_tokens": 265513454.0,
      "step": 7631
    },
    {
      "epoch": 1.4172701949860724,
      "grad_norm": 1.5273005079474429,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8731741309165955,
      "num_tokens": 265550891.0,
      "step": 7632
    },
    {
      "epoch": 1.4174558960074282,
      "grad_norm": 1.5635136101179081,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8824524283409119,
      "num_tokens": 265580633.0,
      "step": 7633
    },
    {
      "epoch": 1.4176415970287837,
      "grad_norm": 1.721959913971911,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8638226389884949,
      "num_tokens": 265612273.0,
      "step": 7634
    },
    {
      "epoch": 1.4178272980501392,
      "grad_norm": 1.510902234394742,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8820115327835083,
      "num_tokens": 265644817.0,
      "step": 7635
    },
    {
      "epoch": 1.418012999071495,
      "grad_norm": 1.4600587677967447,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8721018433570862,
      "num_tokens": 265681132.0,
      "step": 7636
    },
    {
      "epoch": 1.4181987000928504,
      "grad_norm": 1.4775322148995238,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8793419599533081,
      "num_tokens": 265714956.0,
      "step": 7637
    },
    {
      "epoch": 1.4183844011142062,
      "grad_norm": 1.5270552259488241,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8839117884635925,
      "num_tokens": 265748265.0,
      "step": 7638
    },
    {
      "epoch": 1.4185701021355617,
      "grad_norm": 1.4142925793776537,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8913470506668091,
      "num_tokens": 265780307.0,
      "step": 7639
    },
    {
      "epoch": 1.4187558031569174,
      "grad_norm": 1.4920599852701883,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.875131368637085,
      "num_tokens": 265816710.0,
      "step": 7640
    },
    {
      "epoch": 1.418941504178273,
      "grad_norm": 1.5022702579328324,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8711514472961426,
      "num_tokens": 265855213.0,
      "step": 7641
    },
    {
      "epoch": 1.4191272051996286,
      "grad_norm": 1.4305217158039296,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8771853446960449,
      "num_tokens": 265893049.0,
      "step": 7642
    },
    {
      "epoch": 1.4193129062209842,
      "grad_norm": 1.549558738440304,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8739652633666992,
      "num_tokens": 265924231.0,
      "step": 7643
    },
    {
      "epoch": 1.41949860724234,
      "grad_norm": 1.3380945580505483,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.882844090461731,
      "num_tokens": 265964020.0,
      "step": 7644
    },
    {
      "epoch": 1.4196843082636954,
      "grad_norm": 1.4730195576359373,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8676373958587646,
      "num_tokens": 265999396.0,
      "step": 7645
    },
    {
      "epoch": 1.4198700092850511,
      "grad_norm": 1.4584992421747567,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8863416910171509,
      "num_tokens": 266030675.0,
      "step": 7646
    },
    {
      "epoch": 1.4200557103064066,
      "grad_norm": 1.3592011834106228,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8797667622566223,
      "num_tokens": 266070327.0,
      "step": 7647
    },
    {
      "epoch": 1.4202414113277624,
      "grad_norm": 1.3441379814122263,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8771530389785767,
      "num_tokens": 266114754.0,
      "step": 7648
    },
    {
      "epoch": 1.420427112349118,
      "grad_norm": 1.4456021076397871,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.884072482585907,
      "num_tokens": 266150279.0,
      "step": 7649
    },
    {
      "epoch": 1.4206128133704734,
      "grad_norm": 1.559814490634395,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8766264915466309,
      "num_tokens": 266180376.0,
      "step": 7650
    },
    {
      "epoch": 1.4207985143918291,
      "grad_norm": 1.403972959811544,
      "learning_rate": 1e-06,
      "loss": 0.294,
      "mean_token_accuracy": 0.8977582454681396,
      "num_tokens": 266215813.0,
      "step": 7651
    },
    {
      "epoch": 1.4209842154131849,
      "grad_norm": 1.6866390232128676,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8841313719749451,
      "num_tokens": 266241247.0,
      "step": 7652
    },
    {
      "epoch": 1.4211699164345404,
      "grad_norm": 1.5381724800685996,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8804740905761719,
      "num_tokens": 266275124.0,
      "step": 7653
    },
    {
      "epoch": 1.421355617455896,
      "grad_norm": 1.8515726718186896,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8697484135627747,
      "num_tokens": 266305966.0,
      "step": 7654
    },
    {
      "epoch": 1.4215413184772516,
      "grad_norm": 1.616413528755543,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8744487762451172,
      "num_tokens": 266338074.0,
      "step": 7655
    },
    {
      "epoch": 1.4217270194986074,
      "grad_norm": 1.5593616308954632,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.884484052658081,
      "num_tokens": 266373104.0,
      "step": 7656
    },
    {
      "epoch": 1.4219127205199629,
      "grad_norm": 1.512509130129589,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8764150142669678,
      "num_tokens": 266407252.0,
      "step": 7657
    },
    {
      "epoch": 1.4220984215413184,
      "grad_norm": 1.4627384196334503,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8724199533462524,
      "num_tokens": 266445011.0,
      "step": 7658
    },
    {
      "epoch": 1.4222841225626741,
      "grad_norm": 1.6357660561509944,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8842952251434326,
      "num_tokens": 266472571.0,
      "step": 7659
    },
    {
      "epoch": 1.4224698235840298,
      "grad_norm": 1.6516820084404207,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8747277855873108,
      "num_tokens": 266501117.0,
      "step": 7660
    },
    {
      "epoch": 1.4226555246053854,
      "grad_norm": 1.5499416752920285,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8793495893478394,
      "num_tokens": 266532846.0,
      "step": 7661
    },
    {
      "epoch": 1.4228412256267409,
      "grad_norm": 1.4649580824627146,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8812918663024902,
      "num_tokens": 266573604.0,
      "step": 7662
    },
    {
      "epoch": 1.4230269266480966,
      "grad_norm": 1.4445121390800302,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8717470765113831,
      "num_tokens": 266614670.0,
      "step": 7663
    },
    {
      "epoch": 1.4232126276694521,
      "grad_norm": 1.4529592347341493,
      "learning_rate": 1e-06,
      "loss": 0.3145,
      "mean_token_accuracy": 0.8906185626983643,
      "num_tokens": 266648171.0,
      "step": 7664
    },
    {
      "epoch": 1.4233983286908078,
      "grad_norm": 1.313339054267894,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8865994811058044,
      "num_tokens": 266688780.0,
      "step": 7665
    },
    {
      "epoch": 1.4235840297121634,
      "grad_norm": 1.5060448826969017,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8889837265014648,
      "num_tokens": 266723932.0,
      "step": 7666
    },
    {
      "epoch": 1.423769730733519,
      "grad_norm": 1.5627122690762145,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8753643035888672,
      "num_tokens": 266759123.0,
      "step": 7667
    },
    {
      "epoch": 1.4239554317548746,
      "grad_norm": 1.5081824552595255,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8762121200561523,
      "num_tokens": 266797890.0,
      "step": 7668
    },
    {
      "epoch": 1.4241411327762303,
      "grad_norm": 1.3573887513463858,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8835512399673462,
      "num_tokens": 266840747.0,
      "step": 7669
    },
    {
      "epoch": 1.4243268337975858,
      "grad_norm": 1.5764721947116218,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8708847761154175,
      "num_tokens": 266875932.0,
      "step": 7670
    },
    {
      "epoch": 1.4245125348189416,
      "grad_norm": 1.4997789551474003,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8763774633407593,
      "num_tokens": 266909733.0,
      "step": 7671
    },
    {
      "epoch": 1.424698235840297,
      "grad_norm": 1.438193941295413,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8867048025131226,
      "num_tokens": 266944879.0,
      "step": 7672
    },
    {
      "epoch": 1.4248839368616528,
      "grad_norm": 1.5602158343175363,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8761650323867798,
      "num_tokens": 266980429.0,
      "step": 7673
    },
    {
      "epoch": 1.4250696378830083,
      "grad_norm": 1.6935603827656687,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8691086769104004,
      "num_tokens": 267010711.0,
      "step": 7674
    },
    {
      "epoch": 1.425255338904364,
      "grad_norm": 1.498076385620966,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8799517750740051,
      "num_tokens": 267046581.0,
      "step": 7675
    },
    {
      "epoch": 1.4254410399257196,
      "grad_norm": 1.5890524444498901,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8673993945121765,
      "num_tokens": 267078496.0,
      "step": 7676
    },
    {
      "epoch": 1.425626740947075,
      "grad_norm": 1.5865986528112273,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8697785139083862,
      "num_tokens": 267113689.0,
      "step": 7677
    },
    {
      "epoch": 1.4258124419684308,
      "grad_norm": 1.5828364034810734,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8794974088668823,
      "num_tokens": 267143164.0,
      "step": 7678
    },
    {
      "epoch": 1.4259981429897866,
      "grad_norm": 1.6030854232249836,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8732864856719971,
      "num_tokens": 267177314.0,
      "step": 7679
    },
    {
      "epoch": 1.426183844011142,
      "grad_norm": 1.394758169876266,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8785948753356934,
      "num_tokens": 267213758.0,
      "step": 7680
    },
    {
      "epoch": 1.4263695450324976,
      "grad_norm": 1.5829826061198657,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8846160173416138,
      "num_tokens": 267243665.0,
      "step": 7681
    },
    {
      "epoch": 1.4265552460538533,
      "grad_norm": 1.4242212379237047,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8864389657974243,
      "num_tokens": 267282041.0,
      "step": 7682
    },
    {
      "epoch": 1.426740947075209,
      "grad_norm": 1.568097064257747,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8822165727615356,
      "num_tokens": 267314184.0,
      "step": 7683
    },
    {
      "epoch": 1.4269266480965646,
      "grad_norm": 1.4912716602614244,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8807322978973389,
      "num_tokens": 267347820.0,
      "step": 7684
    },
    {
      "epoch": 1.42711234911792,
      "grad_norm": 1.4853835027455415,
      "learning_rate": 1e-06,
      "loss": 0.4073,
      "mean_token_accuracy": 0.8629902601242065,
      "num_tokens": 267389217.0,
      "step": 7685
    },
    {
      "epoch": 1.4272980501392758,
      "grad_norm": 1.4153591356406512,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8857133388519287,
      "num_tokens": 267425139.0,
      "step": 7686
    },
    {
      "epoch": 1.4274837511606313,
      "grad_norm": 1.5287349613175851,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8810165524482727,
      "num_tokens": 267461720.0,
      "step": 7687
    },
    {
      "epoch": 1.427669452181987,
      "grad_norm": 1.4621035265764724,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8706059455871582,
      "num_tokens": 267499999.0,
      "step": 7688
    },
    {
      "epoch": 1.4278551532033426,
      "grad_norm": 1.4003633979407857,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8734711408615112,
      "num_tokens": 267540420.0,
      "step": 7689
    },
    {
      "epoch": 1.4280408542246983,
      "grad_norm": 1.5622660040563594,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8706125020980835,
      "num_tokens": 267575700.0,
      "step": 7690
    },
    {
      "epoch": 1.4282265552460538,
      "grad_norm": 1.5787676935371937,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8669792413711548,
      "num_tokens": 267607654.0,
      "step": 7691
    },
    {
      "epoch": 1.4284122562674095,
      "grad_norm": 1.4687232421069536,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8799954056739807,
      "num_tokens": 267646095.0,
      "step": 7692
    },
    {
      "epoch": 1.428597957288765,
      "grad_norm": 1.341446678594281,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8801303505897522,
      "num_tokens": 267688636.0,
      "step": 7693
    },
    {
      "epoch": 1.4287836583101208,
      "grad_norm": 1.635704615482297,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.8955820202827454,
      "num_tokens": 267711770.0,
      "step": 7694
    },
    {
      "epoch": 1.4289693593314763,
      "grad_norm": 1.5886529692667615,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8727396726608276,
      "num_tokens": 267744858.0,
      "step": 7695
    },
    {
      "epoch": 1.429155060352832,
      "grad_norm": 1.3776040773642626,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8883795738220215,
      "num_tokens": 267784659.0,
      "step": 7696
    },
    {
      "epoch": 1.4293407613741875,
      "grad_norm": 1.5479648517474933,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8802921175956726,
      "num_tokens": 267817891.0,
      "step": 7697
    },
    {
      "epoch": 1.4295264623955433,
      "grad_norm": 1.4377771335742098,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8720948696136475,
      "num_tokens": 267852709.0,
      "step": 7698
    },
    {
      "epoch": 1.4297121634168988,
      "grad_norm": 1.51121104277487,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8834048509597778,
      "num_tokens": 267883507.0,
      "step": 7699
    },
    {
      "epoch": 1.4298978644382543,
      "grad_norm": 1.5452083480606105,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8704540729522705,
      "num_tokens": 267921570.0,
      "step": 7700
    },
    {
      "epoch": 1.43008356545961,
      "grad_norm": 1.4839408263579124,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8787280917167664,
      "num_tokens": 267957893.0,
      "step": 7701
    },
    {
      "epoch": 1.4302692664809658,
      "grad_norm": 1.514094512991351,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8627933263778687,
      "num_tokens": 267996782.0,
      "step": 7702
    },
    {
      "epoch": 1.4304549675023213,
      "grad_norm": 1.5941794031138075,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8751124143600464,
      "num_tokens": 268028233.0,
      "step": 7703
    },
    {
      "epoch": 1.4306406685236768,
      "grad_norm": 1.5278977629263468,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8702235221862793,
      "num_tokens": 268063089.0,
      "step": 7704
    },
    {
      "epoch": 1.4308263695450325,
      "grad_norm": 1.4326578554453588,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8761582970619202,
      "num_tokens": 268098731.0,
      "step": 7705
    },
    {
      "epoch": 1.4310120705663882,
      "grad_norm": 1.6798285623290035,
      "learning_rate": 1e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8575048446655273,
      "num_tokens": 268132440.0,
      "step": 7706
    },
    {
      "epoch": 1.4311977715877438,
      "grad_norm": 1.400555834778079,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8861980438232422,
      "num_tokens": 268167010.0,
      "step": 7707
    },
    {
      "epoch": 1.4313834726090993,
      "grad_norm": 1.65177646356695,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.86556077003479,
      "num_tokens": 268196549.0,
      "step": 7708
    },
    {
      "epoch": 1.431569173630455,
      "grad_norm": 1.5213419789910592,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8653358817100525,
      "num_tokens": 268234271.0,
      "step": 7709
    },
    {
      "epoch": 1.4317548746518105,
      "grad_norm": 1.536861707373078,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8892823457717896,
      "num_tokens": 268265948.0,
      "step": 7710
    },
    {
      "epoch": 1.4319405756731662,
      "grad_norm": 1.5270078777978662,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8683972358703613,
      "num_tokens": 268300893.0,
      "step": 7711
    },
    {
      "epoch": 1.4321262766945217,
      "grad_norm": 1.506290905495929,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8734235763549805,
      "num_tokens": 268336405.0,
      "step": 7712
    },
    {
      "epoch": 1.4323119777158775,
      "grad_norm": 1.4462116825097269,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.881665825843811,
      "num_tokens": 268372564.0,
      "step": 7713
    },
    {
      "epoch": 1.432497678737233,
      "grad_norm": 1.6611247714271449,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8657100200653076,
      "num_tokens": 268404013.0,
      "step": 7714
    },
    {
      "epoch": 1.4326833797585887,
      "grad_norm": 1.479263393300892,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8819635510444641,
      "num_tokens": 268438235.0,
      "step": 7715
    },
    {
      "epoch": 1.4328690807799442,
      "grad_norm": 1.6607803750903323,
      "learning_rate": 1e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8544365763664246,
      "num_tokens": 268469733.0,
      "step": 7716
    },
    {
      "epoch": 1.4330547818013,
      "grad_norm": 1.3369297094380612,
      "learning_rate": 1e-06,
      "loss": 0.3042,
      "mean_token_accuracy": 0.8922054171562195,
      "num_tokens": 268509705.0,
      "step": 7717
    },
    {
      "epoch": 1.4332404828226555,
      "grad_norm": 1.5785556963854623,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.882473349571228,
      "num_tokens": 268542254.0,
      "step": 7718
    },
    {
      "epoch": 1.4334261838440112,
      "grad_norm": 1.616428230696734,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.874975323677063,
      "num_tokens": 268574144.0,
      "step": 7719
    },
    {
      "epoch": 1.4336118848653667,
      "grad_norm": 1.410092135804799,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8856809139251709,
      "num_tokens": 268611369.0,
      "step": 7720
    },
    {
      "epoch": 1.4337975858867225,
      "grad_norm": 1.5404127239763759,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.864770770072937,
      "num_tokens": 268647422.0,
      "step": 7721
    },
    {
      "epoch": 1.433983286908078,
      "grad_norm": 1.347029744978834,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8854026794433594,
      "num_tokens": 268686913.0,
      "step": 7722
    },
    {
      "epoch": 1.4341689879294335,
      "grad_norm": 1.3594602687520665,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8921819925308228,
      "num_tokens": 268721809.0,
      "step": 7723
    },
    {
      "epoch": 1.4343546889507892,
      "grad_norm": 1.5508412960607243,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8698051571846008,
      "num_tokens": 268753981.0,
      "step": 7724
    },
    {
      "epoch": 1.434540389972145,
      "grad_norm": 1.4823263606199464,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8707147836685181,
      "num_tokens": 268791439.0,
      "step": 7725
    },
    {
      "epoch": 1.4347260909935005,
      "grad_norm": 1.647774596027759,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8688814640045166,
      "num_tokens": 268821560.0,
      "step": 7726
    },
    {
      "epoch": 1.434911792014856,
      "grad_norm": 1.535330645044923,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8730429410934448,
      "num_tokens": 268856333.0,
      "step": 7727
    },
    {
      "epoch": 1.4350974930362117,
      "grad_norm": 1.4510555692253293,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8714299201965332,
      "num_tokens": 268892187.0,
      "step": 7728
    },
    {
      "epoch": 1.4352831940575674,
      "grad_norm": 1.4684591370825246,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8750208616256714,
      "num_tokens": 268928365.0,
      "step": 7729
    },
    {
      "epoch": 1.435468895078923,
      "grad_norm": 1.4534517589599865,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8832492828369141,
      "num_tokens": 268966717.0,
      "step": 7730
    },
    {
      "epoch": 1.4356545961002785,
      "grad_norm": 1.775072579321847,
      "learning_rate": 1e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8687610030174255,
      "num_tokens": 268992502.0,
      "step": 7731
    },
    {
      "epoch": 1.4358402971216342,
      "grad_norm": 1.4549254000393381,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8810045123100281,
      "num_tokens": 269028547.0,
      "step": 7732
    },
    {
      "epoch": 1.43602599814299,
      "grad_norm": 1.5551770561101212,
      "learning_rate": 1e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8613675832748413,
      "num_tokens": 269066508.0,
      "step": 7733
    },
    {
      "epoch": 1.4362116991643454,
      "grad_norm": 1.4685375003020877,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8679357171058655,
      "num_tokens": 269106697.0,
      "step": 7734
    },
    {
      "epoch": 1.436397400185701,
      "grad_norm": 1.7222820191005423,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.876842737197876,
      "num_tokens": 269133890.0,
      "step": 7735
    },
    {
      "epoch": 1.4365831012070567,
      "grad_norm": 1.6202600563556029,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8650168776512146,
      "num_tokens": 269167314.0,
      "step": 7736
    },
    {
      "epoch": 1.4367688022284122,
      "grad_norm": 1.407705719584035,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8771541118621826,
      "num_tokens": 269205348.0,
      "step": 7737
    },
    {
      "epoch": 1.436954503249768,
      "grad_norm": 1.4320159023682195,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8812829852104187,
      "num_tokens": 269244316.0,
      "step": 7738
    },
    {
      "epoch": 1.4371402042711234,
      "grad_norm": 1.5518672481231017,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8706941604614258,
      "num_tokens": 269281273.0,
      "step": 7739
    },
    {
      "epoch": 1.4373259052924792,
      "grad_norm": 1.3611368690810288,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.888526439666748,
      "num_tokens": 269318035.0,
      "step": 7740
    },
    {
      "epoch": 1.4375116063138347,
      "grad_norm": 1.3733664518752429,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8746541738510132,
      "num_tokens": 269357600.0,
      "step": 7741
    },
    {
      "epoch": 1.4376973073351904,
      "grad_norm": 1.4382431999309908,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8643121719360352,
      "num_tokens": 269397505.0,
      "step": 7742
    },
    {
      "epoch": 1.437883008356546,
      "grad_norm": 1.5612481353545378,
      "learning_rate": 1e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8598424196243286,
      "num_tokens": 269432367.0,
      "step": 7743
    },
    {
      "epoch": 1.4380687093779017,
      "grad_norm": 1.5844952010779558,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8680727481842041,
      "num_tokens": 269464355.0,
      "step": 7744
    },
    {
      "epoch": 1.4382544103992572,
      "grad_norm": 1.5877462742637902,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8733797669410706,
      "num_tokens": 269498004.0,
      "step": 7745
    },
    {
      "epoch": 1.4384401114206127,
      "grad_norm": 1.395459303977796,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8703449964523315,
      "num_tokens": 269538665.0,
      "step": 7746
    },
    {
      "epoch": 1.4386258124419684,
      "grad_norm": 1.5781096290291081,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8818053007125854,
      "num_tokens": 269570587.0,
      "step": 7747
    },
    {
      "epoch": 1.4388115134633241,
      "grad_norm": 1.6708949677008136,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8715617060661316,
      "num_tokens": 269598222.0,
      "step": 7748
    },
    {
      "epoch": 1.4389972144846797,
      "grad_norm": 1.3789020672227357,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8825509548187256,
      "num_tokens": 269635055.0,
      "step": 7749
    },
    {
      "epoch": 1.4391829155060352,
      "grad_norm": 1.7346037830542433,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8727253079414368,
      "num_tokens": 269667177.0,
      "step": 7750
    },
    {
      "epoch": 1.439368616527391,
      "grad_norm": 1.5695263084428615,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8934924006462097,
      "num_tokens": 269697978.0,
      "step": 7751
    },
    {
      "epoch": 1.4395543175487466,
      "grad_norm": 1.6060518735107359,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8748379945755005,
      "num_tokens": 269732430.0,
      "step": 7752
    },
    {
      "epoch": 1.4397400185701021,
      "grad_norm": 1.5216944151471399,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8703624606132507,
      "num_tokens": 269768333.0,
      "step": 7753
    },
    {
      "epoch": 1.4399257195914577,
      "grad_norm": 1.398181982210063,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8758150935173035,
      "num_tokens": 269803587.0,
      "step": 7754
    },
    {
      "epoch": 1.4401114206128134,
      "grad_norm": 1.36184909809557,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8774706125259399,
      "num_tokens": 269847795.0,
      "step": 7755
    },
    {
      "epoch": 1.4402971216341691,
      "grad_norm": 1.4302758518545116,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8810855746269226,
      "num_tokens": 269887100.0,
      "step": 7756
    },
    {
      "epoch": 1.4404828226555246,
      "grad_norm": 1.4886143430446717,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8909945487976074,
      "num_tokens": 269920724.0,
      "step": 7757
    },
    {
      "epoch": 1.4406685236768801,
      "grad_norm": 1.736465854241993,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8662464618682861,
      "num_tokens": 269951301.0,
      "step": 7758
    },
    {
      "epoch": 1.4408542246982359,
      "grad_norm": 1.4318168776810427,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.884748101234436,
      "num_tokens": 269988200.0,
      "step": 7759
    },
    {
      "epoch": 1.4410399257195914,
      "grad_norm": 1.683571905985778,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8851935863494873,
      "num_tokens": 270015501.0,
      "step": 7760
    },
    {
      "epoch": 1.4412256267409471,
      "grad_norm": 1.469223687056196,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.8914864659309387,
      "num_tokens": 270050915.0,
      "step": 7761
    },
    {
      "epoch": 1.4414113277623026,
      "grad_norm": 1.72244519535808,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8719707727432251,
      "num_tokens": 270078722.0,
      "step": 7762
    },
    {
      "epoch": 1.4415970287836584,
      "grad_norm": 1.5901342036527637,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8726521134376526,
      "num_tokens": 270113103.0,
      "step": 7763
    },
    {
      "epoch": 1.4417827298050139,
      "grad_norm": 1.5168980380955999,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8813953995704651,
      "num_tokens": 270146857.0,
      "step": 7764
    },
    {
      "epoch": 1.4419684308263696,
      "grad_norm": 1.5423824804906257,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8747332096099854,
      "num_tokens": 270183691.0,
      "step": 7765
    },
    {
      "epoch": 1.4421541318477251,
      "grad_norm": 1.5503120160460302,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8775181174278259,
      "num_tokens": 270214340.0,
      "step": 7766
    },
    {
      "epoch": 1.4423398328690809,
      "grad_norm": 1.4076524213470987,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8803834915161133,
      "num_tokens": 270255003.0,
      "step": 7767
    },
    {
      "epoch": 1.4425255338904364,
      "grad_norm": 1.6044683749805628,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8909010887145996,
      "num_tokens": 270283718.0,
      "step": 7768
    },
    {
      "epoch": 1.442711234911792,
      "grad_norm": 1.5067879428181001,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.882809579372406,
      "num_tokens": 270315934.0,
      "step": 7769
    },
    {
      "epoch": 1.4428969359331476,
      "grad_norm": 1.49763696253935,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8865816593170166,
      "num_tokens": 270350748.0,
      "step": 7770
    },
    {
      "epoch": 1.4430826369545033,
      "grad_norm": 1.3896998355051764,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8736422061920166,
      "num_tokens": 270389518.0,
      "step": 7771
    },
    {
      "epoch": 1.4432683379758589,
      "grad_norm": 1.5997394973264212,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8694076538085938,
      "num_tokens": 270419852.0,
      "step": 7772
    },
    {
      "epoch": 1.4434540389972144,
      "grad_norm": 1.3544902090672726,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8776741027832031,
      "num_tokens": 270462058.0,
      "step": 7773
    },
    {
      "epoch": 1.44363974001857,
      "grad_norm": 1.5714197545426474,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8724845051765442,
      "num_tokens": 270495128.0,
      "step": 7774
    },
    {
      "epoch": 1.4438254410399258,
      "grad_norm": 1.5080022181836787,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.882325291633606,
      "num_tokens": 270529053.0,
      "step": 7775
    },
    {
      "epoch": 1.4440111420612813,
      "grad_norm": 1.3168203718029277,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8924390077590942,
      "num_tokens": 270574590.0,
      "step": 7776
    },
    {
      "epoch": 1.4441968430826368,
      "grad_norm": 1.4925289524151413,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8854163885116577,
      "num_tokens": 270609924.0,
      "step": 7777
    },
    {
      "epoch": 1.4443825441039926,
      "grad_norm": 1.446489156600701,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8816167116165161,
      "num_tokens": 270644730.0,
      "step": 7778
    },
    {
      "epoch": 1.4445682451253483,
      "grad_norm": 1.575101823854283,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8834155201911926,
      "num_tokens": 270673761.0,
      "step": 7779
    },
    {
      "epoch": 1.4447539461467038,
      "grad_norm": 1.3357290591818467,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.878450870513916,
      "num_tokens": 270717317.0,
      "step": 7780
    },
    {
      "epoch": 1.4449396471680593,
      "grad_norm": 1.478139609593874,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8844489455223083,
      "num_tokens": 270750736.0,
      "step": 7781
    },
    {
      "epoch": 1.445125348189415,
      "grad_norm": 1.547407804067831,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8702297806739807,
      "num_tokens": 270787193.0,
      "step": 7782
    },
    {
      "epoch": 1.4453110492107706,
      "grad_norm": 1.645017523599111,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8768980503082275,
      "num_tokens": 270817388.0,
      "step": 7783
    },
    {
      "epoch": 1.4454967502321263,
      "grad_norm": 1.3984841028681658,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8923777341842651,
      "num_tokens": 270853257.0,
      "step": 7784
    },
    {
      "epoch": 1.4456824512534818,
      "grad_norm": 1.3681823393572046,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8925763368606567,
      "num_tokens": 270889337.0,
      "step": 7785
    },
    {
      "epoch": 1.4458681522748376,
      "grad_norm": 1.5442824340170271,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8725560903549194,
      "num_tokens": 270923960.0,
      "step": 7786
    },
    {
      "epoch": 1.446053853296193,
      "grad_norm": 1.5334462172217744,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8883984088897705,
      "num_tokens": 270955503.0,
      "step": 7787
    },
    {
      "epoch": 1.4462395543175488,
      "grad_norm": 1.5566620410576735,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.87396240234375,
      "num_tokens": 270990669.0,
      "step": 7788
    },
    {
      "epoch": 1.4464252553389043,
      "grad_norm": 1.532968490701871,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.8836893439292908,
      "num_tokens": 271021788.0,
      "step": 7789
    },
    {
      "epoch": 1.44661095636026,
      "grad_norm": 1.486055138364338,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8775100708007812,
      "num_tokens": 271056022.0,
      "step": 7790
    },
    {
      "epoch": 1.4467966573816156,
      "grad_norm": 1.4418823837678687,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8817470669746399,
      "num_tokens": 271092679.0,
      "step": 7791
    },
    {
      "epoch": 1.4469823584029713,
      "grad_norm": 1.5754618359161472,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8799878358840942,
      "num_tokens": 271122319.0,
      "step": 7792
    },
    {
      "epoch": 1.4471680594243268,
      "grad_norm": 1.5351706375791725,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8679815530776978,
      "num_tokens": 271160583.0,
      "step": 7793
    },
    {
      "epoch": 1.4473537604456825,
      "grad_norm": 1.472925418339687,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8838303685188293,
      "num_tokens": 271195152.0,
      "step": 7794
    },
    {
      "epoch": 1.447539461467038,
      "grad_norm": 1.5507639308786798,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8734256029129028,
      "num_tokens": 271229649.0,
      "step": 7795
    },
    {
      "epoch": 1.4477251624883936,
      "grad_norm": 1.3893434033952243,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.872979998588562,
      "num_tokens": 271272544.0,
      "step": 7796
    },
    {
      "epoch": 1.4479108635097493,
      "grad_norm": 1.5477087846155253,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.892730712890625,
      "num_tokens": 271300929.0,
      "step": 7797
    },
    {
      "epoch": 1.448096564531105,
      "grad_norm": 1.5959204587829003,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8766062259674072,
      "num_tokens": 271334418.0,
      "step": 7798
    },
    {
      "epoch": 1.4482822655524605,
      "grad_norm": 1.5339115180057254,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.891100287437439,
      "num_tokens": 271364814.0,
      "step": 7799
    },
    {
      "epoch": 1.448467966573816,
      "grad_norm": 1.6950062753218111,
      "learning_rate": 1e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8551331162452698,
      "num_tokens": 271394401.0,
      "step": 7800
    },
    {
      "epoch": 1.4486536675951718,
      "grad_norm": 1.6162282351959387,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8742746114730835,
      "num_tokens": 271423357.0,
      "step": 7801
    },
    {
      "epoch": 1.4488393686165275,
      "grad_norm": 1.6630220536316171,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8670116662979126,
      "num_tokens": 271454011.0,
      "step": 7802
    },
    {
      "epoch": 1.449025069637883,
      "grad_norm": 1.489895068019652,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8693198561668396,
      "num_tokens": 271490662.0,
      "step": 7803
    },
    {
      "epoch": 1.4492107706592385,
      "grad_norm": 1.7345230157750475,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8802167177200317,
      "num_tokens": 271524457.0,
      "step": 7804
    },
    {
      "epoch": 1.4493964716805943,
      "grad_norm": 1.5613167381627013,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8780560493469238,
      "num_tokens": 271557434.0,
      "step": 7805
    },
    {
      "epoch": 1.4495821727019498,
      "grad_norm": 1.4394819987055305,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8730551600456238,
      "num_tokens": 271594982.0,
      "step": 7806
    },
    {
      "epoch": 1.4497678737233055,
      "grad_norm": 1.6114149342746815,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8741852045059204,
      "num_tokens": 271628714.0,
      "step": 7807
    },
    {
      "epoch": 1.449953574744661,
      "grad_norm": 1.5622963302620014,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8653551340103149,
      "num_tokens": 271665529.0,
      "step": 7808
    },
    {
      "epoch": 1.4501392757660168,
      "grad_norm": 1.461606157535553,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8789438009262085,
      "num_tokens": 271700524.0,
      "step": 7809
    },
    {
      "epoch": 1.4503249767873723,
      "grad_norm": 1.6576592009511795,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.876623809337616,
      "num_tokens": 271727574.0,
      "step": 7810
    },
    {
      "epoch": 1.450510677808728,
      "grad_norm": 1.510471155262052,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8762165307998657,
      "num_tokens": 271761668.0,
      "step": 7811
    },
    {
      "epoch": 1.4506963788300835,
      "grad_norm": 1.5241458337512084,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8785170912742615,
      "num_tokens": 271794922.0,
      "step": 7812
    },
    {
      "epoch": 1.4508820798514392,
      "grad_norm": 1.4775009167346744,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8817756175994873,
      "num_tokens": 271829626.0,
      "step": 7813
    },
    {
      "epoch": 1.4510677808727948,
      "grad_norm": 1.4435237753509844,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8803320527076721,
      "num_tokens": 271865369.0,
      "step": 7814
    },
    {
      "epoch": 1.4512534818941505,
      "grad_norm": 1.4471788175409845,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8781226873397827,
      "num_tokens": 271902348.0,
      "step": 7815
    },
    {
      "epoch": 1.451439182915506,
      "grad_norm": 1.5870357417480436,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8802512884140015,
      "num_tokens": 271932832.0,
      "step": 7816
    },
    {
      "epoch": 1.4516248839368617,
      "grad_norm": 1.4360068202457272,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8754597902297974,
      "num_tokens": 271970230.0,
      "step": 7817
    },
    {
      "epoch": 1.4518105849582172,
      "grad_norm": 1.517827133702644,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8679186701774597,
      "num_tokens": 272004452.0,
      "step": 7818
    },
    {
      "epoch": 1.4519962859795728,
      "grad_norm": 1.4401987103261147,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8726156949996948,
      "num_tokens": 272044395.0,
      "step": 7819
    },
    {
      "epoch": 1.4521819870009285,
      "grad_norm": 1.6522822751926094,
      "learning_rate": 1e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8624510765075684,
      "num_tokens": 272075338.0,
      "step": 7820
    },
    {
      "epoch": 1.4523676880222842,
      "grad_norm": 1.5223544183997404,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8731303215026855,
      "num_tokens": 272110416.0,
      "step": 7821
    },
    {
      "epoch": 1.4525533890436397,
      "grad_norm": 1.5115624253756326,
      "learning_rate": 1e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8573311567306519,
      "num_tokens": 272146090.0,
      "step": 7822
    },
    {
      "epoch": 1.4527390900649952,
      "grad_norm": 1.4769035012095404,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8707252144813538,
      "num_tokens": 272184917.0,
      "step": 7823
    },
    {
      "epoch": 1.452924791086351,
      "grad_norm": 1.4746439561046116,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8816528916358948,
      "num_tokens": 272221871.0,
      "step": 7824
    },
    {
      "epoch": 1.4531104921077067,
      "grad_norm": 1.4668057543378625,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8915771842002869,
      "num_tokens": 272252515.0,
      "step": 7825
    },
    {
      "epoch": 1.4532961931290622,
      "grad_norm": 1.5021324271896783,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8800630569458008,
      "num_tokens": 272285548.0,
      "step": 7826
    },
    {
      "epoch": 1.4534818941504177,
      "grad_norm": 1.4953019261558078,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.878874659538269,
      "num_tokens": 272320692.0,
      "step": 7827
    },
    {
      "epoch": 1.4536675951717735,
      "grad_norm": 1.5143435365179079,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8798258900642395,
      "num_tokens": 272355740.0,
      "step": 7828
    },
    {
      "epoch": 1.4538532961931292,
      "grad_norm": 1.6024083146233463,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8790508508682251,
      "num_tokens": 272389777.0,
      "step": 7829
    },
    {
      "epoch": 1.4540389972144847,
      "grad_norm": 1.6338942083730426,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8802560567855835,
      "num_tokens": 272422714.0,
      "step": 7830
    },
    {
      "epoch": 1.4542246982358402,
      "grad_norm": 1.578803043182503,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8689262866973877,
      "num_tokens": 272456032.0,
      "step": 7831
    },
    {
      "epoch": 1.454410399257196,
      "grad_norm": 1.5199500230542182,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8841688632965088,
      "num_tokens": 272487193.0,
      "step": 7832
    },
    {
      "epoch": 1.4545961002785515,
      "grad_norm": 1.4842963296309408,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8905155062675476,
      "num_tokens": 272518967.0,
      "step": 7833
    },
    {
      "epoch": 1.4547818012999072,
      "grad_norm": 1.5386937220384986,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8769205808639526,
      "num_tokens": 272555283.0,
      "step": 7834
    },
    {
      "epoch": 1.4549675023212627,
      "grad_norm": 1.7031024988163315,
      "learning_rate": 1e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8574271202087402,
      "num_tokens": 272585856.0,
      "step": 7835
    },
    {
      "epoch": 1.4551532033426184,
      "grad_norm": 1.7431163768790845,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8855822086334229,
      "num_tokens": 272611399.0,
      "step": 7836
    },
    {
      "epoch": 1.455338904363974,
      "grad_norm": 1.603216341777523,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8729755878448486,
      "num_tokens": 272644292.0,
      "step": 7837
    },
    {
      "epoch": 1.4555246053853297,
      "grad_norm": 1.5215265593920244,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.883979082107544,
      "num_tokens": 272674449.0,
      "step": 7838
    },
    {
      "epoch": 1.4557103064066852,
      "grad_norm": 1.514300421520888,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8726904392242432,
      "num_tokens": 272712083.0,
      "step": 7839
    },
    {
      "epoch": 1.455896007428041,
      "grad_norm": 1.561104437333261,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8750019073486328,
      "num_tokens": 272747865.0,
      "step": 7840
    },
    {
      "epoch": 1.4560817084493964,
      "grad_norm": 1.6004451020807269,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8677366375923157,
      "num_tokens": 272783738.0,
      "step": 7841
    },
    {
      "epoch": 1.4562674094707522,
      "grad_norm": 1.4962745399520723,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8721309900283813,
      "num_tokens": 272820098.0,
      "step": 7842
    },
    {
      "epoch": 1.4564531104921077,
      "grad_norm": 1.4557341684010974,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8722302913665771,
      "num_tokens": 272858334.0,
      "step": 7843
    },
    {
      "epoch": 1.4566388115134634,
      "grad_norm": 1.6300143226116803,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8724719285964966,
      "num_tokens": 272889986.0,
      "step": 7844
    },
    {
      "epoch": 1.456824512534819,
      "grad_norm": 1.6416500100416382,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8767580986022949,
      "num_tokens": 272918212.0,
      "step": 7845
    },
    {
      "epoch": 1.4570102135561744,
      "grad_norm": 1.5049778474178104,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8803505897521973,
      "num_tokens": 272952476.0,
      "step": 7846
    },
    {
      "epoch": 1.4571959145775302,
      "grad_norm": 1.5941388207389091,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8719995617866516,
      "num_tokens": 272988211.0,
      "step": 7847
    },
    {
      "epoch": 1.457381615598886,
      "grad_norm": 1.5199136313305526,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8673367500305176,
      "num_tokens": 273025002.0,
      "step": 7848
    },
    {
      "epoch": 1.4575673166202414,
      "grad_norm": 1.4176915160840868,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.877851665019989,
      "num_tokens": 273063079.0,
      "step": 7849
    },
    {
      "epoch": 1.457753017641597,
      "grad_norm": 1.5318727403841805,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8924564123153687,
      "num_tokens": 273093611.0,
      "step": 7850
    },
    {
      "epoch": 1.4579387186629527,
      "grad_norm": 1.5431311715405982,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8751987814903259,
      "num_tokens": 273125662.0,
      "step": 7851
    },
    {
      "epoch": 1.4581244196843084,
      "grad_norm": 1.578447247220057,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8765754103660583,
      "num_tokens": 273161611.0,
      "step": 7852
    },
    {
      "epoch": 1.458310120705664,
      "grad_norm": 1.5715233049306947,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8740837574005127,
      "num_tokens": 273194315.0,
      "step": 7853
    },
    {
      "epoch": 1.4584958217270194,
      "grad_norm": 1.5897353051498404,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8711800575256348,
      "num_tokens": 273226360.0,
      "step": 7854
    },
    {
      "epoch": 1.4586815227483751,
      "grad_norm": 1.6210903032509218,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8696373105049133,
      "num_tokens": 273258876.0,
      "step": 7855
    },
    {
      "epoch": 1.4588672237697307,
      "grad_norm": 1.659171324116283,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8795082569122314,
      "num_tokens": 273289067.0,
      "step": 7856
    },
    {
      "epoch": 1.4590529247910864,
      "grad_norm": 1.499986458619808,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8716590404510498,
      "num_tokens": 273324008.0,
      "step": 7857
    },
    {
      "epoch": 1.459238625812442,
      "grad_norm": 1.4124025324622307,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.8789481520652771,
      "num_tokens": 273362859.0,
      "step": 7858
    },
    {
      "epoch": 1.4594243268337976,
      "grad_norm": 1.5326059773033813,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8943455219268799,
      "num_tokens": 273395652.0,
      "step": 7859
    },
    {
      "epoch": 1.4596100278551531,
      "grad_norm": 1.4458083232294139,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8818263411521912,
      "num_tokens": 273427493.0,
      "step": 7860
    },
    {
      "epoch": 1.4597957288765089,
      "grad_norm": 1.7426017482115117,
      "learning_rate": 1e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8621851801872253,
      "num_tokens": 273456933.0,
      "step": 7861
    },
    {
      "epoch": 1.4599814298978644,
      "grad_norm": 1.5412460735327844,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.893230676651001,
      "num_tokens": 273489195.0,
      "step": 7862
    },
    {
      "epoch": 1.4601671309192201,
      "grad_norm": 1.4744943621501672,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8734731674194336,
      "num_tokens": 273523734.0,
      "step": 7863
    },
    {
      "epoch": 1.4603528319405756,
      "grad_norm": 1.5122765700664522,
      "learning_rate": 1e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8562582731246948,
      "num_tokens": 273562547.0,
      "step": 7864
    },
    {
      "epoch": 1.4605385329619314,
      "grad_norm": 1.6563807468067504,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.8931844234466553,
      "num_tokens": 273584837.0,
      "step": 7865
    },
    {
      "epoch": 1.4607242339832869,
      "grad_norm": 1.4171507171115736,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8939622640609741,
      "num_tokens": 273621965.0,
      "step": 7866
    },
    {
      "epoch": 1.4609099350046426,
      "grad_norm": 1.5127075993748307,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8690738677978516,
      "num_tokens": 273660752.0,
      "step": 7867
    },
    {
      "epoch": 1.4610956360259981,
      "grad_norm": 1.4698385171456703,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8801562786102295,
      "num_tokens": 273694571.0,
      "step": 7868
    },
    {
      "epoch": 1.4612813370473536,
      "grad_norm": 1.6243060347354437,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8750175833702087,
      "num_tokens": 273730049.0,
      "step": 7869
    },
    {
      "epoch": 1.4614670380687094,
      "grad_norm": 1.5792592452898064,
      "learning_rate": 1e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8570299744606018,
      "num_tokens": 273765916.0,
      "step": 7870
    },
    {
      "epoch": 1.461652739090065,
      "grad_norm": 1.5386295174133258,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8668504953384399,
      "num_tokens": 273799185.0,
      "step": 7871
    },
    {
      "epoch": 1.4618384401114206,
      "grad_norm": 1.439442447095003,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8756374716758728,
      "num_tokens": 273836185.0,
      "step": 7872
    },
    {
      "epoch": 1.4620241411327761,
      "grad_norm": 1.5691865302098704,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8708196878433228,
      "num_tokens": 273867639.0,
      "step": 7873
    },
    {
      "epoch": 1.4622098421541319,
      "grad_norm": 1.55483325563577,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8749011158943176,
      "num_tokens": 273901141.0,
      "step": 7874
    },
    {
      "epoch": 1.4623955431754876,
      "grad_norm": 1.56109741709589,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8821815848350525,
      "num_tokens": 273935064.0,
      "step": 7875
    },
    {
      "epoch": 1.462581244196843,
      "grad_norm": 1.5174562639792901,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8679524660110474,
      "num_tokens": 273969947.0,
      "step": 7876
    },
    {
      "epoch": 1.4627669452181986,
      "grad_norm": 1.4168931943642282,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8803113698959351,
      "num_tokens": 274006469.0,
      "step": 7877
    },
    {
      "epoch": 1.4629526462395543,
      "grad_norm": 1.5430897587299663,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8818585872650146,
      "num_tokens": 274041556.0,
      "step": 7878
    },
    {
      "epoch": 1.4631383472609099,
      "grad_norm": 1.4698854824378962,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8721096515655518,
      "num_tokens": 274079389.0,
      "step": 7879
    },
    {
      "epoch": 1.4633240482822656,
      "grad_norm": 1.6731141314171722,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8844597935676575,
      "num_tokens": 274109401.0,
      "step": 7880
    },
    {
      "epoch": 1.463509749303621,
      "grad_norm": 1.5866385160090273,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8756905198097229,
      "num_tokens": 274141397.0,
      "step": 7881
    },
    {
      "epoch": 1.4636954503249768,
      "grad_norm": 1.5286495860227145,
      "learning_rate": 1e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8519064784049988,
      "num_tokens": 274181961.0,
      "step": 7882
    },
    {
      "epoch": 1.4638811513463323,
      "grad_norm": 1.4324507133733462,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8834004402160645,
      "num_tokens": 274218298.0,
      "step": 7883
    },
    {
      "epoch": 1.464066852367688,
      "grad_norm": 1.6375467897743772,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8893471956253052,
      "num_tokens": 274248715.0,
      "step": 7884
    },
    {
      "epoch": 1.4642525533890436,
      "grad_norm": 1.6191342863691576,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8727912902832031,
      "num_tokens": 274281244.0,
      "step": 7885
    },
    {
      "epoch": 1.4644382544103993,
      "grad_norm": 1.5192968659432295,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8772299289703369,
      "num_tokens": 274316146.0,
      "step": 7886
    },
    {
      "epoch": 1.4646239554317548,
      "grad_norm": 1.6046724515822692,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8790242075920105,
      "num_tokens": 274348497.0,
      "step": 7887
    },
    {
      "epoch": 1.4648096564531106,
      "grad_norm": 1.4881338657268603,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.870617151260376,
      "num_tokens": 274386436.0,
      "step": 7888
    },
    {
      "epoch": 1.464995357474466,
      "grad_norm": 1.4479657121367466,
      "learning_rate": 1e-06,
      "loss": 0.2921,
      "mean_token_accuracy": 0.8971048593521118,
      "num_tokens": 274418424.0,
      "step": 7889
    },
    {
      "epoch": 1.4651810584958218,
      "grad_norm": 1.4671174968004563,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8747309446334839,
      "num_tokens": 274455234.0,
      "step": 7890
    },
    {
      "epoch": 1.4653667595171773,
      "grad_norm": 1.522721101568722,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8848024010658264,
      "num_tokens": 274489465.0,
      "step": 7891
    },
    {
      "epoch": 1.4655524605385328,
      "grad_norm": 1.5120578112832668,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8785674571990967,
      "num_tokens": 274524888.0,
      "step": 7892
    },
    {
      "epoch": 1.4657381615598886,
      "grad_norm": 1.5988720211797416,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8641393184661865,
      "num_tokens": 274556805.0,
      "step": 7893
    },
    {
      "epoch": 1.4659238625812443,
      "grad_norm": 1.4770371767572594,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8868948221206665,
      "num_tokens": 274588431.0,
      "step": 7894
    },
    {
      "epoch": 1.4661095636025998,
      "grad_norm": 1.3957101487720225,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8867161870002747,
      "num_tokens": 274626136.0,
      "step": 7895
    },
    {
      "epoch": 1.4662952646239553,
      "grad_norm": 1.5673361770057885,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8807550668716431,
      "num_tokens": 274658818.0,
      "step": 7896
    },
    {
      "epoch": 1.466480965645311,
      "grad_norm": 1.548820047670697,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8691263198852539,
      "num_tokens": 274692882.0,
      "step": 7897
    },
    {
      "epoch": 1.4666666666666668,
      "grad_norm": 1.658082159812113,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8641959428787231,
      "num_tokens": 274725189.0,
      "step": 7898
    },
    {
      "epoch": 1.4668523676880223,
      "grad_norm": 1.6243220739438116,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8807402849197388,
      "num_tokens": 274756500.0,
      "step": 7899
    },
    {
      "epoch": 1.4670380687093778,
      "grad_norm": 1.6186877689613002,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.870810329914093,
      "num_tokens": 274786828.0,
      "step": 7900
    },
    {
      "epoch": 1.4672237697307335,
      "grad_norm": 1.637456771060307,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8812681436538696,
      "num_tokens": 274815084.0,
      "step": 7901
    },
    {
      "epoch": 1.4674094707520893,
      "grad_norm": 1.4315624034891683,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8736735582351685,
      "num_tokens": 274852797.0,
      "step": 7902
    },
    {
      "epoch": 1.4675951717734448,
      "grad_norm": 1.4687468892411035,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8808590769767761,
      "num_tokens": 274886018.0,
      "step": 7903
    },
    {
      "epoch": 1.4677808727948003,
      "grad_norm": 1.6031153145443693,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.861422598361969,
      "num_tokens": 274919029.0,
      "step": 7904
    },
    {
      "epoch": 1.467966573816156,
      "grad_norm": 1.515917070692791,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8721569776535034,
      "num_tokens": 274953774.0,
      "step": 7905
    },
    {
      "epoch": 1.4681522748375115,
      "grad_norm": 1.3913722969418583,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8837407827377319,
      "num_tokens": 274991545.0,
      "step": 7906
    },
    {
      "epoch": 1.4683379758588673,
      "grad_norm": 1.3953117949907174,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.882888913154602,
      "num_tokens": 275031056.0,
      "step": 7907
    },
    {
      "epoch": 1.4685236768802228,
      "grad_norm": 1.5260532714625092,
      "learning_rate": 1e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8641948103904724,
      "num_tokens": 275068094.0,
      "step": 7908
    },
    {
      "epoch": 1.4687093779015785,
      "grad_norm": 1.306073730467738,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.882887065410614,
      "num_tokens": 275111279.0,
      "step": 7909
    },
    {
      "epoch": 1.468895078922934,
      "grad_norm": 1.5208154146569062,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8761056661605835,
      "num_tokens": 275144974.0,
      "step": 7910
    },
    {
      "epoch": 1.4690807799442898,
      "grad_norm": 1.4428472148521643,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8832625150680542,
      "num_tokens": 275180092.0,
      "step": 7911
    },
    {
      "epoch": 1.4692664809656453,
      "grad_norm": 1.3804828171150432,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8688466548919678,
      "num_tokens": 275220655.0,
      "step": 7912
    },
    {
      "epoch": 1.469452181987001,
      "grad_norm": 1.6143238319315416,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8667848706245422,
      "num_tokens": 275255217.0,
      "step": 7913
    },
    {
      "epoch": 1.4696378830083565,
      "grad_norm": 1.4613049018839166,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8762905597686768,
      "num_tokens": 275292005.0,
      "step": 7914
    },
    {
      "epoch": 1.469823584029712,
      "grad_norm": 1.4834623363365997,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8762587904930115,
      "num_tokens": 275325215.0,
      "step": 7915
    },
    {
      "epoch": 1.4700092850510678,
      "grad_norm": 1.4461626228739173,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8820316791534424,
      "num_tokens": 275360038.0,
      "step": 7916
    },
    {
      "epoch": 1.4701949860724235,
      "grad_norm": 1.4383852245147606,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.882411003112793,
      "num_tokens": 275397449.0,
      "step": 7917
    },
    {
      "epoch": 1.470380687093779,
      "grad_norm": 1.4124484632469947,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8924734592437744,
      "num_tokens": 275430383.0,
      "step": 7918
    },
    {
      "epoch": 1.4705663881151345,
      "grad_norm": 1.4595081407730253,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8723527193069458,
      "num_tokens": 275468396.0,
      "step": 7919
    },
    {
      "epoch": 1.4707520891364902,
      "grad_norm": 1.4454792841471433,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8761109709739685,
      "num_tokens": 275507380.0,
      "step": 7920
    },
    {
      "epoch": 1.470937790157846,
      "grad_norm": 1.6441542629568722,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8663020730018616,
      "num_tokens": 275540905.0,
      "step": 7921
    },
    {
      "epoch": 1.4711234911792015,
      "grad_norm": 1.5689944139399796,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8827215433120728,
      "num_tokens": 275569973.0,
      "step": 7922
    },
    {
      "epoch": 1.471309192200557,
      "grad_norm": 1.527236881825001,
      "learning_rate": 1e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8712774515151978,
      "num_tokens": 275607367.0,
      "step": 7923
    },
    {
      "epoch": 1.4714948932219127,
      "grad_norm": 1.478710771078737,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8643208742141724,
      "num_tokens": 275646373.0,
      "step": 7924
    },
    {
      "epoch": 1.4716805942432685,
      "grad_norm": 1.6309701178054805,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.862122654914856,
      "num_tokens": 275677374.0,
      "step": 7925
    },
    {
      "epoch": 1.471866295264624,
      "grad_norm": 1.7932299796260687,
      "learning_rate": 1e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.854507565498352,
      "num_tokens": 275704131.0,
      "step": 7926
    },
    {
      "epoch": 1.4720519962859795,
      "grad_norm": 1.4934328088619968,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8829814195632935,
      "num_tokens": 275741288.0,
      "step": 7927
    },
    {
      "epoch": 1.4722376973073352,
      "grad_norm": 1.47763740065283,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8706684112548828,
      "num_tokens": 275780058.0,
      "step": 7928
    },
    {
      "epoch": 1.4724233983286907,
      "grad_norm": 1.4276428761091906,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8792921304702759,
      "num_tokens": 275814672.0,
      "step": 7929
    },
    {
      "epoch": 1.4726090993500465,
      "grad_norm": 1.445928486785915,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8747828602790833,
      "num_tokens": 275848220.0,
      "step": 7930
    },
    {
      "epoch": 1.472794800371402,
      "grad_norm": 1.5024194939650326,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8785727024078369,
      "num_tokens": 275881272.0,
      "step": 7931
    },
    {
      "epoch": 1.4729805013927577,
      "grad_norm": 1.503254322692638,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8799785375595093,
      "num_tokens": 275917456.0,
      "step": 7932
    },
    {
      "epoch": 1.4731662024141132,
      "grad_norm": 1.4516031699560605,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8836100101470947,
      "num_tokens": 275951071.0,
      "step": 7933
    },
    {
      "epoch": 1.473351903435469,
      "grad_norm": 1.4617554585062729,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8688303828239441,
      "num_tokens": 275987584.0,
      "step": 7934
    },
    {
      "epoch": 1.4735376044568245,
      "grad_norm": 1.4055576735785833,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8823674917221069,
      "num_tokens": 276022550.0,
      "step": 7935
    },
    {
      "epoch": 1.4737233054781802,
      "grad_norm": 1.3136160885214563,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8754585981369019,
      "num_tokens": 276064035.0,
      "step": 7936
    },
    {
      "epoch": 1.4739090064995357,
      "grad_norm": 1.4666379738235484,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8865823745727539,
      "num_tokens": 276096969.0,
      "step": 7937
    },
    {
      "epoch": 1.4740947075208914,
      "grad_norm": 1.4722321754966685,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8702129125595093,
      "num_tokens": 276128996.0,
      "step": 7938
    },
    {
      "epoch": 1.474280408542247,
      "grad_norm": 1.4728253870271728,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.878288745880127,
      "num_tokens": 276164095.0,
      "step": 7939
    },
    {
      "epoch": 1.4744661095636027,
      "grad_norm": 1.3652804388939306,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.877289891242981,
      "num_tokens": 276202982.0,
      "step": 7940
    },
    {
      "epoch": 1.4746518105849582,
      "grad_norm": 1.6181640623004412,
      "learning_rate": 1e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8481929302215576,
      "num_tokens": 276237433.0,
      "step": 7941
    },
    {
      "epoch": 1.4748375116063137,
      "grad_norm": 1.5374387167017989,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8721417188644409,
      "num_tokens": 276272866.0,
      "step": 7942
    },
    {
      "epoch": 1.4750232126276694,
      "grad_norm": 1.461655318854771,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8709637522697449,
      "num_tokens": 276308603.0,
      "step": 7943
    },
    {
      "epoch": 1.4752089136490252,
      "grad_norm": 1.4239765723144335,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8805859088897705,
      "num_tokens": 276343233.0,
      "step": 7944
    },
    {
      "epoch": 1.4753946146703807,
      "grad_norm": 1.4225168876228276,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8815297484397888,
      "num_tokens": 276384482.0,
      "step": 7945
    },
    {
      "epoch": 1.4755803156917362,
      "grad_norm": 1.4408046210945025,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8663442730903625,
      "num_tokens": 276423021.0,
      "step": 7946
    },
    {
      "epoch": 1.475766016713092,
      "grad_norm": 1.3762411154405874,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8891271352767944,
      "num_tokens": 276463322.0,
      "step": 7947
    },
    {
      "epoch": 1.4759517177344477,
      "grad_norm": 1.5106400501072201,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8885692358016968,
      "num_tokens": 276496665.0,
      "step": 7948
    },
    {
      "epoch": 1.4761374187558032,
      "grad_norm": 1.4498626893348474,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.884617805480957,
      "num_tokens": 276532448.0,
      "step": 7949
    },
    {
      "epoch": 1.4763231197771587,
      "grad_norm": 1.3705208335746688,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8832400441169739,
      "num_tokens": 276571040.0,
      "step": 7950
    },
    {
      "epoch": 1.4765088207985144,
      "grad_norm": 1.4720599247042803,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8777857422828674,
      "num_tokens": 276606303.0,
      "step": 7951
    },
    {
      "epoch": 1.47669452181987,
      "grad_norm": 1.4828906600100353,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8736739158630371,
      "num_tokens": 276640652.0,
      "step": 7952
    },
    {
      "epoch": 1.4768802228412257,
      "grad_norm": 1.446746086765898,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8770766854286194,
      "num_tokens": 276676836.0,
      "step": 7953
    },
    {
      "epoch": 1.4770659238625812,
      "grad_norm": 1.4724809810519424,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8638961315155029,
      "num_tokens": 276719801.0,
      "step": 7954
    },
    {
      "epoch": 1.477251624883937,
      "grad_norm": 1.5606919179531673,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8737095594406128,
      "num_tokens": 276754796.0,
      "step": 7955
    },
    {
      "epoch": 1.4774373259052924,
      "grad_norm": 1.567553416533198,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8697117567062378,
      "num_tokens": 276789043.0,
      "step": 7956
    },
    {
      "epoch": 1.4776230269266482,
      "grad_norm": 1.5998439534271112,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8702144622802734,
      "num_tokens": 276819358.0,
      "step": 7957
    },
    {
      "epoch": 1.4778087279480037,
      "grad_norm": 1.5177429469665071,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8705105781555176,
      "num_tokens": 276855395.0,
      "step": 7958
    },
    {
      "epoch": 1.4779944289693594,
      "grad_norm": 1.4870035812334037,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8829567432403564,
      "num_tokens": 276888646.0,
      "step": 7959
    },
    {
      "epoch": 1.478180129990715,
      "grad_norm": 1.432016402107516,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8933937549591064,
      "num_tokens": 276924519.0,
      "step": 7960
    },
    {
      "epoch": 1.4783658310120706,
      "grad_norm": 1.5324193613657835,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8817651271820068,
      "num_tokens": 276959211.0,
      "step": 7961
    },
    {
      "epoch": 1.4785515320334262,
      "grad_norm": 1.7180905604667562,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8716647624969482,
      "num_tokens": 276987905.0,
      "step": 7962
    },
    {
      "epoch": 1.4787372330547819,
      "grad_norm": 1.4754007396976303,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8821201920509338,
      "num_tokens": 277022311.0,
      "step": 7963
    },
    {
      "epoch": 1.4789229340761374,
      "grad_norm": 1.5740131286616044,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8786988258361816,
      "num_tokens": 277051581.0,
      "step": 7964
    },
    {
      "epoch": 1.479108635097493,
      "grad_norm": 1.6394236576548238,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8695909976959229,
      "num_tokens": 277083840.0,
      "step": 7965
    },
    {
      "epoch": 1.4792943361188486,
      "grad_norm": 1.459079715575612,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8708885908126831,
      "num_tokens": 277120695.0,
      "step": 7966
    },
    {
      "epoch": 1.4794800371402044,
      "grad_norm": 1.5080675762823752,
      "learning_rate": 1e-06,
      "loss": 0.2968,
      "mean_token_accuracy": 0.8961266875267029,
      "num_tokens": 277161614.0,
      "step": 7967
    },
    {
      "epoch": 1.4796657381615599,
      "grad_norm": 1.5179102973316927,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8689997792243958,
      "num_tokens": 277195717.0,
      "step": 7968
    },
    {
      "epoch": 1.4798514391829154,
      "grad_norm": 1.4865982166011964,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8661602735519409,
      "num_tokens": 277231192.0,
      "step": 7969
    },
    {
      "epoch": 1.4800371402042711,
      "grad_norm": 1.51761872488849,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8813514709472656,
      "num_tokens": 277263867.0,
      "step": 7970
    },
    {
      "epoch": 1.4802228412256269,
      "grad_norm": 1.5593456493525821,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8957225680351257,
      "num_tokens": 277293724.0,
      "step": 7971
    },
    {
      "epoch": 1.4804085422469824,
      "grad_norm": 1.531988281638248,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8661211729049683,
      "num_tokens": 277330210.0,
      "step": 7972
    },
    {
      "epoch": 1.4805942432683379,
      "grad_norm": 1.491268601462393,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8858146667480469,
      "num_tokens": 277362009.0,
      "step": 7973
    },
    {
      "epoch": 1.4807799442896936,
      "grad_norm": 1.4928238604926238,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8780943751335144,
      "num_tokens": 277398322.0,
      "step": 7974
    },
    {
      "epoch": 1.4809656453110491,
      "grad_norm": 1.3476553185746731,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8905364274978638,
      "num_tokens": 277434742.0,
      "step": 7975
    },
    {
      "epoch": 1.4811513463324049,
      "grad_norm": 1.537400884444393,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8620448708534241,
      "num_tokens": 277470372.0,
      "step": 7976
    },
    {
      "epoch": 1.4813370473537604,
      "grad_norm": 1.4428074338153036,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8746610283851624,
      "num_tokens": 277506916.0,
      "step": 7977
    },
    {
      "epoch": 1.481522748375116,
      "grad_norm": 1.6137613172843617,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8748966455459595,
      "num_tokens": 277535683.0,
      "step": 7978
    },
    {
      "epoch": 1.4817084493964716,
      "grad_norm": 1.5350736408202434,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8678311109542847,
      "num_tokens": 277570739.0,
      "step": 7979
    },
    {
      "epoch": 1.4818941504178273,
      "grad_norm": 1.4802904379754729,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8788769841194153,
      "num_tokens": 277605322.0,
      "step": 7980
    },
    {
      "epoch": 1.4820798514391829,
      "grad_norm": 1.450300171445163,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.870358943939209,
      "num_tokens": 277646035.0,
      "step": 7981
    },
    {
      "epoch": 1.4822655524605386,
      "grad_norm": 1.554964868067685,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8813724517822266,
      "num_tokens": 277680442.0,
      "step": 7982
    },
    {
      "epoch": 1.482451253481894,
      "grad_norm": 1.5245493042435567,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8866330981254578,
      "num_tokens": 277713554.0,
      "step": 7983
    },
    {
      "epoch": 1.4826369545032498,
      "grad_norm": 1.4941551504818635,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8844866752624512,
      "num_tokens": 277744692.0,
      "step": 7984
    },
    {
      "epoch": 1.4828226555246053,
      "grad_norm": 1.6268831080530046,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8795582056045532,
      "num_tokens": 277775092.0,
      "step": 7985
    },
    {
      "epoch": 1.483008356545961,
      "grad_norm": 1.4953553891052156,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.874299168586731,
      "num_tokens": 277809065.0,
      "step": 7986
    },
    {
      "epoch": 1.4831940575673166,
      "grad_norm": 1.431264166524754,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8802249431610107,
      "num_tokens": 277846426.0,
      "step": 7987
    },
    {
      "epoch": 1.483379758588672,
      "grad_norm": 1.489623439876071,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8738387823104858,
      "num_tokens": 277880287.0,
      "step": 7988
    },
    {
      "epoch": 1.4835654596100278,
      "grad_norm": 1.6807471737794486,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8665380477905273,
      "num_tokens": 277910594.0,
      "step": 7989
    },
    {
      "epoch": 1.4837511606313836,
      "grad_norm": 1.6319244934757133,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8598654270172119,
      "num_tokens": 277943936.0,
      "step": 7990
    },
    {
      "epoch": 1.483936861652739,
      "grad_norm": 1.3590081645194725,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8781437873840332,
      "num_tokens": 277983328.0,
      "step": 7991
    },
    {
      "epoch": 1.4841225626740946,
      "grad_norm": 1.7149518114079834,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8751891255378723,
      "num_tokens": 278015256.0,
      "step": 7992
    },
    {
      "epoch": 1.4843082636954503,
      "grad_norm": 1.5165727812153103,
      "learning_rate": 1e-06,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8764423131942749,
      "num_tokens": 278049666.0,
      "step": 7993
    },
    {
      "epoch": 1.484493964716806,
      "grad_norm": 1.5897998441701082,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8878357410430908,
      "num_tokens": 278076290.0,
      "step": 7994
    },
    {
      "epoch": 1.4846796657381616,
      "grad_norm": 1.5253184488592482,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.887016236782074,
      "num_tokens": 278104870.0,
      "step": 7995
    },
    {
      "epoch": 1.484865366759517,
      "grad_norm": 1.4698038171395902,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8835047483444214,
      "num_tokens": 278142425.0,
      "step": 7996
    },
    {
      "epoch": 1.4850510677808728,
      "grad_norm": 1.7114923774129238,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8624624609947205,
      "num_tokens": 278170510.0,
      "step": 7997
    },
    {
      "epoch": 1.4852367688022285,
      "grad_norm": 1.6381803227900722,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8787178993225098,
      "num_tokens": 278200154.0,
      "step": 7998
    },
    {
      "epoch": 1.485422469823584,
      "grad_norm": 1.4460516724030144,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8800410032272339,
      "num_tokens": 278233886.0,
      "step": 7999
    },
    {
      "epoch": 1.4856081708449396,
      "grad_norm": 1.6509781171016586,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8766788840293884,
      "num_tokens": 278261356.0,
      "step": 8000
    },
    {
      "epoch": 1.4857938718662953,
      "grad_norm": 1.65289043292779,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8705999255180359,
      "num_tokens": 278293685.0,
      "step": 8001
    },
    {
      "epoch": 1.4859795728876508,
      "grad_norm": 1.3988527268877788,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8798421621322632,
      "num_tokens": 278329392.0,
      "step": 8002
    },
    {
      "epoch": 1.4861652739090065,
      "grad_norm": 1.3030763461173915,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8922961354255676,
      "num_tokens": 278368135.0,
      "step": 8003
    },
    {
      "epoch": 1.486350974930362,
      "grad_norm": 1.4789310937921394,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8700500726699829,
      "num_tokens": 278402877.0,
      "step": 8004
    },
    {
      "epoch": 1.4865366759517178,
      "grad_norm": 1.8000243492151529,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8626328110694885,
      "num_tokens": 278430885.0,
      "step": 8005
    },
    {
      "epoch": 1.4867223769730733,
      "grad_norm": 1.4564410762248732,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8773729801177979,
      "num_tokens": 278466935.0,
      "step": 8006
    },
    {
      "epoch": 1.486908077994429,
      "grad_norm": 1.5453598282462926,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8702870607376099,
      "num_tokens": 278501525.0,
      "step": 8007
    },
    {
      "epoch": 1.4870937790157845,
      "grad_norm": 1.3363939348433547,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8823264837265015,
      "num_tokens": 278544868.0,
      "step": 8008
    },
    {
      "epoch": 1.4872794800371403,
      "grad_norm": 1.5196492993918365,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8717353940010071,
      "num_tokens": 278580881.0,
      "step": 8009
    },
    {
      "epoch": 1.4874651810584958,
      "grad_norm": 1.5620361875873694,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8767166137695312,
      "num_tokens": 278613411.0,
      "step": 8010
    },
    {
      "epoch": 1.4876508820798515,
      "grad_norm": 1.442941778555189,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8790494203567505,
      "num_tokens": 278649168.0,
      "step": 8011
    },
    {
      "epoch": 1.487836583101207,
      "grad_norm": 1.5415767189826897,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8833004236221313,
      "num_tokens": 278684734.0,
      "step": 8012
    },
    {
      "epoch": 1.4880222841225628,
      "grad_norm": 1.5686187423077715,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8812910318374634,
      "num_tokens": 278716638.0,
      "step": 8013
    },
    {
      "epoch": 1.4882079851439183,
      "grad_norm": 1.4843361251252678,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8794592618942261,
      "num_tokens": 278749852.0,
      "step": 8014
    },
    {
      "epoch": 1.4883936861652738,
      "grad_norm": 1.6160970461996955,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8759069442749023,
      "num_tokens": 278778219.0,
      "step": 8015
    },
    {
      "epoch": 1.4885793871866295,
      "grad_norm": 1.6233639455840818,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8774799108505249,
      "num_tokens": 278810067.0,
      "step": 8016
    },
    {
      "epoch": 1.4887650882079853,
      "grad_norm": 1.9058421468157423,
      "learning_rate": 1e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8615512847900391,
      "num_tokens": 278837158.0,
      "step": 8017
    },
    {
      "epoch": 1.4889507892293408,
      "grad_norm": 1.4623795194024742,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.870541512966156,
      "num_tokens": 278877466.0,
      "step": 8018
    },
    {
      "epoch": 1.4891364902506963,
      "grad_norm": 1.5386495084017031,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8748141527175903,
      "num_tokens": 278915322.0,
      "step": 8019
    },
    {
      "epoch": 1.489322191272052,
      "grad_norm": 1.3621648758661729,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8642915487289429,
      "num_tokens": 278963050.0,
      "step": 8020
    },
    {
      "epoch": 1.4895078922934077,
      "grad_norm": 1.5926886449254893,
      "learning_rate": 1e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8581262230873108,
      "num_tokens": 279001342.0,
      "step": 8021
    },
    {
      "epoch": 1.4896935933147633,
      "grad_norm": 1.6365537805446477,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8797003030776978,
      "num_tokens": 279033561.0,
      "step": 8022
    },
    {
      "epoch": 1.4898792943361188,
      "grad_norm": 1.691995879352894,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8640176057815552,
      "num_tokens": 279067849.0,
      "step": 8023
    },
    {
      "epoch": 1.4900649953574745,
      "grad_norm": 1.448736783591327,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.883130669593811,
      "num_tokens": 279104323.0,
      "step": 8024
    },
    {
      "epoch": 1.49025069637883,
      "grad_norm": 1.4683268814842267,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8777725696563721,
      "num_tokens": 279137680.0,
      "step": 8025
    },
    {
      "epoch": 1.4904363974001857,
      "grad_norm": 1.490125820930647,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8854432106018066,
      "num_tokens": 279171525.0,
      "step": 8026
    },
    {
      "epoch": 1.4906220984215413,
      "grad_norm": 1.587440255379941,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8789412975311279,
      "num_tokens": 279205178.0,
      "step": 8027
    },
    {
      "epoch": 1.490807799442897,
      "grad_norm": 1.5910764941460742,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8776007890701294,
      "num_tokens": 279236153.0,
      "step": 8028
    },
    {
      "epoch": 1.4909935004642525,
      "grad_norm": 1.4985765462143386,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8660949468612671,
      "num_tokens": 279274916.0,
      "step": 8029
    },
    {
      "epoch": 1.4911792014856082,
      "grad_norm": 1.4423987070709043,
      "learning_rate": 1e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8687914609909058,
      "num_tokens": 279313565.0,
      "step": 8030
    },
    {
      "epoch": 1.4913649025069637,
      "grad_norm": 1.4935507005375968,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8645002841949463,
      "num_tokens": 279352039.0,
      "step": 8031
    },
    {
      "epoch": 1.4915506035283195,
      "grad_norm": 1.6521169227138488,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.884987473487854,
      "num_tokens": 279381406.0,
      "step": 8032
    },
    {
      "epoch": 1.491736304549675,
      "grad_norm": 1.4684661350645927,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8796464204788208,
      "num_tokens": 279415083.0,
      "step": 8033
    },
    {
      "epoch": 1.4919220055710307,
      "grad_norm": 1.5343178781277786,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8740331530570984,
      "num_tokens": 279449132.0,
      "step": 8034
    },
    {
      "epoch": 1.4921077065923862,
      "grad_norm": 1.3943484359321872,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8840138912200928,
      "num_tokens": 279489185.0,
      "step": 8035
    },
    {
      "epoch": 1.492293407613742,
      "grad_norm": 1.5206569958390606,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8858634233474731,
      "num_tokens": 279521061.0,
      "step": 8036
    },
    {
      "epoch": 1.4924791086350975,
      "grad_norm": 1.495242935436816,
      "learning_rate": 1e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8594610691070557,
      "num_tokens": 279559382.0,
      "step": 8037
    },
    {
      "epoch": 1.492664809656453,
      "grad_norm": 1.4917454437048447,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8696877360343933,
      "num_tokens": 279594659.0,
      "step": 8038
    },
    {
      "epoch": 1.4928505106778087,
      "grad_norm": 1.5299647510124026,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8835340738296509,
      "num_tokens": 279624813.0,
      "step": 8039
    },
    {
      "epoch": 1.4930362116991645,
      "grad_norm": 1.3748782281589118,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8939000964164734,
      "num_tokens": 279660700.0,
      "step": 8040
    },
    {
      "epoch": 1.49322191272052,
      "grad_norm": 1.4478966246604363,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.874885082244873,
      "num_tokens": 279699527.0,
      "step": 8041
    },
    {
      "epoch": 1.4934076137418755,
      "grad_norm": 1.537638933339997,
      "learning_rate": 1e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8677012920379639,
      "num_tokens": 279734965.0,
      "step": 8042
    },
    {
      "epoch": 1.4935933147632312,
      "grad_norm": 1.5872371745153404,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8798907995223999,
      "num_tokens": 279772106.0,
      "step": 8043
    },
    {
      "epoch": 1.493779015784587,
      "grad_norm": 1.3714996311016892,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.878070592880249,
      "num_tokens": 279809821.0,
      "step": 8044
    },
    {
      "epoch": 1.4939647168059424,
      "grad_norm": 1.5827786019457533,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8786733150482178,
      "num_tokens": 279841203.0,
      "step": 8045
    },
    {
      "epoch": 1.494150417827298,
      "grad_norm": 1.4435829776278322,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8907122611999512,
      "num_tokens": 279877742.0,
      "step": 8046
    },
    {
      "epoch": 1.4943361188486537,
      "grad_norm": 1.4946200744124922,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8729640245437622,
      "num_tokens": 279911782.0,
      "step": 8047
    },
    {
      "epoch": 1.4945218198700092,
      "grad_norm": 1.458234929541749,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8786730766296387,
      "num_tokens": 279951603.0,
      "step": 8048
    },
    {
      "epoch": 1.494707520891365,
      "grad_norm": 1.5840298258497552,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8624232411384583,
      "num_tokens": 279987340.0,
      "step": 8049
    },
    {
      "epoch": 1.4948932219127204,
      "grad_norm": 1.4916180301714212,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8823648691177368,
      "num_tokens": 280018522.0,
      "step": 8050
    },
    {
      "epoch": 1.4950789229340762,
      "grad_norm": 1.4413713065799283,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8709683418273926,
      "num_tokens": 280057180.0,
      "step": 8051
    },
    {
      "epoch": 1.4952646239554317,
      "grad_norm": 1.4985684334008398,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8744664192199707,
      "num_tokens": 280091564.0,
      "step": 8052
    },
    {
      "epoch": 1.4954503249767874,
      "grad_norm": 1.4935363540959923,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8774034976959229,
      "num_tokens": 280125232.0,
      "step": 8053
    },
    {
      "epoch": 1.495636025998143,
      "grad_norm": 1.4952278496846234,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.879662036895752,
      "num_tokens": 280164020.0,
      "step": 8054
    },
    {
      "epoch": 1.4958217270194987,
      "grad_norm": 1.8632513511857898,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8779938817024231,
      "num_tokens": 280187898.0,
      "step": 8055
    },
    {
      "epoch": 1.4960074280408542,
      "grad_norm": 1.3569857374814607,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8888049125671387,
      "num_tokens": 280228857.0,
      "step": 8056
    },
    {
      "epoch": 1.49619312906221,
      "grad_norm": 1.469136587975899,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8775894641876221,
      "num_tokens": 280262917.0,
      "step": 8057
    },
    {
      "epoch": 1.4963788300835654,
      "grad_norm": 1.6297290388065326,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8757753372192383,
      "num_tokens": 280291464.0,
      "step": 8058
    },
    {
      "epoch": 1.4965645311049212,
      "grad_norm": 1.4382726395577052,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8790603876113892,
      "num_tokens": 280329161.0,
      "step": 8059
    },
    {
      "epoch": 1.4967502321262767,
      "grad_norm": 1.4851756606403481,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8900423049926758,
      "num_tokens": 280361850.0,
      "step": 8060
    },
    {
      "epoch": 1.4969359331476322,
      "grad_norm": 1.4785324122132406,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8871693015098572,
      "num_tokens": 280395064.0,
      "step": 8061
    },
    {
      "epoch": 1.497121634168988,
      "grad_norm": 1.5856961435424086,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8694130182266235,
      "num_tokens": 280430194.0,
      "step": 8062
    },
    {
      "epoch": 1.4973073351903436,
      "grad_norm": 1.4572753191629153,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8840404748916626,
      "num_tokens": 280465332.0,
      "step": 8063
    },
    {
      "epoch": 1.4974930362116992,
      "grad_norm": 1.4843596104645398,
      "learning_rate": 1e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8644943237304688,
      "num_tokens": 280500969.0,
      "step": 8064
    },
    {
      "epoch": 1.4976787372330547,
      "grad_norm": 1.5684517620607792,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8736352324485779,
      "num_tokens": 280533132.0,
      "step": 8065
    },
    {
      "epoch": 1.4978644382544104,
      "grad_norm": 1.524473412712821,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8701601624488831,
      "num_tokens": 280565071.0,
      "step": 8066
    },
    {
      "epoch": 1.4980501392757661,
      "grad_norm": 1.3579937729606395,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8776119947433472,
      "num_tokens": 280606045.0,
      "step": 8067
    },
    {
      "epoch": 1.4982358402971216,
      "grad_norm": 1.307046125346997,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8841527700424194,
      "num_tokens": 280650695.0,
      "step": 8068
    },
    {
      "epoch": 1.4984215413184772,
      "grad_norm": 1.5814100660133394,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.868903636932373,
      "num_tokens": 280684437.0,
      "step": 8069
    },
    {
      "epoch": 1.498607242339833,
      "grad_norm": 1.6917165776467664,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8762559294700623,
      "num_tokens": 280714182.0,
      "step": 8070
    },
    {
      "epoch": 1.4987929433611886,
      "grad_norm": 1.5196898772133778,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8684045672416687,
      "num_tokens": 280750295.0,
      "step": 8071
    },
    {
      "epoch": 1.4989786443825441,
      "grad_norm": 1.5348357489435676,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8667088150978088,
      "num_tokens": 280785647.0,
      "step": 8072
    },
    {
      "epoch": 1.4991643454038996,
      "grad_norm": 1.752078205680478,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8798166513442993,
      "num_tokens": 280812615.0,
      "step": 8073
    },
    {
      "epoch": 1.4993500464252554,
      "grad_norm": 1.4145320046243814,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8908544778823853,
      "num_tokens": 280850904.0,
      "step": 8074
    },
    {
      "epoch": 1.499535747446611,
      "grad_norm": 1.474610021689617,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8910764455795288,
      "num_tokens": 280882916.0,
      "step": 8075
    },
    {
      "epoch": 1.4997214484679666,
      "grad_norm": 1.4093719568895553,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8786667585372925,
      "num_tokens": 280921399.0,
      "step": 8076
    },
    {
      "epoch": 1.4999071494893221,
      "grad_norm": 1.4117820410525483,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8824757933616638,
      "num_tokens": 280958430.0,
      "step": 8077
    },
    {
      "epoch": 1.5000928505106779,
      "grad_norm": 1.563128898244267,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8833826780319214,
      "num_tokens": 280992584.0,
      "step": 8078
    },
    {
      "epoch": 1.5002785515320334,
      "grad_norm": 1.5191413779746374,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8778821229934692,
      "num_tokens": 281030746.0,
      "step": 8079
    },
    {
      "epoch": 1.5004642525533889,
      "grad_norm": 1.3394053382046418,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.8952038288116455,
      "num_tokens": 281065814.0,
      "step": 8080
    },
    {
      "epoch": 1.5006499535747446,
      "grad_norm": 1.4071213617771179,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.870846688747406,
      "num_tokens": 281106039.0,
      "step": 8081
    },
    {
      "epoch": 1.5008356545961004,
      "grad_norm": 1.5820038689258633,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8760563135147095,
      "num_tokens": 281140060.0,
      "step": 8082
    },
    {
      "epoch": 1.5010213556174559,
      "grad_norm": 1.4819176516318395,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8665308952331543,
      "num_tokens": 281176973.0,
      "step": 8083
    },
    {
      "epoch": 1.5012070566388114,
      "grad_norm": 1.3374324458871196,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.885710597038269,
      "num_tokens": 281217738.0,
      "step": 8084
    },
    {
      "epoch": 1.501392757660167,
      "grad_norm": 1.4605486620986097,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8894282579421997,
      "num_tokens": 281252504.0,
      "step": 8085
    },
    {
      "epoch": 1.5015784586815228,
      "grad_norm": 1.4908146912385942,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8742603063583374,
      "num_tokens": 281286173.0,
      "step": 8086
    },
    {
      "epoch": 1.5017641597028784,
      "grad_norm": 1.4180487170566438,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8715658187866211,
      "num_tokens": 281328685.0,
      "step": 8087
    },
    {
      "epoch": 1.5019498607242339,
      "grad_norm": 1.590053596815923,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.88186115026474,
      "num_tokens": 281356519.0,
      "step": 8088
    },
    {
      "epoch": 1.5021355617455896,
      "grad_norm": 1.511355099639656,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8863310813903809,
      "num_tokens": 281389638.0,
      "step": 8089
    },
    {
      "epoch": 1.5023212627669453,
      "grad_norm": 1.541404869615924,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8771159648895264,
      "num_tokens": 281422811.0,
      "step": 8090
    },
    {
      "epoch": 1.5025069637883008,
      "grad_norm": 1.633443724569242,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.869512677192688,
      "num_tokens": 281455329.0,
      "step": 8091
    },
    {
      "epoch": 1.5026926648096564,
      "grad_norm": 1.5487277201914857,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8793919086456299,
      "num_tokens": 281487215.0,
      "step": 8092
    },
    {
      "epoch": 1.502878365831012,
      "grad_norm": 1.4387384457259422,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8749992847442627,
      "num_tokens": 281526920.0,
      "step": 8093
    },
    {
      "epoch": 1.5030640668523678,
      "grad_norm": 1.4563634748879823,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8648232221603394,
      "num_tokens": 281566434.0,
      "step": 8094
    },
    {
      "epoch": 1.5032497678737233,
      "grad_norm": 1.6242437737720317,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8800402879714966,
      "num_tokens": 281599648.0,
      "step": 8095
    },
    {
      "epoch": 1.5034354688950788,
      "grad_norm": 1.4613136260842532,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8778990507125854,
      "num_tokens": 281634516.0,
      "step": 8096
    },
    {
      "epoch": 1.5036211699164346,
      "grad_norm": 1.6041994566460929,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8645811080932617,
      "num_tokens": 281666314.0,
      "step": 8097
    },
    {
      "epoch": 1.5038068709377903,
      "grad_norm": 1.4739776923228793,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8897839188575745,
      "num_tokens": 281699010.0,
      "step": 8098
    },
    {
      "epoch": 1.5039925719591458,
      "grad_norm": 1.4720780409742555,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.883126437664032,
      "num_tokens": 281732956.0,
      "step": 8099
    },
    {
      "epoch": 1.5041782729805013,
      "grad_norm": 1.4384608137843902,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8761700391769409,
      "num_tokens": 281766887.0,
      "step": 8100
    },
    {
      "epoch": 1.504363974001857,
      "grad_norm": 1.5841567066616375,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8721665143966675,
      "num_tokens": 281798051.0,
      "step": 8101
    },
    {
      "epoch": 1.5045496750232126,
      "grad_norm": 1.550182323251779,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8758817911148071,
      "num_tokens": 281830491.0,
      "step": 8102
    },
    {
      "epoch": 1.504735376044568,
      "grad_norm": 1.5228809108672259,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.875739574432373,
      "num_tokens": 281863357.0,
      "step": 8103
    },
    {
      "epoch": 1.5049210770659238,
      "grad_norm": 1.6710804088949074,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8731045722961426,
      "num_tokens": 281892797.0,
      "step": 8104
    },
    {
      "epoch": 1.5051067780872796,
      "grad_norm": 1.5507277134798825,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8685745000839233,
      "num_tokens": 281928428.0,
      "step": 8105
    },
    {
      "epoch": 1.505292479108635,
      "grad_norm": 1.4346640859927433,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8769057393074036,
      "num_tokens": 281965991.0,
      "step": 8106
    },
    {
      "epoch": 1.5054781801299906,
      "grad_norm": 1.6786418005028536,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8860763907432556,
      "num_tokens": 281993051.0,
      "step": 8107
    },
    {
      "epoch": 1.5056638811513463,
      "grad_norm": 1.5169309689168533,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8586146831512451,
      "num_tokens": 282032805.0,
      "step": 8108
    },
    {
      "epoch": 1.505849582172702,
      "grad_norm": 1.366749237721634,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8815295696258545,
      "num_tokens": 282075150.0,
      "step": 8109
    },
    {
      "epoch": 1.5060352831940576,
      "grad_norm": 1.4996072515748915,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.877139687538147,
      "num_tokens": 282110963.0,
      "step": 8110
    },
    {
      "epoch": 1.506220984215413,
      "grad_norm": 1.4734640331123483,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8790755271911621,
      "num_tokens": 282146454.0,
      "step": 8111
    },
    {
      "epoch": 1.5064066852367688,
      "grad_norm": 1.4775972312011738,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8740615844726562,
      "num_tokens": 282182452.0,
      "step": 8112
    },
    {
      "epoch": 1.5065923862581245,
      "grad_norm": 1.374751046602294,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8765243291854858,
      "num_tokens": 282224399.0,
      "step": 8113
    },
    {
      "epoch": 1.50677808727948,
      "grad_norm": 1.5379310957559715,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8715587258338928,
      "num_tokens": 282261033.0,
      "step": 8114
    },
    {
      "epoch": 1.5069637883008355,
      "grad_norm": 1.5250756620519696,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8826248645782471,
      "num_tokens": 282292630.0,
      "step": 8115
    },
    {
      "epoch": 1.5071494893221913,
      "grad_norm": 1.4386432395173214,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8815507888793945,
      "num_tokens": 282329495.0,
      "step": 8116
    },
    {
      "epoch": 1.507335190343547,
      "grad_norm": 1.3937895449474307,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.873489499092102,
      "num_tokens": 282371667.0,
      "step": 8117
    },
    {
      "epoch": 1.5075208913649025,
      "grad_norm": 1.5824212781125455,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8742466568946838,
      "num_tokens": 282403396.0,
      "step": 8118
    },
    {
      "epoch": 1.507706592386258,
      "grad_norm": 1.539636643278677,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8713761568069458,
      "num_tokens": 282441009.0,
      "step": 8119
    },
    {
      "epoch": 1.5078922934076138,
      "grad_norm": 1.632461080068622,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8826216459274292,
      "num_tokens": 282470553.0,
      "step": 8120
    },
    {
      "epoch": 1.5080779944289695,
      "grad_norm": 1.553876979518218,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8723971843719482,
      "num_tokens": 282504387.0,
      "step": 8121
    },
    {
      "epoch": 1.508263695450325,
      "grad_norm": 1.2775500395414388,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8960847854614258,
      "num_tokens": 282541809.0,
      "step": 8122
    },
    {
      "epoch": 1.5084493964716805,
      "grad_norm": 1.538341168969403,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8860800266265869,
      "num_tokens": 282571653.0,
      "step": 8123
    },
    {
      "epoch": 1.5086350974930363,
      "grad_norm": 1.462843127526363,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8867043256759644,
      "num_tokens": 282605538.0,
      "step": 8124
    },
    {
      "epoch": 1.508820798514392,
      "grad_norm": 1.4844804456036276,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8718675971031189,
      "num_tokens": 282645930.0,
      "step": 8125
    },
    {
      "epoch": 1.5090064995357473,
      "grad_norm": 1.4219192863615235,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.882442831993103,
      "num_tokens": 282684044.0,
      "step": 8126
    },
    {
      "epoch": 1.509192200557103,
      "grad_norm": 1.5289051415679158,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8697022199630737,
      "num_tokens": 282719831.0,
      "step": 8127
    },
    {
      "epoch": 1.5093779015784587,
      "grad_norm": 1.4858416312908165,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.880122184753418,
      "num_tokens": 282752754.0,
      "step": 8128
    },
    {
      "epoch": 1.5095636025998143,
      "grad_norm": 1.436745345037761,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8916839957237244,
      "num_tokens": 282784908.0,
      "step": 8129
    },
    {
      "epoch": 1.5097493036211698,
      "grad_norm": 1.5216558367133095,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8847417831420898,
      "num_tokens": 282820871.0,
      "step": 8130
    },
    {
      "epoch": 1.5099350046425255,
      "grad_norm": 1.4280497540487131,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8837399482727051,
      "num_tokens": 282856754.0,
      "step": 8131
    },
    {
      "epoch": 1.5101207056638812,
      "grad_norm": 1.5590890316554313,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8671915531158447,
      "num_tokens": 282890235.0,
      "step": 8132
    },
    {
      "epoch": 1.5103064066852367,
      "grad_norm": 1.407021566164942,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8732634782791138,
      "num_tokens": 282934347.0,
      "step": 8133
    },
    {
      "epoch": 1.5104921077065923,
      "grad_norm": 1.4919948664044753,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.866147518157959,
      "num_tokens": 282975925.0,
      "step": 8134
    },
    {
      "epoch": 1.510677808727948,
      "grad_norm": 1.479624840732388,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8834353685379028,
      "num_tokens": 283008464.0,
      "step": 8135
    },
    {
      "epoch": 1.5108635097493037,
      "grad_norm": 1.3883982205021315,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8892898559570312,
      "num_tokens": 283043365.0,
      "step": 8136
    },
    {
      "epoch": 1.5110492107706592,
      "grad_norm": 1.5697413988718836,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8789921998977661,
      "num_tokens": 283075920.0,
      "step": 8137
    },
    {
      "epoch": 1.5112349117920147,
      "grad_norm": 1.535804098485775,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8620849847793579,
      "num_tokens": 283113276.0,
      "step": 8138
    },
    {
      "epoch": 1.5114206128133705,
      "grad_norm": 1.4358437362384575,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8758668899536133,
      "num_tokens": 283150509.0,
      "step": 8139
    },
    {
      "epoch": 1.5116063138347262,
      "grad_norm": 1.568203583513147,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8781319856643677,
      "num_tokens": 283181430.0,
      "step": 8140
    },
    {
      "epoch": 1.5117920148560817,
      "grad_norm": 1.5319862197020309,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8637892603874207,
      "num_tokens": 283215777.0,
      "step": 8141
    },
    {
      "epoch": 1.5119777158774372,
      "grad_norm": 1.5988699798496842,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.874448299407959,
      "num_tokens": 283245144.0,
      "step": 8142
    },
    {
      "epoch": 1.512163416898793,
      "grad_norm": 1.5232211853897895,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8641775846481323,
      "num_tokens": 283281292.0,
      "step": 8143
    },
    {
      "epoch": 1.5123491179201487,
      "grad_norm": 1.5260914877764509,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8684365749359131,
      "num_tokens": 283317261.0,
      "step": 8144
    },
    {
      "epoch": 1.5125348189415042,
      "grad_norm": 1.381345762256829,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8832473754882812,
      "num_tokens": 283358199.0,
      "step": 8145
    },
    {
      "epoch": 1.5127205199628597,
      "grad_norm": 1.6096600241401957,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8819331526756287,
      "num_tokens": 283387926.0,
      "step": 8146
    },
    {
      "epoch": 1.5129062209842155,
      "grad_norm": 1.5260153198210507,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.870517373085022,
      "num_tokens": 283424258.0,
      "step": 8147
    },
    {
      "epoch": 1.5130919220055712,
      "grad_norm": 1.5305527997877795,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8823581337928772,
      "num_tokens": 283456249.0,
      "step": 8148
    },
    {
      "epoch": 1.5132776230269267,
      "grad_norm": 1.5702136303344612,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8614514470100403,
      "num_tokens": 283493086.0,
      "step": 8149
    },
    {
      "epoch": 1.5134633240482822,
      "grad_norm": 1.6027842128663,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8761423230171204,
      "num_tokens": 283525558.0,
      "step": 8150
    },
    {
      "epoch": 1.513649025069638,
      "grad_norm": 1.4485290116076308,
      "learning_rate": 1e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.865034818649292,
      "num_tokens": 283567757.0,
      "step": 8151
    },
    {
      "epoch": 1.5138347260909935,
      "grad_norm": 1.4341166992219025,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8883283734321594,
      "num_tokens": 283603373.0,
      "step": 8152
    },
    {
      "epoch": 1.514020427112349,
      "grad_norm": 1.7157871217020713,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8718892335891724,
      "num_tokens": 283629559.0,
      "step": 8153
    },
    {
      "epoch": 1.5142061281337047,
      "grad_norm": 1.5831244542723906,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8684008717536926,
      "num_tokens": 283659701.0,
      "step": 8154
    },
    {
      "epoch": 1.5143918291550604,
      "grad_norm": 1.5571086034798725,
      "learning_rate": 1e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8607670664787292,
      "num_tokens": 283695754.0,
      "step": 8155
    },
    {
      "epoch": 1.514577530176416,
      "grad_norm": 1.4843098755853972,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8836249709129333,
      "num_tokens": 283728569.0,
      "step": 8156
    },
    {
      "epoch": 1.5147632311977715,
      "grad_norm": 1.5672283781701921,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8845674991607666,
      "num_tokens": 283762788.0,
      "step": 8157
    },
    {
      "epoch": 1.5149489322191272,
      "grad_norm": 1.5621830700133048,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.87343430519104,
      "num_tokens": 283797316.0,
      "step": 8158
    },
    {
      "epoch": 1.515134633240483,
      "grad_norm": 1.5468080860360809,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8745992183685303,
      "num_tokens": 283831055.0,
      "step": 8159
    },
    {
      "epoch": 1.5153203342618384,
      "grad_norm": 1.3611207598590618,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8894660472869873,
      "num_tokens": 283871018.0,
      "step": 8160
    },
    {
      "epoch": 1.515506035283194,
      "grad_norm": 1.459522289057567,
      "learning_rate": 1e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.863638699054718,
      "num_tokens": 283913755.0,
      "step": 8161
    },
    {
      "epoch": 1.5156917363045497,
      "grad_norm": 1.755992930948632,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8623352646827698,
      "num_tokens": 283942007.0,
      "step": 8162
    },
    {
      "epoch": 1.5158774373259054,
      "grad_norm": 1.401732767119633,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8864552974700928,
      "num_tokens": 283977647.0,
      "step": 8163
    },
    {
      "epoch": 1.516063138347261,
      "grad_norm": 1.4465182486404091,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.874576985836029,
      "num_tokens": 284013442.0,
      "step": 8164
    },
    {
      "epoch": 1.5162488393686164,
      "grad_norm": 1.647640251554855,
      "learning_rate": 1e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8622543811798096,
      "num_tokens": 284045397.0,
      "step": 8165
    },
    {
      "epoch": 1.5164345403899722,
      "grad_norm": 1.6388309743182996,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8755427598953247,
      "num_tokens": 284077231.0,
      "step": 8166
    },
    {
      "epoch": 1.516620241411328,
      "grad_norm": 1.4421363824368514,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8700658082962036,
      "num_tokens": 284113779.0,
      "step": 8167
    },
    {
      "epoch": 1.5168059424326834,
      "grad_norm": 1.5136265347469222,
      "learning_rate": 1e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.8576492667198181,
      "num_tokens": 284152970.0,
      "step": 8168
    },
    {
      "epoch": 1.516991643454039,
      "grad_norm": 1.549419837158392,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8690835237503052,
      "num_tokens": 284189203.0,
      "step": 8169
    },
    {
      "epoch": 1.5171773444753947,
      "grad_norm": 1.437569910683671,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8683459758758545,
      "num_tokens": 284228923.0,
      "step": 8170
    },
    {
      "epoch": 1.5173630454967504,
      "grad_norm": 1.5544546556161332,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8657435178756714,
      "num_tokens": 284264799.0,
      "step": 8171
    },
    {
      "epoch": 1.517548746518106,
      "grad_norm": 1.5348755137101973,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8687649965286255,
      "num_tokens": 284296672.0,
      "step": 8172
    },
    {
      "epoch": 1.5177344475394614,
      "grad_norm": 1.4568479623527328,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8692739009857178,
      "num_tokens": 284334521.0,
      "step": 8173
    },
    {
      "epoch": 1.5179201485608171,
      "grad_norm": 1.475559308387585,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8859667778015137,
      "num_tokens": 284367398.0,
      "step": 8174
    },
    {
      "epoch": 1.5181058495821727,
      "grad_norm": 1.6578860374698847,
      "learning_rate": 1e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.856131374835968,
      "num_tokens": 284398349.0,
      "step": 8175
    },
    {
      "epoch": 1.5182915506035282,
      "grad_norm": 1.5704148823319903,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8726927042007446,
      "num_tokens": 284430403.0,
      "step": 8176
    },
    {
      "epoch": 1.518477251624884,
      "grad_norm": 1.4715023788351387,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8730276226997375,
      "num_tokens": 284469209.0,
      "step": 8177
    },
    {
      "epoch": 1.5186629526462396,
      "grad_norm": 1.5079818377857446,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8787689208984375,
      "num_tokens": 284502634.0,
      "step": 8178
    },
    {
      "epoch": 1.5188486536675951,
      "grad_norm": 1.5119684337160775,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8759198784828186,
      "num_tokens": 284535158.0,
      "step": 8179
    },
    {
      "epoch": 1.5190343546889506,
      "grad_norm": 1.5121509438417773,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8822435140609741,
      "num_tokens": 284567337.0,
      "step": 8180
    },
    {
      "epoch": 1.5192200557103064,
      "grad_norm": 1.5764068809552658,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8695701956748962,
      "num_tokens": 284601021.0,
      "step": 8181
    },
    {
      "epoch": 1.5194057567316621,
      "grad_norm": 1.582583189972439,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8699224591255188,
      "num_tokens": 284636830.0,
      "step": 8182
    },
    {
      "epoch": 1.5195914577530176,
      "grad_norm": 1.6597944554838584,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8797498941421509,
      "num_tokens": 284665957.0,
      "step": 8183
    },
    {
      "epoch": 1.5197771587743731,
      "grad_norm": 1.511602302941962,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.88416588306427,
      "num_tokens": 284702136.0,
      "step": 8184
    },
    {
      "epoch": 1.5199628597957289,
      "grad_norm": 1.4393478628130634,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8700093030929565,
      "num_tokens": 284740224.0,
      "step": 8185
    },
    {
      "epoch": 1.5201485608170846,
      "grad_norm": 1.641782833016333,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8705577850341797,
      "num_tokens": 284767579.0,
      "step": 8186
    },
    {
      "epoch": 1.5203342618384401,
      "grad_norm": 1.3487588712751428,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8885877132415771,
      "num_tokens": 284807721.0,
      "step": 8187
    },
    {
      "epoch": 1.5205199628597956,
      "grad_norm": 1.5212372931116724,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8802040815353394,
      "num_tokens": 284845139.0,
      "step": 8188
    },
    {
      "epoch": 1.5207056638811514,
      "grad_norm": 1.4990520022627873,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8741875290870667,
      "num_tokens": 284879652.0,
      "step": 8189
    },
    {
      "epoch": 1.520891364902507,
      "grad_norm": 1.4854157176190352,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8641021847724915,
      "num_tokens": 284916261.0,
      "step": 8190
    },
    {
      "epoch": 1.5210770659238626,
      "grad_norm": 1.5411778172877761,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8728598952293396,
      "num_tokens": 284951118.0,
      "step": 8191
    },
    {
      "epoch": 1.5212627669452181,
      "grad_norm": 1.4515176156241503,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.884609043598175,
      "num_tokens": 284987949.0,
      "step": 8192
    },
    {
      "epoch": 1.5214484679665738,
      "grad_norm": 1.6806468337548752,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8736460208892822,
      "num_tokens": 285017733.0,
      "step": 8193
    },
    {
      "epoch": 1.5216341689879296,
      "grad_norm": 1.4822589381880866,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8717309236526489,
      "num_tokens": 285054959.0,
      "step": 8194
    },
    {
      "epoch": 1.521819870009285,
      "grad_norm": 1.5703179620503835,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.873727560043335,
      "num_tokens": 285087752.0,
      "step": 8195
    },
    {
      "epoch": 1.5220055710306406,
      "grad_norm": 1.4216870591491186,
      "learning_rate": 1e-06,
      "loss": 0.3683,
      "mean_token_accuracy": 0.8707908391952515,
      "num_tokens": 285126744.0,
      "step": 8196
    },
    {
      "epoch": 1.5221912720519963,
      "grad_norm": 1.5340881945708968,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8743331432342529,
      "num_tokens": 285164910.0,
      "step": 8197
    },
    {
      "epoch": 1.522376973073352,
      "grad_norm": 1.5873227946009223,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8807737827301025,
      "num_tokens": 285194589.0,
      "step": 8198
    },
    {
      "epoch": 1.5225626740947074,
      "grad_norm": 1.4720877189445696,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8762543201446533,
      "num_tokens": 285229424.0,
      "step": 8199
    },
    {
      "epoch": 1.522748375116063,
      "grad_norm": 1.4668877338065127,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8828206658363342,
      "num_tokens": 285265726.0,
      "step": 8200
    },
    {
      "epoch": 1.5229340761374188,
      "grad_norm": 1.4728904147723187,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8715593814849854,
      "num_tokens": 285299892.0,
      "step": 8201
    },
    {
      "epoch": 1.5231197771587743,
      "grad_norm": 1.560753108207471,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8745049834251404,
      "num_tokens": 285333819.0,
      "step": 8202
    },
    {
      "epoch": 1.5233054781801298,
      "grad_norm": 1.5387430328494092,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8714789152145386,
      "num_tokens": 285369531.0,
      "step": 8203
    },
    {
      "epoch": 1.5234911792014856,
      "grad_norm": 1.4951195412278002,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8803740739822388,
      "num_tokens": 285403328.0,
      "step": 8204
    },
    {
      "epoch": 1.5236768802228413,
      "grad_norm": 1.3641082513757963,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8914207220077515,
      "num_tokens": 285442850.0,
      "step": 8205
    },
    {
      "epoch": 1.5238625812441968,
      "grad_norm": 1.4799132502654175,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8813676834106445,
      "num_tokens": 285475420.0,
      "step": 8206
    },
    {
      "epoch": 1.5240482822655523,
      "grad_norm": 1.5804239012018104,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8762885928153992,
      "num_tokens": 285509147.0,
      "step": 8207
    },
    {
      "epoch": 1.524233983286908,
      "grad_norm": 1.4635882866389838,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.878849446773529,
      "num_tokens": 285547162.0,
      "step": 8208
    },
    {
      "epoch": 1.5244196843082638,
      "grad_norm": 1.614261091777485,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8627835512161255,
      "num_tokens": 285580143.0,
      "step": 8209
    },
    {
      "epoch": 1.5246053853296193,
      "grad_norm": 1.6006205565679175,
      "learning_rate": 1e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8624075055122375,
      "num_tokens": 285612175.0,
      "step": 8210
    },
    {
      "epoch": 1.5247910863509748,
      "grad_norm": 1.4600794693589378,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8746553063392639,
      "num_tokens": 285649343.0,
      "step": 8211
    },
    {
      "epoch": 1.5249767873723306,
      "grad_norm": 1.4205059873805561,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8738377094268799,
      "num_tokens": 285688987.0,
      "step": 8212
    },
    {
      "epoch": 1.5251624883936863,
      "grad_norm": 1.569599002648228,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.871230959892273,
      "num_tokens": 285720849.0,
      "step": 8213
    },
    {
      "epoch": 1.5253481894150418,
      "grad_norm": 1.467581190377012,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.871433675289154,
      "num_tokens": 285757179.0,
      "step": 8214
    },
    {
      "epoch": 1.5255338904363973,
      "grad_norm": 1.5682928959155045,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.86857008934021,
      "num_tokens": 285792485.0,
      "step": 8215
    },
    {
      "epoch": 1.525719591457753,
      "grad_norm": 1.5187719956682344,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8722473382949829,
      "num_tokens": 285823411.0,
      "step": 8216
    },
    {
      "epoch": 1.5259052924791088,
      "grad_norm": 1.5072073732481186,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8814871311187744,
      "num_tokens": 285857691.0,
      "step": 8217
    },
    {
      "epoch": 1.5260909935004643,
      "grad_norm": 1.6170389463295702,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8647195100784302,
      "num_tokens": 285889964.0,
      "step": 8218
    },
    {
      "epoch": 1.5262766945218198,
      "grad_norm": 1.5028670448629218,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8784671425819397,
      "num_tokens": 285922957.0,
      "step": 8219
    },
    {
      "epoch": 1.5264623955431755,
      "grad_norm": 1.4990300634146185,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8753782510757446,
      "num_tokens": 285961869.0,
      "step": 8220
    },
    {
      "epoch": 1.5266480965645313,
      "grad_norm": 1.5129132558359153,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8760524988174438,
      "num_tokens": 285997016.0,
      "step": 8221
    },
    {
      "epoch": 1.5268337975858868,
      "grad_norm": 1.4461065657917807,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8925467729568481,
      "num_tokens": 286035022.0,
      "step": 8222
    },
    {
      "epoch": 1.5270194986072423,
      "grad_norm": 1.4636730446749102,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.889080286026001,
      "num_tokens": 286065633.0,
      "step": 8223
    },
    {
      "epoch": 1.527205199628598,
      "grad_norm": 1.6742586427185786,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8804275989532471,
      "num_tokens": 286095656.0,
      "step": 8224
    },
    {
      "epoch": 1.5273909006499535,
      "grad_norm": 1.790810137201269,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8693995475769043,
      "num_tokens": 286127739.0,
      "step": 8225
    },
    {
      "epoch": 1.527576601671309,
      "grad_norm": 1.6256352699715855,
      "learning_rate": 1e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8622853755950928,
      "num_tokens": 286160254.0,
      "step": 8226
    },
    {
      "epoch": 1.5277623026926648,
      "grad_norm": 1.5794538789010044,
      "learning_rate": 1e-06,
      "loss": 0.3929,
      "mean_token_accuracy": 0.8681817054748535,
      "num_tokens": 286191820.0,
      "step": 8227
    },
    {
      "epoch": 1.5279480037140205,
      "grad_norm": 1.5484420932352445,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.876961350440979,
      "num_tokens": 286226102.0,
      "step": 8228
    },
    {
      "epoch": 1.528133704735376,
      "grad_norm": 1.4713728614865753,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8804311752319336,
      "num_tokens": 286260033.0,
      "step": 8229
    },
    {
      "epoch": 1.5283194057567315,
      "grad_norm": 1.4699150215607135,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8797745108604431,
      "num_tokens": 286294575.0,
      "step": 8230
    },
    {
      "epoch": 1.5285051067780873,
      "grad_norm": 1.5049967870896532,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8869304060935974,
      "num_tokens": 286328959.0,
      "step": 8231
    },
    {
      "epoch": 1.528690807799443,
      "grad_norm": 1.5632786781898502,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8874702453613281,
      "num_tokens": 286362633.0,
      "step": 8232
    },
    {
      "epoch": 1.5288765088207985,
      "grad_norm": 1.5952888953799993,
      "learning_rate": 1e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8645355701446533,
      "num_tokens": 286392583.0,
      "step": 8233
    },
    {
      "epoch": 1.529062209842154,
      "grad_norm": 1.4992154990221565,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8668636083602905,
      "num_tokens": 286429177.0,
      "step": 8234
    },
    {
      "epoch": 1.5292479108635098,
      "grad_norm": 1.6930073802143888,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.874229907989502,
      "num_tokens": 286457102.0,
      "step": 8235
    },
    {
      "epoch": 1.5294336118848655,
      "grad_norm": 1.606373640902951,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8659518957138062,
      "num_tokens": 286489081.0,
      "step": 8236
    },
    {
      "epoch": 1.529619312906221,
      "grad_norm": 1.5104825459521953,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8714742064476013,
      "num_tokens": 286522967.0,
      "step": 8237
    },
    {
      "epoch": 1.5298050139275765,
      "grad_norm": 1.670316657827281,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8675755262374878,
      "num_tokens": 286551776.0,
      "step": 8238
    },
    {
      "epoch": 1.5299907149489322,
      "grad_norm": 1.4652430132338872,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8828259706497192,
      "num_tokens": 286589737.0,
      "step": 8239
    },
    {
      "epoch": 1.530176415970288,
      "grad_norm": 1.5616393773256685,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8681223392486572,
      "num_tokens": 286624378.0,
      "step": 8240
    },
    {
      "epoch": 1.5303621169916435,
      "grad_norm": 1.6865625343563964,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.87314373254776,
      "num_tokens": 286654966.0,
      "step": 8241
    },
    {
      "epoch": 1.530547818012999,
      "grad_norm": 1.3983051846816024,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8909580707550049,
      "num_tokens": 286691010.0,
      "step": 8242
    },
    {
      "epoch": 1.5307335190343547,
      "grad_norm": 1.6072109690433982,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8799570202827454,
      "num_tokens": 286720616.0,
      "step": 8243
    },
    {
      "epoch": 1.5309192200557105,
      "grad_norm": 1.529202310317556,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8808923363685608,
      "num_tokens": 286752523.0,
      "step": 8244
    },
    {
      "epoch": 1.531104921077066,
      "grad_norm": 1.640349519392285,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8678324222564697,
      "num_tokens": 286784746.0,
      "step": 8245
    },
    {
      "epoch": 1.5312906220984215,
      "grad_norm": 1.5515908882482485,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8702048659324646,
      "num_tokens": 286818370.0,
      "step": 8246
    },
    {
      "epoch": 1.5314763231197772,
      "grad_norm": 1.6740322727593357,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8681259155273438,
      "num_tokens": 286847957.0,
      "step": 8247
    },
    {
      "epoch": 1.5316620241411327,
      "grad_norm": 1.4710703756753667,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8763006925582886,
      "num_tokens": 286887103.0,
      "step": 8248
    },
    {
      "epoch": 1.5318477251624882,
      "grad_norm": 1.5527075223928495,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8796910643577576,
      "num_tokens": 286919582.0,
      "step": 8249
    },
    {
      "epoch": 1.532033426183844,
      "grad_norm": 1.787371060820898,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8620918393135071,
      "num_tokens": 286946689.0,
      "step": 8250
    },
    {
      "epoch": 1.5322191272051997,
      "grad_norm": 1.552089920980342,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8794366121292114,
      "num_tokens": 286975589.0,
      "step": 8251
    },
    {
      "epoch": 1.5324048282265552,
      "grad_norm": 1.5588208138275397,
      "learning_rate": 1e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8664573431015015,
      "num_tokens": 287009527.0,
      "step": 8252
    },
    {
      "epoch": 1.5325905292479107,
      "grad_norm": 1.6579053786193816,
      "learning_rate": 1e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8499735593795776,
      "num_tokens": 287041880.0,
      "step": 8253
    },
    {
      "epoch": 1.5327762302692665,
      "grad_norm": 1.4694903953982013,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8854110240936279,
      "num_tokens": 287076198.0,
      "step": 8254
    },
    {
      "epoch": 1.5329619312906222,
      "grad_norm": 1.588344020376577,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8745702505111694,
      "num_tokens": 287109007.0,
      "step": 8255
    },
    {
      "epoch": 1.5331476323119777,
      "grad_norm": 1.5935957120991886,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8684847950935364,
      "num_tokens": 287147356.0,
      "step": 8256
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 1.5203548086242011,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8666266202926636,
      "num_tokens": 287185722.0,
      "step": 8257
    },
    {
      "epoch": 1.533519034354689,
      "grad_norm": 1.5650697842130694,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.890040934085846,
      "num_tokens": 287215830.0,
      "step": 8258
    },
    {
      "epoch": 1.5337047353760447,
      "grad_norm": 1.507872972796406,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8760678172111511,
      "num_tokens": 287248085.0,
      "step": 8259
    },
    {
      "epoch": 1.5338904363974002,
      "grad_norm": 1.3859465831613729,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.8952100276947021,
      "num_tokens": 287282591.0,
      "step": 8260
    },
    {
      "epoch": 1.5340761374187557,
      "grad_norm": 1.5397428861663978,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8789635896682739,
      "num_tokens": 287312915.0,
      "step": 8261
    },
    {
      "epoch": 1.5342618384401114,
      "grad_norm": 1.394917985315991,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8770735859870911,
      "num_tokens": 287354437.0,
      "step": 8262
    },
    {
      "epoch": 1.5344475394614672,
      "grad_norm": 1.3785316470745297,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8848557472229004,
      "num_tokens": 287390601.0,
      "step": 8263
    },
    {
      "epoch": 1.5346332404828227,
      "grad_norm": 1.4264836914901808,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8732573986053467,
      "num_tokens": 287427112.0,
      "step": 8264
    },
    {
      "epoch": 1.5348189415041782,
      "grad_norm": 1.5397677911600363,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8679249286651611,
      "num_tokens": 287459207.0,
      "step": 8265
    },
    {
      "epoch": 1.535004642525534,
      "grad_norm": 1.59136262860227,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8856716156005859,
      "num_tokens": 287490184.0,
      "step": 8266
    },
    {
      "epoch": 1.5351903435468897,
      "grad_norm": 1.569270247553882,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.882365345954895,
      "num_tokens": 287521970.0,
      "step": 8267
    },
    {
      "epoch": 1.5353760445682452,
      "grad_norm": 1.425000823488059,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8825210332870483,
      "num_tokens": 287559427.0,
      "step": 8268
    },
    {
      "epoch": 1.5355617455896007,
      "grad_norm": 1.4500328054677822,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8760162591934204,
      "num_tokens": 287594230.0,
      "step": 8269
    },
    {
      "epoch": 1.5357474466109564,
      "grad_norm": 1.457113524162689,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8774677515029907,
      "num_tokens": 287630712.0,
      "step": 8270
    },
    {
      "epoch": 1.535933147632312,
      "grad_norm": 1.496968100262457,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.882818341255188,
      "num_tokens": 287666145.0,
      "step": 8271
    },
    {
      "epoch": 1.5361188486536674,
      "grad_norm": 1.4920436727177788,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8884179592132568,
      "num_tokens": 287700637.0,
      "step": 8272
    },
    {
      "epoch": 1.5363045496750232,
      "grad_norm": 1.555055719126336,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8739565014839172,
      "num_tokens": 287733962.0,
      "step": 8273
    },
    {
      "epoch": 1.536490250696379,
      "grad_norm": 1.6135179091826746,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8647414445877075,
      "num_tokens": 287768740.0,
      "step": 8274
    },
    {
      "epoch": 1.5366759517177344,
      "grad_norm": 1.6278745788771736,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8682689666748047,
      "num_tokens": 287801364.0,
      "step": 8275
    },
    {
      "epoch": 1.53686165273909,
      "grad_norm": 1.4803455983735896,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8768429756164551,
      "num_tokens": 287834217.0,
      "step": 8276
    },
    {
      "epoch": 1.5370473537604457,
      "grad_norm": 1.5757410068294657,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8720338344573975,
      "num_tokens": 287867058.0,
      "step": 8277
    },
    {
      "epoch": 1.5372330547818014,
      "grad_norm": 1.5352433943676032,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8775311708450317,
      "num_tokens": 287899705.0,
      "step": 8278
    },
    {
      "epoch": 1.537418755803157,
      "grad_norm": 1.5400662496585018,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8703300952911377,
      "num_tokens": 287938957.0,
      "step": 8279
    },
    {
      "epoch": 1.5376044568245124,
      "grad_norm": 1.654095996822214,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8792740106582642,
      "num_tokens": 287973873.0,
      "step": 8280
    },
    {
      "epoch": 1.5377901578458681,
      "grad_norm": 1.4591352035057372,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8650121092796326,
      "num_tokens": 288012785.0,
      "step": 8281
    },
    {
      "epoch": 1.5379758588672239,
      "grad_norm": 1.4099879630192196,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8892732858657837,
      "num_tokens": 288052734.0,
      "step": 8282
    },
    {
      "epoch": 1.5381615598885794,
      "grad_norm": 1.6299088511038442,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8674664497375488,
      "num_tokens": 288086966.0,
      "step": 8283
    },
    {
      "epoch": 1.538347260909935,
      "grad_norm": 1.5517086486053233,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.8776988983154297,
      "num_tokens": 288119945.0,
      "step": 8284
    },
    {
      "epoch": 1.5385329619312906,
      "grad_norm": 1.332888747769121,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8929826021194458,
      "num_tokens": 288157749.0,
      "step": 8285
    },
    {
      "epoch": 1.5387186629526464,
      "grad_norm": 1.4488611641015159,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8719260096549988,
      "num_tokens": 288195830.0,
      "step": 8286
    },
    {
      "epoch": 1.5389043639740019,
      "grad_norm": 1.611021837635153,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8735476732254028,
      "num_tokens": 288225567.0,
      "step": 8287
    },
    {
      "epoch": 1.5390900649953574,
      "grad_norm": 1.39556547427993,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8841179609298706,
      "num_tokens": 288266440.0,
      "step": 8288
    },
    {
      "epoch": 1.5392757660167131,
      "grad_norm": 1.5310613855173958,
      "learning_rate": 1e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8593412637710571,
      "num_tokens": 288304514.0,
      "step": 8289
    },
    {
      "epoch": 1.5394614670380689,
      "grad_norm": 1.473889850152721,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8688751459121704,
      "num_tokens": 288341711.0,
      "step": 8290
    },
    {
      "epoch": 1.5396471680594244,
      "grad_norm": 1.4870305760995626,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8843334317207336,
      "num_tokens": 288375776.0,
      "step": 8291
    },
    {
      "epoch": 1.5398328690807799,
      "grad_norm": 1.4422772804005644,
      "learning_rate": 1e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8613114356994629,
      "num_tokens": 288416962.0,
      "step": 8292
    },
    {
      "epoch": 1.5400185701021356,
      "grad_norm": 1.5350536532518753,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8777913451194763,
      "num_tokens": 288449511.0,
      "step": 8293
    },
    {
      "epoch": 1.5402042711234913,
      "grad_norm": 1.4291795273131724,
      "learning_rate": 1e-06,
      "loss": 0.2927,
      "mean_token_accuracy": 0.9033052921295166,
      "num_tokens": 288485058.0,
      "step": 8294
    },
    {
      "epoch": 1.5403899721448466,
      "grad_norm": 1.5950944216135285,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8747566938400269,
      "num_tokens": 288517595.0,
      "step": 8295
    },
    {
      "epoch": 1.5405756731662024,
      "grad_norm": 1.4931211757278615,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8762427568435669,
      "num_tokens": 288556433.0,
      "step": 8296
    },
    {
      "epoch": 1.540761374187558,
      "grad_norm": 1.527340721448085,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8700007200241089,
      "num_tokens": 288591539.0,
      "step": 8297
    },
    {
      "epoch": 1.5409470752089136,
      "grad_norm": 1.5607474727370712,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8727024793624878,
      "num_tokens": 288628180.0,
      "step": 8298
    },
    {
      "epoch": 1.5411327762302691,
      "grad_norm": 1.3786640864724309,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8844415545463562,
      "num_tokens": 288667152.0,
      "step": 8299
    },
    {
      "epoch": 1.5413184772516249,
      "grad_norm": 1.4093820267063994,
      "learning_rate": 1e-06,
      "loss": 0.4048,
      "mean_token_accuracy": 0.8617488741874695,
      "num_tokens": 288705924.0,
      "step": 8300
    },
    {
      "epoch": 1.5415041782729806,
      "grad_norm": 1.6197354460693514,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8708410263061523,
      "num_tokens": 288741551.0,
      "step": 8301
    },
    {
      "epoch": 1.541689879294336,
      "grad_norm": 1.5987731340038158,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8761723041534424,
      "num_tokens": 288771610.0,
      "step": 8302
    },
    {
      "epoch": 1.5418755803156916,
      "grad_norm": 1.471098991421303,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8831890225410461,
      "num_tokens": 288807940.0,
      "step": 8303
    },
    {
      "epoch": 1.5420612813370473,
      "grad_norm": 1.607186048639758,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8828384280204773,
      "num_tokens": 288837947.0,
      "step": 8304
    },
    {
      "epoch": 1.542246982358403,
      "grad_norm": 1.6598637720325715,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8765091300010681,
      "num_tokens": 288868648.0,
      "step": 8305
    },
    {
      "epoch": 1.5424326833797586,
      "grad_norm": 1.3833939452641706,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8820359706878662,
      "num_tokens": 288903850.0,
      "step": 8306
    },
    {
      "epoch": 1.542618384401114,
      "grad_norm": 1.3607722930664263,
      "learning_rate": 1e-06,
      "loss": 0.2931,
      "mean_token_accuracy": 0.8965673446655273,
      "num_tokens": 288939351.0,
      "step": 8307
    },
    {
      "epoch": 1.5428040854224698,
      "grad_norm": 1.5694760932469094,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.869044840335846,
      "num_tokens": 288977196.0,
      "step": 8308
    },
    {
      "epoch": 1.5429897864438256,
      "grad_norm": 1.5108419941608264,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8841872215270996,
      "num_tokens": 289013464.0,
      "step": 8309
    },
    {
      "epoch": 1.543175487465181,
      "grad_norm": 1.4499978357087826,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8909169435501099,
      "num_tokens": 289047294.0,
      "step": 8310
    },
    {
      "epoch": 1.5433611884865366,
      "grad_norm": 1.5845847102255706,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8704845905303955,
      "num_tokens": 289077716.0,
      "step": 8311
    },
    {
      "epoch": 1.5435468895078923,
      "grad_norm": 1.4917619071283474,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8778584003448486,
      "num_tokens": 289109624.0,
      "step": 8312
    },
    {
      "epoch": 1.543732590529248,
      "grad_norm": 1.6066584670118371,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8747792840003967,
      "num_tokens": 289143849.0,
      "step": 8313
    },
    {
      "epoch": 1.5439182915506036,
      "grad_norm": 1.5537223383598182,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8628456592559814,
      "num_tokens": 289180650.0,
      "step": 8314
    },
    {
      "epoch": 1.544103992571959,
      "grad_norm": 1.6654274432186724,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8652442097663879,
      "num_tokens": 289212970.0,
      "step": 8315
    },
    {
      "epoch": 1.5442896935933148,
      "grad_norm": 1.6075010924969488,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8711186051368713,
      "num_tokens": 289250101.0,
      "step": 8316
    },
    {
      "epoch": 1.5444753946146705,
      "grad_norm": 1.431441116664193,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8914002180099487,
      "num_tokens": 289286780.0,
      "step": 8317
    },
    {
      "epoch": 1.544661095636026,
      "grad_norm": 1.50944735921159,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8628296852111816,
      "num_tokens": 289321097.0,
      "step": 8318
    },
    {
      "epoch": 1.5448467966573816,
      "grad_norm": 1.5300077877326406,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8710969686508179,
      "num_tokens": 289353308.0,
      "step": 8319
    },
    {
      "epoch": 1.5450324976787373,
      "grad_norm": 1.4647720853395452,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8846213221549988,
      "num_tokens": 289386271.0,
      "step": 8320
    },
    {
      "epoch": 1.5452181987000928,
      "grad_norm": 1.4893405391407868,
      "learning_rate": 1e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8597769737243652,
      "num_tokens": 289424771.0,
      "step": 8321
    },
    {
      "epoch": 1.5454038997214483,
      "grad_norm": 1.5228094754622992,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8822994828224182,
      "num_tokens": 289456426.0,
      "step": 8322
    },
    {
      "epoch": 1.545589600742804,
      "grad_norm": 1.596417877925912,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8785907626152039,
      "num_tokens": 289486350.0,
      "step": 8323
    },
    {
      "epoch": 1.5457753017641598,
      "grad_norm": 1.4203066723755247,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8942243456840515,
      "num_tokens": 289518966.0,
      "step": 8324
    },
    {
      "epoch": 1.5459610027855153,
      "grad_norm": 1.4382989005134499,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8775001168251038,
      "num_tokens": 289555214.0,
      "step": 8325
    },
    {
      "epoch": 1.5461467038068708,
      "grad_norm": 1.4495390642364558,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8900655508041382,
      "num_tokens": 289590110.0,
      "step": 8326
    },
    {
      "epoch": 1.5463324048282265,
      "grad_norm": 1.4887917294728312,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.8943227529525757,
      "num_tokens": 289622303.0,
      "step": 8327
    },
    {
      "epoch": 1.5465181058495823,
      "grad_norm": 1.561126393665457,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8656302690505981,
      "num_tokens": 289657240.0,
      "step": 8328
    },
    {
      "epoch": 1.5467038068709378,
      "grad_norm": 1.5168380253149651,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8924216032028198,
      "num_tokens": 289691861.0,
      "step": 8329
    },
    {
      "epoch": 1.5468895078922933,
      "grad_norm": 1.625387252512172,
      "learning_rate": 1e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8610302805900574,
      "num_tokens": 289727495.0,
      "step": 8330
    },
    {
      "epoch": 1.547075208913649,
      "grad_norm": 1.5932158026134073,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8768218159675598,
      "num_tokens": 289759081.0,
      "step": 8331
    },
    {
      "epoch": 1.5472609099350048,
      "grad_norm": 1.5400340843134006,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8814495205879211,
      "num_tokens": 289792173.0,
      "step": 8332
    },
    {
      "epoch": 1.5474466109563603,
      "grad_norm": 1.5333370874074956,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8829572796821594,
      "num_tokens": 289827639.0,
      "step": 8333
    },
    {
      "epoch": 1.5476323119777158,
      "grad_norm": 1.5212931267851035,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8686642646789551,
      "num_tokens": 289870302.0,
      "step": 8334
    },
    {
      "epoch": 1.5478180129990715,
      "grad_norm": 1.3418296589318637,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8709763288497925,
      "num_tokens": 289917629.0,
      "step": 8335
    },
    {
      "epoch": 1.5480037140204272,
      "grad_norm": 1.4027145447239728,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8837940096855164,
      "num_tokens": 289955171.0,
      "step": 8336
    },
    {
      "epoch": 1.5481894150417828,
      "grad_norm": 1.487524427104461,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.863332211971283,
      "num_tokens": 289994740.0,
      "step": 8337
    },
    {
      "epoch": 1.5483751160631383,
      "grad_norm": 1.4154878769802726,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8877376317977905,
      "num_tokens": 290032542.0,
      "step": 8338
    },
    {
      "epoch": 1.548560817084494,
      "grad_norm": 1.4212163128133342,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8937852382659912,
      "num_tokens": 290071027.0,
      "step": 8339
    },
    {
      "epoch": 1.5487465181058497,
      "grad_norm": 1.5616489023842164,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.880838930606842,
      "num_tokens": 290105001.0,
      "step": 8340
    },
    {
      "epoch": 1.5489322191272052,
      "grad_norm": 1.5347042755416653,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8804174065589905,
      "num_tokens": 290137502.0,
      "step": 8341
    },
    {
      "epoch": 1.5491179201485608,
      "grad_norm": 1.5053522165215485,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8778756856918335,
      "num_tokens": 290174604.0,
      "step": 8342
    },
    {
      "epoch": 1.5493036211699165,
      "grad_norm": 1.542680892958278,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.885809600353241,
      "num_tokens": 290212705.0,
      "step": 8343
    },
    {
      "epoch": 1.549489322191272,
      "grad_norm": 1.459697920276339,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8866505026817322,
      "num_tokens": 290245565.0,
      "step": 8344
    },
    {
      "epoch": 1.5496750232126275,
      "grad_norm": 1.429189506329267,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8731988668441772,
      "num_tokens": 290286169.0,
      "step": 8345
    },
    {
      "epoch": 1.5498607242339832,
      "grad_norm": 1.7716689081649302,
      "learning_rate": 1e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8567163944244385,
      "num_tokens": 290322918.0,
      "step": 8346
    },
    {
      "epoch": 1.550046425255339,
      "grad_norm": 1.4082856988074501,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8749203681945801,
      "num_tokens": 290363121.0,
      "step": 8347
    },
    {
      "epoch": 1.5502321262766945,
      "grad_norm": 1.6399327635477141,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8802431225776672,
      "num_tokens": 290394122.0,
      "step": 8348
    },
    {
      "epoch": 1.55041782729805,
      "grad_norm": 1.5409496168715853,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8706955313682556,
      "num_tokens": 290429167.0,
      "step": 8349
    },
    {
      "epoch": 1.5506035283194057,
      "grad_norm": 1.6393467696379724,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8708875775337219,
      "num_tokens": 290459494.0,
      "step": 8350
    },
    {
      "epoch": 1.5507892293407615,
      "grad_norm": 1.6058005512969298,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8793492317199707,
      "num_tokens": 290492840.0,
      "step": 8351
    },
    {
      "epoch": 1.550974930362117,
      "grad_norm": 1.4748168426468733,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8855699300765991,
      "num_tokens": 290528247.0,
      "step": 8352
    },
    {
      "epoch": 1.5511606313834725,
      "grad_norm": 1.694967037520548,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8785933256149292,
      "num_tokens": 290558360.0,
      "step": 8353
    },
    {
      "epoch": 1.5513463324048282,
      "grad_norm": 1.5238746693392544,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8600320219993591,
      "num_tokens": 290597110.0,
      "step": 8354
    },
    {
      "epoch": 1.551532033426184,
      "grad_norm": 1.5444884087507913,
      "learning_rate": 1e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8605398535728455,
      "num_tokens": 290633644.0,
      "step": 8355
    },
    {
      "epoch": 1.5517177344475395,
      "grad_norm": 1.5297663774815564,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.8745973110198975,
      "num_tokens": 290669799.0,
      "step": 8356
    },
    {
      "epoch": 1.551903435468895,
      "grad_norm": 1.6959194877238575,
      "learning_rate": 1e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8714693784713745,
      "num_tokens": 290702651.0,
      "step": 8357
    },
    {
      "epoch": 1.5520891364902507,
      "grad_norm": 1.5165934301499706,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8885733485221863,
      "num_tokens": 290732652.0,
      "step": 8358
    },
    {
      "epoch": 1.5522748375116064,
      "grad_norm": 1.5907355752824284,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.875,
      "num_tokens": 290761402.0,
      "step": 8359
    },
    {
      "epoch": 1.552460538532962,
      "grad_norm": 1.4219333312690159,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8733576536178589,
      "num_tokens": 290800454.0,
      "step": 8360
    },
    {
      "epoch": 1.5526462395543175,
      "grad_norm": 1.4395918042150986,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8650740385055542,
      "num_tokens": 290840412.0,
      "step": 8361
    },
    {
      "epoch": 1.5528319405756732,
      "grad_norm": 1.430567346602936,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8837592601776123,
      "num_tokens": 290873941.0,
      "step": 8362
    },
    {
      "epoch": 1.553017641597029,
      "grad_norm": 1.4531146896329585,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8796428442001343,
      "num_tokens": 290907524.0,
      "step": 8363
    },
    {
      "epoch": 1.5532033426183844,
      "grad_norm": 1.35696872662313,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8748680949211121,
      "num_tokens": 290949416.0,
      "step": 8364
    },
    {
      "epoch": 1.55338904363974,
      "grad_norm": 1.5492362017352288,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8733034133911133,
      "num_tokens": 290982040.0,
      "step": 8365
    },
    {
      "epoch": 1.5535747446610957,
      "grad_norm": 1.5540481152240717,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8800468444824219,
      "num_tokens": 291013025.0,
      "step": 8366
    },
    {
      "epoch": 1.5537604456824514,
      "grad_norm": 1.5785731307411188,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8652403354644775,
      "num_tokens": 291048487.0,
      "step": 8367
    },
    {
      "epoch": 1.5539461467038067,
      "grad_norm": 1.55402236828842,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8704357147216797,
      "num_tokens": 291082184.0,
      "step": 8368
    },
    {
      "epoch": 1.5541318477251624,
      "grad_norm": 1.4619810699827926,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8884174823760986,
      "num_tokens": 291116088.0,
      "step": 8369
    },
    {
      "epoch": 1.5543175487465182,
      "grad_norm": 1.530966309250346,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8766643404960632,
      "num_tokens": 291150495.0,
      "step": 8370
    },
    {
      "epoch": 1.5545032497678737,
      "grad_norm": 1.5269808258649125,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8840669393539429,
      "num_tokens": 291185723.0,
      "step": 8371
    },
    {
      "epoch": 1.5546889507892292,
      "grad_norm": 1.525866877661567,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.874911904335022,
      "num_tokens": 291222219.0,
      "step": 8372
    },
    {
      "epoch": 1.554874651810585,
      "grad_norm": 1.552112816630886,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8724882006645203,
      "num_tokens": 291255874.0,
      "step": 8373
    },
    {
      "epoch": 1.5550603528319407,
      "grad_norm": 1.36184677787869,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8790565729141235,
      "num_tokens": 291295502.0,
      "step": 8374
    },
    {
      "epoch": 1.5552460538532962,
      "grad_norm": 1.4791209676006767,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8759100437164307,
      "num_tokens": 291330167.0,
      "step": 8375
    },
    {
      "epoch": 1.5554317548746517,
      "grad_norm": 1.4756536393746436,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8814114332199097,
      "num_tokens": 291367056.0,
      "step": 8376
    },
    {
      "epoch": 1.5556174558960074,
      "grad_norm": 1.5735158454886458,
      "learning_rate": 1e-06,
      "loss": 0.3886,
      "mean_token_accuracy": 0.8660810589790344,
      "num_tokens": 291399814.0,
      "step": 8377
    },
    {
      "epoch": 1.5558031569173632,
      "grad_norm": 1.622002301660279,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8798792362213135,
      "num_tokens": 291428953.0,
      "step": 8378
    },
    {
      "epoch": 1.5559888579387187,
      "grad_norm": 1.720596134411838,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8776262402534485,
      "num_tokens": 291454683.0,
      "step": 8379
    },
    {
      "epoch": 1.5561745589600742,
      "grad_norm": 1.5105445431920574,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.875802218914032,
      "num_tokens": 291489718.0,
      "step": 8380
    },
    {
      "epoch": 1.55636025998143,
      "grad_norm": 1.4792194514524617,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8802044987678528,
      "num_tokens": 291523224.0,
      "step": 8381
    },
    {
      "epoch": 1.5565459610027856,
      "grad_norm": 1.5463816151827139,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8698481321334839,
      "num_tokens": 291559281.0,
      "step": 8382
    },
    {
      "epoch": 1.5567316620241411,
      "grad_norm": 1.4963620758400167,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8769979476928711,
      "num_tokens": 291592347.0,
      "step": 8383
    },
    {
      "epoch": 1.5569173630454967,
      "grad_norm": 1.3568078917894224,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8879632353782654,
      "num_tokens": 291631396.0,
      "step": 8384
    },
    {
      "epoch": 1.5571030640668524,
      "grad_norm": 1.5439960846116267,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8649914860725403,
      "num_tokens": 291668073.0,
      "step": 8385
    },
    {
      "epoch": 1.5572887650882081,
      "grad_norm": 1.557407505418402,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8825459480285645,
      "num_tokens": 291700034.0,
      "step": 8386
    },
    {
      "epoch": 1.5574744661095636,
      "grad_norm": 1.5735879022735562,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8833951950073242,
      "num_tokens": 291729278.0,
      "step": 8387
    },
    {
      "epoch": 1.5576601671309191,
      "grad_norm": 1.6153643768887582,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8655651807785034,
      "num_tokens": 291760673.0,
      "step": 8388
    },
    {
      "epoch": 1.5578458681522749,
      "grad_norm": 1.441321304532145,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.880338728427887,
      "num_tokens": 291797045.0,
      "step": 8389
    },
    {
      "epoch": 1.5580315691736306,
      "grad_norm": 1.3661000462334965,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8813856840133667,
      "num_tokens": 291837494.0,
      "step": 8390
    },
    {
      "epoch": 1.5582172701949861,
      "grad_norm": 1.620427968163001,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8677295446395874,
      "num_tokens": 291867330.0,
      "step": 8391
    },
    {
      "epoch": 1.5584029712163416,
      "grad_norm": 1.4959268885740717,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8745744228363037,
      "num_tokens": 291903214.0,
      "step": 8392
    },
    {
      "epoch": 1.5585886722376974,
      "grad_norm": 1.3720756271023347,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8855365514755249,
      "num_tokens": 291942110.0,
      "step": 8393
    },
    {
      "epoch": 1.5587743732590529,
      "grad_norm": 1.5322405046158074,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8694421052932739,
      "num_tokens": 291978351.0,
      "step": 8394
    },
    {
      "epoch": 1.5589600742804084,
      "grad_norm": 1.4320404225226475,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8882172107696533,
      "num_tokens": 292013662.0,
      "step": 8395
    },
    {
      "epoch": 1.5591457753017641,
      "grad_norm": 1.4581874569895228,
      "learning_rate": 1e-06,
      "loss": 0.3926,
      "mean_token_accuracy": 0.8699434995651245,
      "num_tokens": 292050094.0,
      "step": 8396
    },
    {
      "epoch": 1.5593314763231199,
      "grad_norm": 1.540760924516187,
      "learning_rate": 1e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8694692850112915,
      "num_tokens": 292079880.0,
      "step": 8397
    },
    {
      "epoch": 1.5595171773444754,
      "grad_norm": 1.3866747911703927,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8756887912750244,
      "num_tokens": 292116374.0,
      "step": 8398
    },
    {
      "epoch": 1.5597028783658309,
      "grad_norm": 1.7828823536567413,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8801314234733582,
      "num_tokens": 292140749.0,
      "step": 8399
    },
    {
      "epoch": 1.5598885793871866,
      "grad_norm": 1.4821537131056475,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8861109018325806,
      "num_tokens": 292175314.0,
      "step": 8400
    },
    {
      "epoch": 1.5600742804085423,
      "grad_norm": 1.588967046773342,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8764025568962097,
      "num_tokens": 292205349.0,
      "step": 8401
    },
    {
      "epoch": 1.5602599814298979,
      "grad_norm": 1.5182504572265205,
      "learning_rate": 1e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8655938506126404,
      "num_tokens": 292241693.0,
      "step": 8402
    },
    {
      "epoch": 1.5604456824512534,
      "grad_norm": 1.8400731598941051,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8672808408737183,
      "num_tokens": 292264962.0,
      "step": 8403
    },
    {
      "epoch": 1.560631383472609,
      "grad_norm": 1.517514970414971,
      "learning_rate": 1e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8689426183700562,
      "num_tokens": 292302606.0,
      "step": 8404
    },
    {
      "epoch": 1.5608170844939648,
      "grad_norm": 1.5591528201362506,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8845426440238953,
      "num_tokens": 292333019.0,
      "step": 8405
    },
    {
      "epoch": 1.5610027855153203,
      "grad_norm": 1.4806066291975106,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8818138837814331,
      "num_tokens": 292368295.0,
      "step": 8406
    },
    {
      "epoch": 1.5611884865366759,
      "grad_norm": 1.563071007789881,
      "learning_rate": 1e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8504447340965271,
      "num_tokens": 292405796.0,
      "step": 8407
    },
    {
      "epoch": 1.5613741875580316,
      "grad_norm": 1.4647601877788479,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8743281960487366,
      "num_tokens": 292440651.0,
      "step": 8408
    },
    {
      "epoch": 1.5615598885793873,
      "grad_norm": 1.5086814719384594,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8803509473800659,
      "num_tokens": 292473922.0,
      "step": 8409
    },
    {
      "epoch": 1.5617455896007428,
      "grad_norm": 1.5577341233430442,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.887138843536377,
      "num_tokens": 292505768.0,
      "step": 8410
    },
    {
      "epoch": 1.5619312906220983,
      "grad_norm": 1.5431075168338926,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8759523630142212,
      "num_tokens": 292541020.0,
      "step": 8411
    },
    {
      "epoch": 1.562116991643454,
      "grad_norm": 1.3769841129871025,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8905789852142334,
      "num_tokens": 292574532.0,
      "step": 8412
    },
    {
      "epoch": 1.5623026926648098,
      "grad_norm": 1.5466797642794388,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.87218177318573,
      "num_tokens": 292611832.0,
      "step": 8413
    },
    {
      "epoch": 1.5624883936861653,
      "grad_norm": 1.445506004466311,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8733454942703247,
      "num_tokens": 292654116.0,
      "step": 8414
    },
    {
      "epoch": 1.5626740947075208,
      "grad_norm": 1.622934071324913,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8709019422531128,
      "num_tokens": 292690050.0,
      "step": 8415
    },
    {
      "epoch": 1.5628597957288766,
      "grad_norm": 1.6630475822749313,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.880835771560669,
      "num_tokens": 292720376.0,
      "step": 8416
    },
    {
      "epoch": 1.563045496750232,
      "grad_norm": 1.4615829650841354,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8756543397903442,
      "num_tokens": 292756788.0,
      "step": 8417
    },
    {
      "epoch": 1.5632311977715876,
      "grad_norm": 1.568271062100131,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8748859167098999,
      "num_tokens": 292789723.0,
      "step": 8418
    },
    {
      "epoch": 1.5634168987929433,
      "grad_norm": 1.3921372724301735,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8855528831481934,
      "num_tokens": 292828728.0,
      "step": 8419
    },
    {
      "epoch": 1.563602599814299,
      "grad_norm": 1.606400423052563,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8802312612533569,
      "num_tokens": 292861062.0,
      "step": 8420
    },
    {
      "epoch": 1.5637883008356546,
      "grad_norm": 1.4281045401826407,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8672627210617065,
      "num_tokens": 292900899.0,
      "step": 8421
    },
    {
      "epoch": 1.56397400185701,
      "grad_norm": 1.4127569571249863,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8955811262130737,
      "num_tokens": 292934595.0,
      "step": 8422
    },
    {
      "epoch": 1.5641597028783658,
      "grad_norm": 1.3534844265816661,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8837082386016846,
      "num_tokens": 292977381.0,
      "step": 8423
    },
    {
      "epoch": 1.5643454038997215,
      "grad_norm": 1.4638251343675348,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8696391582489014,
      "num_tokens": 293018244.0,
      "step": 8424
    },
    {
      "epoch": 1.564531104921077,
      "grad_norm": 1.4716522925147928,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8729642629623413,
      "num_tokens": 293055135.0,
      "step": 8425
    },
    {
      "epoch": 1.5647168059424326,
      "grad_norm": 1.4375371722471804,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8864201307296753,
      "num_tokens": 293090607.0,
      "step": 8426
    },
    {
      "epoch": 1.5649025069637883,
      "grad_norm": 1.4483261929053295,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.877040684223175,
      "num_tokens": 293126515.0,
      "step": 8427
    },
    {
      "epoch": 1.565088207985144,
      "grad_norm": 1.3608984092239662,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8874979019165039,
      "num_tokens": 293164874.0,
      "step": 8428
    },
    {
      "epoch": 1.5652739090064995,
      "grad_norm": 1.5783102720363733,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.880062997341156,
      "num_tokens": 293195909.0,
      "step": 8429
    },
    {
      "epoch": 1.565459610027855,
      "grad_norm": 1.5016304501994082,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.866519033908844,
      "num_tokens": 293235701.0,
      "step": 8430
    },
    {
      "epoch": 1.5656453110492108,
      "grad_norm": 1.5147126395165287,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8795618414878845,
      "num_tokens": 293270121.0,
      "step": 8431
    },
    {
      "epoch": 1.5658310120705665,
      "grad_norm": 1.559049216514573,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8780015707015991,
      "num_tokens": 293305215.0,
      "step": 8432
    },
    {
      "epoch": 1.566016713091922,
      "grad_norm": 1.443492140367729,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8762093186378479,
      "num_tokens": 293339980.0,
      "step": 8433
    },
    {
      "epoch": 1.5662024141132775,
      "grad_norm": 1.5359194131212646,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8679965138435364,
      "num_tokens": 293376092.0,
      "step": 8434
    },
    {
      "epoch": 1.5663881151346333,
      "grad_norm": 1.5695328247148252,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8664516806602478,
      "num_tokens": 293408291.0,
      "step": 8435
    },
    {
      "epoch": 1.566573816155989,
      "grad_norm": 1.3383341921630285,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8771926164627075,
      "num_tokens": 293452762.0,
      "step": 8436
    },
    {
      "epoch": 1.5667595171773445,
      "grad_norm": 1.4733756714322468,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8683865070343018,
      "num_tokens": 293490686.0,
      "step": 8437
    },
    {
      "epoch": 1.5669452181987,
      "grad_norm": 1.542289230161426,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.873903751373291,
      "num_tokens": 293523735.0,
      "step": 8438
    },
    {
      "epoch": 1.5671309192200558,
      "grad_norm": 1.5066411655171275,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8685965538024902,
      "num_tokens": 293558896.0,
      "step": 8439
    },
    {
      "epoch": 1.5673166202414113,
      "grad_norm": 1.5891619313940653,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8763688206672668,
      "num_tokens": 293588717.0,
      "step": 8440
    },
    {
      "epoch": 1.5675023212627668,
      "grad_norm": 1.723128667205149,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8998186588287354,
      "num_tokens": 293614147.0,
      "step": 8441
    },
    {
      "epoch": 1.5676880222841225,
      "grad_norm": 1.4688394671295724,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8759108781814575,
      "num_tokens": 293650699.0,
      "step": 8442
    },
    {
      "epoch": 1.5678737233054783,
      "grad_norm": 1.5785564599105733,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8799696564674377,
      "num_tokens": 293682693.0,
      "step": 8443
    },
    {
      "epoch": 1.5680594243268338,
      "grad_norm": 1.6838488181522864,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8726961612701416,
      "num_tokens": 293709603.0,
      "step": 8444
    },
    {
      "epoch": 1.5682451253481893,
      "grad_norm": 1.6513688303837586,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.87453693151474,
      "num_tokens": 293739063.0,
      "step": 8445
    },
    {
      "epoch": 1.568430826369545,
      "grad_norm": 1.5800266670959435,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8832476139068604,
      "num_tokens": 293767706.0,
      "step": 8446
    },
    {
      "epoch": 1.5686165273909007,
      "grad_norm": 1.5237446513447153,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8864007592201233,
      "num_tokens": 293803757.0,
      "step": 8447
    },
    {
      "epoch": 1.5688022284122562,
      "grad_norm": 1.7280451319111563,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.887578010559082,
      "num_tokens": 293829213.0,
      "step": 8448
    },
    {
      "epoch": 1.5689879294336118,
      "grad_norm": 1.6050216545042424,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8622354865074158,
      "num_tokens": 293861980.0,
      "step": 8449
    },
    {
      "epoch": 1.5691736304549675,
      "grad_norm": 1.5264462836769574,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8833101987838745,
      "num_tokens": 293892536.0,
      "step": 8450
    },
    {
      "epoch": 1.5693593314763232,
      "grad_norm": 1.4774323280625639,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8733485341072083,
      "num_tokens": 293926329.0,
      "step": 8451
    },
    {
      "epoch": 1.5695450324976787,
      "grad_norm": 1.5183557623093964,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8793082237243652,
      "num_tokens": 293959343.0,
      "step": 8452
    },
    {
      "epoch": 1.5697307335190342,
      "grad_norm": 1.5502400302322068,
      "learning_rate": 1e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.861234724521637,
      "num_tokens": 293997189.0,
      "step": 8453
    },
    {
      "epoch": 1.56991643454039,
      "grad_norm": 1.6368536251174457,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8685212731361389,
      "num_tokens": 294030406.0,
      "step": 8454
    },
    {
      "epoch": 1.5701021355617457,
      "grad_norm": 1.512079208092707,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8870858550071716,
      "num_tokens": 294063129.0,
      "step": 8455
    },
    {
      "epoch": 1.5702878365831012,
      "grad_norm": 1.7130562706841914,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8654443621635437,
      "num_tokens": 294093149.0,
      "step": 8456
    },
    {
      "epoch": 1.5704735376044567,
      "grad_norm": 1.4926034441336262,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.875704288482666,
      "num_tokens": 294127407.0,
      "step": 8457
    },
    {
      "epoch": 1.5706592386258125,
      "grad_norm": 1.456352525116874,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8809522390365601,
      "num_tokens": 294160972.0,
      "step": 8458
    },
    {
      "epoch": 1.5708449396471682,
      "grad_norm": 1.8483920525696458,
      "learning_rate": 1e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.854210376739502,
      "num_tokens": 294188645.0,
      "step": 8459
    },
    {
      "epoch": 1.5710306406685237,
      "grad_norm": 1.5804335665835272,
      "learning_rate": 1e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8556630611419678,
      "num_tokens": 294224716.0,
      "step": 8460
    },
    {
      "epoch": 1.5712163416898792,
      "grad_norm": 1.4554335269325267,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8725131750106812,
      "num_tokens": 294264692.0,
      "step": 8461
    },
    {
      "epoch": 1.571402042711235,
      "grad_norm": 1.5015530771260341,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.8962671756744385,
      "num_tokens": 294294765.0,
      "step": 8462
    },
    {
      "epoch": 1.5715877437325907,
      "grad_norm": 1.4728868042081935,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8898100852966309,
      "num_tokens": 294325661.0,
      "step": 8463
    },
    {
      "epoch": 1.571773444753946,
      "grad_norm": 1.404719318421167,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8705649971961975,
      "num_tokens": 294368538.0,
      "step": 8464
    },
    {
      "epoch": 1.5719591457753017,
      "grad_norm": 1.385241810600339,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8828874826431274,
      "num_tokens": 294406837.0,
      "step": 8465
    },
    {
      "epoch": 1.5721448467966574,
      "grad_norm": 1.5073466958651038,
      "learning_rate": 1e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8578513860702515,
      "num_tokens": 294447253.0,
      "step": 8466
    },
    {
      "epoch": 1.572330547818013,
      "grad_norm": 1.5207386845318067,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8717177510261536,
      "num_tokens": 294481500.0,
      "step": 8467
    },
    {
      "epoch": 1.5725162488393685,
      "grad_norm": 1.4418982347695402,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8767788410186768,
      "num_tokens": 294519385.0,
      "step": 8468
    },
    {
      "epoch": 1.5727019498607242,
      "grad_norm": 1.3971804919207778,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.8830467462539673,
      "num_tokens": 294560565.0,
      "step": 8469
    },
    {
      "epoch": 1.57288765088208,
      "grad_norm": 1.4427911406093183,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8682584762573242,
      "num_tokens": 294600192.0,
      "step": 8470
    },
    {
      "epoch": 1.5730733519034354,
      "grad_norm": 1.5206352469732993,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8837313652038574,
      "num_tokens": 294639860.0,
      "step": 8471
    },
    {
      "epoch": 1.573259052924791,
      "grad_norm": 1.6627435660138088,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8677636384963989,
      "num_tokens": 294672647.0,
      "step": 8472
    },
    {
      "epoch": 1.5734447539461467,
      "grad_norm": 1.5419644921603615,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8643709421157837,
      "num_tokens": 294705518.0,
      "step": 8473
    },
    {
      "epoch": 1.5736304549675024,
      "grad_norm": 1.6080506208840717,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8682483434677124,
      "num_tokens": 294740593.0,
      "step": 8474
    },
    {
      "epoch": 1.573816155988858,
      "grad_norm": 1.3987851593028533,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8857033252716064,
      "num_tokens": 294779814.0,
      "step": 8475
    },
    {
      "epoch": 1.5740018570102134,
      "grad_norm": 1.4568077994000943,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.869245171546936,
      "num_tokens": 294815075.0,
      "step": 8476
    },
    {
      "epoch": 1.5741875580315692,
      "grad_norm": 1.5897851644477414,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8857929110527039,
      "num_tokens": 294848854.0,
      "step": 8477
    },
    {
      "epoch": 1.574373259052925,
      "grad_norm": 1.4794590025179541,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8771688342094421,
      "num_tokens": 294884435.0,
      "step": 8478
    },
    {
      "epoch": 1.5745589600742804,
      "grad_norm": 1.5391743643808933,
      "learning_rate": 1e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8692364692687988,
      "num_tokens": 294920003.0,
      "step": 8479
    },
    {
      "epoch": 1.574744661095636,
      "grad_norm": 1.6021726065642927,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8817528486251831,
      "num_tokens": 294956468.0,
      "step": 8480
    },
    {
      "epoch": 1.5749303621169917,
      "grad_norm": 1.4752690381077902,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8884973526000977,
      "num_tokens": 294991432.0,
      "step": 8481
    },
    {
      "epoch": 1.5751160631383474,
      "grad_norm": 1.442213454977858,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.873210072517395,
      "num_tokens": 295025304.0,
      "step": 8482
    },
    {
      "epoch": 1.575301764159703,
      "grad_norm": 1.6979928735946097,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8882026672363281,
      "num_tokens": 295050229.0,
      "step": 8483
    },
    {
      "epoch": 1.5754874651810584,
      "grad_norm": 1.7032756756394716,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8695051670074463,
      "num_tokens": 295079289.0,
      "step": 8484
    },
    {
      "epoch": 1.5756731662024142,
      "grad_norm": 1.5562292252061285,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8716882467269897,
      "num_tokens": 295109354.0,
      "step": 8485
    },
    {
      "epoch": 1.5758588672237699,
      "grad_norm": 1.53995875126251,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8773580193519592,
      "num_tokens": 295144322.0,
      "step": 8486
    },
    {
      "epoch": 1.5760445682451254,
      "grad_norm": 1.4059051487783973,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8762688040733337,
      "num_tokens": 295183023.0,
      "step": 8487
    },
    {
      "epoch": 1.576230269266481,
      "grad_norm": 1.6117578943187951,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8738529086112976,
      "num_tokens": 295215628.0,
      "step": 8488
    },
    {
      "epoch": 1.5764159702878366,
      "grad_norm": 1.6606451138466378,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8797525763511658,
      "num_tokens": 295251797.0,
      "step": 8489
    },
    {
      "epoch": 1.5766016713091922,
      "grad_norm": 1.4529881946412881,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8781824111938477,
      "num_tokens": 295287502.0,
      "step": 8490
    },
    {
      "epoch": 1.5767873723305477,
      "grad_norm": 1.405874410057102,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8808678388595581,
      "num_tokens": 295326267.0,
      "step": 8491
    },
    {
      "epoch": 1.5769730733519034,
      "grad_norm": 1.5695444121367677,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8674815893173218,
      "num_tokens": 295358727.0,
      "step": 8492
    },
    {
      "epoch": 1.5771587743732591,
      "grad_norm": 1.5207261847034637,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8765507936477661,
      "num_tokens": 295390432.0,
      "step": 8493
    },
    {
      "epoch": 1.5773444753946146,
      "grad_norm": 1.4582792531724755,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8788079023361206,
      "num_tokens": 295430601.0,
      "step": 8494
    },
    {
      "epoch": 1.5775301764159702,
      "grad_norm": 1.7152909977388349,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8794471025466919,
      "num_tokens": 295459213.0,
      "step": 8495
    },
    {
      "epoch": 1.5777158774373259,
      "grad_norm": 1.379725343951822,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8932356834411621,
      "num_tokens": 295493498.0,
      "step": 8496
    },
    {
      "epoch": 1.5779015784586816,
      "grad_norm": 1.4988803256078778,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.875003457069397,
      "num_tokens": 295529781.0,
      "step": 8497
    },
    {
      "epoch": 1.5780872794800371,
      "grad_norm": 1.558006997604094,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8733047246932983,
      "num_tokens": 295562824.0,
      "step": 8498
    },
    {
      "epoch": 1.5782729805013926,
      "grad_norm": 1.4892687414507055,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8842645883560181,
      "num_tokens": 295597741.0,
      "step": 8499
    },
    {
      "epoch": 1.5784586815227484,
      "grad_norm": 1.526393816356952,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8614699244499207,
      "num_tokens": 295633866.0,
      "step": 8500
    },
    {
      "epoch": 1.578644382544104,
      "grad_norm": 1.351366680509792,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8843350410461426,
      "num_tokens": 295674074.0,
      "step": 8501
    },
    {
      "epoch": 1.5788300835654596,
      "grad_norm": 1.6513426250312662,
      "learning_rate": 1e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.8750010132789612,
      "num_tokens": 295701839.0,
      "step": 8502
    },
    {
      "epoch": 1.5790157845868151,
      "grad_norm": 1.4274789090320807,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8714631795883179,
      "num_tokens": 295739761.0,
      "step": 8503
    },
    {
      "epoch": 1.5792014856081709,
      "grad_norm": 1.5775023485402377,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.86785888671875,
      "num_tokens": 295771774.0,
      "step": 8504
    },
    {
      "epoch": 1.5793871866295266,
      "grad_norm": 1.61310513873193,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8772858381271362,
      "num_tokens": 295803774.0,
      "step": 8505
    },
    {
      "epoch": 1.579572887650882,
      "grad_norm": 1.5568168099860094,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8787767887115479,
      "num_tokens": 295834141.0,
      "step": 8506
    },
    {
      "epoch": 1.5797585886722376,
      "grad_norm": 1.514121644935839,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8815069794654846,
      "num_tokens": 295863045.0,
      "step": 8507
    },
    {
      "epoch": 1.5799442896935934,
      "grad_norm": 1.6334844871985894,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8728595972061157,
      "num_tokens": 295891572.0,
      "step": 8508
    },
    {
      "epoch": 1.580129990714949,
      "grad_norm": 1.439364040631702,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8788332939147949,
      "num_tokens": 295924765.0,
      "step": 8509
    },
    {
      "epoch": 1.5803156917363046,
      "grad_norm": 1.268309572864278,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8777457475662231,
      "num_tokens": 295968991.0,
      "step": 8510
    },
    {
      "epoch": 1.58050139275766,
      "grad_norm": 1.4320977223643339,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8804868459701538,
      "num_tokens": 296010701.0,
      "step": 8511
    },
    {
      "epoch": 1.5806870937790158,
      "grad_norm": 1.4583152705994218,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8865745067596436,
      "num_tokens": 296046402.0,
      "step": 8512
    },
    {
      "epoch": 1.5808727948003714,
      "grad_norm": 1.693768125490322,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.873487114906311,
      "num_tokens": 296081632.0,
      "step": 8513
    },
    {
      "epoch": 1.5810584958217269,
      "grad_norm": 1.4533983513321267,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8601677417755127,
      "num_tokens": 296123418.0,
      "step": 8514
    },
    {
      "epoch": 1.5812441968430826,
      "grad_norm": 1.5421303996148032,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8871861696243286,
      "num_tokens": 296159246.0,
      "step": 8515
    },
    {
      "epoch": 1.5814298978644383,
      "grad_norm": 1.5787769665874878,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.874996542930603,
      "num_tokens": 296191477.0,
      "step": 8516
    },
    {
      "epoch": 1.5816155988857938,
      "grad_norm": 1.4281611740344475,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8940513134002686,
      "num_tokens": 296226349.0,
      "step": 8517
    },
    {
      "epoch": 1.5818012999071493,
      "grad_norm": 1.4512896647216202,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8848708271980286,
      "num_tokens": 296260931.0,
      "step": 8518
    },
    {
      "epoch": 1.581987000928505,
      "grad_norm": 1.5060333405375492,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8843152523040771,
      "num_tokens": 296295257.0,
      "step": 8519
    },
    {
      "epoch": 1.5821727019498608,
      "grad_norm": 1.459333534809537,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8852104544639587,
      "num_tokens": 296330166.0,
      "step": 8520
    },
    {
      "epoch": 1.5823584029712163,
      "grad_norm": 1.633900173514396,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8730083703994751,
      "num_tokens": 296359119.0,
      "step": 8521
    },
    {
      "epoch": 1.5825441039925718,
      "grad_norm": 1.4428180759565887,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8793669939041138,
      "num_tokens": 296393421.0,
      "step": 8522
    },
    {
      "epoch": 1.5827298050139276,
      "grad_norm": 1.4528559134482708,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8816662430763245,
      "num_tokens": 296428900.0,
      "step": 8523
    },
    {
      "epoch": 1.5829155060352833,
      "grad_norm": 1.523007199794162,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8762235045433044,
      "num_tokens": 296464883.0,
      "step": 8524
    },
    {
      "epoch": 1.5831012070566388,
      "grad_norm": 1.4580501447042,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8806788921356201,
      "num_tokens": 296503226.0,
      "step": 8525
    },
    {
      "epoch": 1.5832869080779943,
      "grad_norm": 1.4975261323233267,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8890608549118042,
      "num_tokens": 296534522.0,
      "step": 8526
    },
    {
      "epoch": 1.58347260909935,
      "grad_norm": 1.5666643346963103,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8765783309936523,
      "num_tokens": 296565120.0,
      "step": 8527
    },
    {
      "epoch": 1.5836583101207058,
      "grad_norm": 1.791759743684604,
      "learning_rate": 1e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8627001047134399,
      "num_tokens": 296594373.0,
      "step": 8528
    },
    {
      "epoch": 1.5838440111420613,
      "grad_norm": 1.6552478618991766,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8689987659454346,
      "num_tokens": 296624735.0,
      "step": 8529
    },
    {
      "epoch": 1.5840297121634168,
      "grad_norm": 1.4722455613942786,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8726502060890198,
      "num_tokens": 296662987.0,
      "step": 8530
    },
    {
      "epoch": 1.5842154131847725,
      "grad_norm": 1.4324912553522406,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8812201619148254,
      "num_tokens": 296700429.0,
      "step": 8531
    },
    {
      "epoch": 1.5844011142061283,
      "grad_norm": 1.4617645140053115,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8821848034858704,
      "num_tokens": 296736401.0,
      "step": 8532
    },
    {
      "epoch": 1.5845868152274838,
      "grad_norm": 1.444843993811553,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8825246691703796,
      "num_tokens": 296771108.0,
      "step": 8533
    },
    {
      "epoch": 1.5847725162488393,
      "grad_norm": 1.6572712140658763,
      "learning_rate": 1e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8624820709228516,
      "num_tokens": 296803383.0,
      "step": 8534
    },
    {
      "epoch": 1.584958217270195,
      "grad_norm": 1.66046673560334,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8708126544952393,
      "num_tokens": 296833397.0,
      "step": 8535
    },
    {
      "epoch": 1.5851439182915508,
      "grad_norm": 1.4904945927896438,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8722823858261108,
      "num_tokens": 296869152.0,
      "step": 8536
    },
    {
      "epoch": 1.585329619312906,
      "grad_norm": 1.6135684901672025,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8728673458099365,
      "num_tokens": 296900045.0,
      "step": 8537
    },
    {
      "epoch": 1.5855153203342618,
      "grad_norm": 1.4141543734483981,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.883664071559906,
      "num_tokens": 296938847.0,
      "step": 8538
    },
    {
      "epoch": 1.5857010213556175,
      "grad_norm": 1.6272671017743017,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8801708221435547,
      "num_tokens": 296967787.0,
      "step": 8539
    },
    {
      "epoch": 1.585886722376973,
      "grad_norm": 1.3463286776027834,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8874967098236084,
      "num_tokens": 297005805.0,
      "step": 8540
    },
    {
      "epoch": 1.5860724233983285,
      "grad_norm": 1.4569656641428448,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8865742683410645,
      "num_tokens": 297037956.0,
      "step": 8541
    },
    {
      "epoch": 1.5862581244196843,
      "grad_norm": 1.478244849487411,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8846617937088013,
      "num_tokens": 297068637.0,
      "step": 8542
    },
    {
      "epoch": 1.58644382544104,
      "grad_norm": 1.5628234612936356,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8829552531242371,
      "num_tokens": 297100448.0,
      "step": 8543
    },
    {
      "epoch": 1.5866295264623955,
      "grad_norm": 1.609606087483658,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8790842890739441,
      "num_tokens": 297134122.0,
      "step": 8544
    },
    {
      "epoch": 1.586815227483751,
      "grad_norm": 1.4130008755397083,
      "learning_rate": 1e-06,
      "loss": 0.3084,
      "mean_token_accuracy": 0.891316294670105,
      "num_tokens": 297169452.0,
      "step": 8545
    },
    {
      "epoch": 1.5870009285051068,
      "grad_norm": 1.404352677460016,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8765307068824768,
      "num_tokens": 297206510.0,
      "step": 8546
    },
    {
      "epoch": 1.5871866295264625,
      "grad_norm": 1.4976134146951245,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8843193650245667,
      "num_tokens": 297241554.0,
      "step": 8547
    },
    {
      "epoch": 1.587372330547818,
      "grad_norm": 1.5834325811576606,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8699953556060791,
      "num_tokens": 297273703.0,
      "step": 8548
    },
    {
      "epoch": 1.5875580315691735,
      "grad_norm": 1.516551813080451,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8743398785591125,
      "num_tokens": 297305140.0,
      "step": 8549
    },
    {
      "epoch": 1.5877437325905293,
      "grad_norm": 1.4365527612024622,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8704574108123779,
      "num_tokens": 297340851.0,
      "step": 8550
    },
    {
      "epoch": 1.587929433611885,
      "grad_norm": 1.4325639741537943,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8764814734458923,
      "num_tokens": 297374103.0,
      "step": 8551
    },
    {
      "epoch": 1.5881151346332405,
      "grad_norm": 1.5483884214549308,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8668245077133179,
      "num_tokens": 297410708.0,
      "step": 8552
    },
    {
      "epoch": 1.588300835654596,
      "grad_norm": 1.4583119224655954,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8868914842605591,
      "num_tokens": 297445553.0,
      "step": 8553
    },
    {
      "epoch": 1.5884865366759517,
      "grad_norm": 1.5467335757413125,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8627230525016785,
      "num_tokens": 297479601.0,
      "step": 8554
    },
    {
      "epoch": 1.5886722376973075,
      "grad_norm": 1.370359907729399,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8628342151641846,
      "num_tokens": 297520349.0,
      "step": 8555
    },
    {
      "epoch": 1.588857938718663,
      "grad_norm": 1.5433512606659936,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8738877177238464,
      "num_tokens": 297555323.0,
      "step": 8556
    },
    {
      "epoch": 1.5890436397400185,
      "grad_norm": 1.540439511151258,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8737950921058655,
      "num_tokens": 297588350.0,
      "step": 8557
    },
    {
      "epoch": 1.5892293407613742,
      "grad_norm": 1.527579943900749,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8741289377212524,
      "num_tokens": 297622122.0,
      "step": 8558
    },
    {
      "epoch": 1.58941504178273,
      "grad_norm": 1.5862711838069627,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8663522005081177,
      "num_tokens": 297651372.0,
      "step": 8559
    },
    {
      "epoch": 1.5896007428040855,
      "grad_norm": 1.362525947706534,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8742766976356506,
      "num_tokens": 297691835.0,
      "step": 8560
    },
    {
      "epoch": 1.589786443825441,
      "grad_norm": 1.5684934595106987,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8750544786453247,
      "num_tokens": 297724683.0,
      "step": 8561
    },
    {
      "epoch": 1.5899721448467967,
      "grad_norm": 1.4467792194155322,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.879766047000885,
      "num_tokens": 297758639.0,
      "step": 8562
    },
    {
      "epoch": 1.5901578458681522,
      "grad_norm": 1.387862370111404,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8842918872833252,
      "num_tokens": 297794831.0,
      "step": 8563
    },
    {
      "epoch": 1.5903435468895077,
      "grad_norm": 1.3951926075163956,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8813445568084717,
      "num_tokens": 297835780.0,
      "step": 8564
    },
    {
      "epoch": 1.5905292479108635,
      "grad_norm": 1.586003167934257,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8618375062942505,
      "num_tokens": 297870966.0,
      "step": 8565
    },
    {
      "epoch": 1.5907149489322192,
      "grad_norm": 1.4183556743711567,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8760197758674622,
      "num_tokens": 297910699.0,
      "step": 8566
    },
    {
      "epoch": 1.5909006499535747,
      "grad_norm": 1.449615739051183,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8883458375930786,
      "num_tokens": 297948188.0,
      "step": 8567
    },
    {
      "epoch": 1.5910863509749302,
      "grad_norm": 1.4825610388271755,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.876018762588501,
      "num_tokens": 297986693.0,
      "step": 8568
    },
    {
      "epoch": 1.591272051996286,
      "grad_norm": 1.506237993949713,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8731203675270081,
      "num_tokens": 298020623.0,
      "step": 8569
    },
    {
      "epoch": 1.5914577530176417,
      "grad_norm": 1.4889720273532263,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8765544295310974,
      "num_tokens": 298057118.0,
      "step": 8570
    },
    {
      "epoch": 1.5916434540389972,
      "grad_norm": 1.6056865022479712,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.883090615272522,
      "num_tokens": 298087483.0,
      "step": 8571
    },
    {
      "epoch": 1.5918291550603527,
      "grad_norm": 1.5192590042935434,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8715455532073975,
      "num_tokens": 298122740.0,
      "step": 8572
    },
    {
      "epoch": 1.5920148560817085,
      "grad_norm": 1.6266734397712912,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8785720467567444,
      "num_tokens": 298152414.0,
      "step": 8573
    },
    {
      "epoch": 1.5922005571030642,
      "grad_norm": 1.400748522151488,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8688148856163025,
      "num_tokens": 298195385.0,
      "step": 8574
    },
    {
      "epoch": 1.5923862581244197,
      "grad_norm": 1.6409648817808626,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8698222041130066,
      "num_tokens": 298226393.0,
      "step": 8575
    },
    {
      "epoch": 1.5925719591457752,
      "grad_norm": 1.4398110921081626,
      "learning_rate": 1e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8669253587722778,
      "num_tokens": 298265368.0,
      "step": 8576
    },
    {
      "epoch": 1.592757660167131,
      "grad_norm": 1.4376621599794936,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8749627470970154,
      "num_tokens": 298302516.0,
      "step": 8577
    },
    {
      "epoch": 1.5929433611884867,
      "grad_norm": 1.4884679490580845,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.876418948173523,
      "num_tokens": 298341937.0,
      "step": 8578
    },
    {
      "epoch": 1.5931290622098422,
      "grad_norm": 1.6241598877076953,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8747348189353943,
      "num_tokens": 298371560.0,
      "step": 8579
    },
    {
      "epoch": 1.5933147632311977,
      "grad_norm": 1.5829689670936986,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8903349041938782,
      "num_tokens": 298399532.0,
      "step": 8580
    },
    {
      "epoch": 1.5935004642525534,
      "grad_norm": 1.3222914264507792,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8716776967048645,
      "num_tokens": 298442796.0,
      "step": 8581
    },
    {
      "epoch": 1.5936861652739092,
      "grad_norm": 1.5397882203657367,
      "learning_rate": 1e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8655128479003906,
      "num_tokens": 298479451.0,
      "step": 8582
    },
    {
      "epoch": 1.5938718662952647,
      "grad_norm": 1.57116814645516,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8832281827926636,
      "num_tokens": 298512375.0,
      "step": 8583
    },
    {
      "epoch": 1.5940575673166202,
      "grad_norm": 1.4265063064784649,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8851963877677917,
      "num_tokens": 298549428.0,
      "step": 8584
    },
    {
      "epoch": 1.594243268337976,
      "grad_norm": 1.5123563937712237,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8907670974731445,
      "num_tokens": 298579513.0,
      "step": 8585
    },
    {
      "epoch": 1.5944289693593314,
      "grad_norm": 1.5930517107817899,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8650859594345093,
      "num_tokens": 298615209.0,
      "step": 8586
    },
    {
      "epoch": 1.594614670380687,
      "grad_norm": 1.4944713392970759,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8825826048851013,
      "num_tokens": 298647094.0,
      "step": 8587
    },
    {
      "epoch": 1.5948003714020427,
      "grad_norm": 1.3979085029837683,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8762023448944092,
      "num_tokens": 298688117.0,
      "step": 8588
    },
    {
      "epoch": 1.5949860724233984,
      "grad_norm": 1.462506793210534,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8696187734603882,
      "num_tokens": 298724795.0,
      "step": 8589
    },
    {
      "epoch": 1.595171773444754,
      "grad_norm": 1.464841291290535,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.877800703048706,
      "num_tokens": 298762214.0,
      "step": 8590
    },
    {
      "epoch": 1.5953574744661094,
      "grad_norm": 1.4092252191354915,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8786768913269043,
      "num_tokens": 298802676.0,
      "step": 8591
    },
    {
      "epoch": 1.5955431754874652,
      "grad_norm": 1.4158565694892364,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8929551839828491,
      "num_tokens": 298836854.0,
      "step": 8592
    },
    {
      "epoch": 1.595728876508821,
      "grad_norm": 1.4347335892290198,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8818867206573486,
      "num_tokens": 298871763.0,
      "step": 8593
    },
    {
      "epoch": 1.5959145775301764,
      "grad_norm": 1.788123360956725,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8726773262023926,
      "num_tokens": 298903588.0,
      "step": 8594
    },
    {
      "epoch": 1.596100278551532,
      "grad_norm": 1.4186265485393101,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8830386996269226,
      "num_tokens": 298940012.0,
      "step": 8595
    },
    {
      "epoch": 1.5962859795728876,
      "grad_norm": 1.3864281769343352,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8821501135826111,
      "num_tokens": 298977657.0,
      "step": 8596
    },
    {
      "epoch": 1.5964716805942434,
      "grad_norm": 1.4430951583801763,
      "learning_rate": 1e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8685296177864075,
      "num_tokens": 299016191.0,
      "step": 8597
    },
    {
      "epoch": 1.596657381615599,
      "grad_norm": 1.4361629939461393,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8757014274597168,
      "num_tokens": 299055054.0,
      "step": 8598
    },
    {
      "epoch": 1.5968430826369544,
      "grad_norm": 1.4804208112192943,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8807708024978638,
      "num_tokens": 299087055.0,
      "step": 8599
    },
    {
      "epoch": 1.5970287836583101,
      "grad_norm": 1.4144073795025722,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8772658705711365,
      "num_tokens": 299126998.0,
      "step": 8600
    },
    {
      "epoch": 1.5972144846796659,
      "grad_norm": 1.5135159667883569,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8697699904441833,
      "num_tokens": 299163059.0,
      "step": 8601
    },
    {
      "epoch": 1.5974001857010214,
      "grad_norm": 1.5621271374194003,
      "learning_rate": 1e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8610634803771973,
      "num_tokens": 299197627.0,
      "step": 8602
    },
    {
      "epoch": 1.597585886722377,
      "grad_norm": 1.4541998871466892,
      "learning_rate": 1e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8639013767242432,
      "num_tokens": 299238644.0,
      "step": 8603
    },
    {
      "epoch": 1.5977715877437326,
      "grad_norm": 1.365419190910286,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8831615447998047,
      "num_tokens": 299277901.0,
      "step": 8604
    },
    {
      "epoch": 1.5979572887650884,
      "grad_norm": 1.4421556197253103,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8790531158447266,
      "num_tokens": 299312419.0,
      "step": 8605
    },
    {
      "epoch": 1.5981429897864439,
      "grad_norm": 1.3819529020292696,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8932077288627625,
      "num_tokens": 299349584.0,
      "step": 8606
    },
    {
      "epoch": 1.5983286908077994,
      "grad_norm": 1.6502906703117883,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8646613359451294,
      "num_tokens": 299382158.0,
      "step": 8607
    },
    {
      "epoch": 1.5985143918291551,
      "grad_norm": 1.4842385972120866,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8675850629806519,
      "num_tokens": 299417904.0,
      "step": 8608
    },
    {
      "epoch": 1.5987000928505106,
      "grad_norm": 1.5551836579644214,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.879543125629425,
      "num_tokens": 299450165.0,
      "step": 8609
    },
    {
      "epoch": 1.5988857938718661,
      "grad_norm": 1.599819983356372,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.877186119556427,
      "num_tokens": 299480474.0,
      "step": 8610
    },
    {
      "epoch": 1.5990714948932219,
      "grad_norm": 1.3544668054950817,
      "learning_rate": 1e-06,
      "loss": 0.3254,
      "mean_token_accuracy": 0.8885716795921326,
      "num_tokens": 299519574.0,
      "step": 8611
    },
    {
      "epoch": 1.5992571959145776,
      "grad_norm": 1.6817311431173976,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.870375394821167,
      "num_tokens": 299551271.0,
      "step": 8612
    },
    {
      "epoch": 1.5994428969359331,
      "grad_norm": 1.5496992688322313,
      "learning_rate": 1e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8658879399299622,
      "num_tokens": 299586133.0,
      "step": 8613
    },
    {
      "epoch": 1.5996285979572886,
      "grad_norm": 1.4810347495771667,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.880761981010437,
      "num_tokens": 299619372.0,
      "step": 8614
    },
    {
      "epoch": 1.5998142989786444,
      "grad_norm": 1.5300651935847318,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8796811699867249,
      "num_tokens": 299651138.0,
      "step": 8615
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.568759678066548,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8652611970901489,
      "num_tokens": 299688820.0,
      "step": 8616
    },
    {
      "epoch": 1.6001857010213556,
      "grad_norm": 1.6139475857048433,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8846173882484436,
      "num_tokens": 299719521.0,
      "step": 8617
    },
    {
      "epoch": 1.600371402042711,
      "grad_norm": 1.4784603851851672,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.879645824432373,
      "num_tokens": 299755404.0,
      "step": 8618
    },
    {
      "epoch": 1.6005571030640668,
      "grad_norm": 1.4199108414502832,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8870707750320435,
      "num_tokens": 299795213.0,
      "step": 8619
    },
    {
      "epoch": 1.6007428040854226,
      "grad_norm": 1.3629992231995405,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8877701163291931,
      "num_tokens": 299832645.0,
      "step": 8620
    },
    {
      "epoch": 1.600928505106778,
      "grad_norm": 1.5381686511241213,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8790438771247864,
      "num_tokens": 299868241.0,
      "step": 8621
    },
    {
      "epoch": 1.6011142061281336,
      "grad_norm": 1.661978957066758,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8750689029693604,
      "num_tokens": 299900321.0,
      "step": 8622
    },
    {
      "epoch": 1.6012999071494893,
      "grad_norm": 1.4687586675433761,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8842481374740601,
      "num_tokens": 299936570.0,
      "step": 8623
    },
    {
      "epoch": 1.601485608170845,
      "grad_norm": 1.4478863744040555,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8768308162689209,
      "num_tokens": 299978337.0,
      "step": 8624
    },
    {
      "epoch": 1.6016713091922006,
      "grad_norm": 1.4573109978290057,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.87040114402771,
      "num_tokens": 300014330.0,
      "step": 8625
    },
    {
      "epoch": 1.601857010213556,
      "grad_norm": 1.4234810196875889,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8768441081047058,
      "num_tokens": 300051475.0,
      "step": 8626
    },
    {
      "epoch": 1.6020427112349118,
      "grad_norm": 1.4778561328345128,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8799299001693726,
      "num_tokens": 300083937.0,
      "step": 8627
    },
    {
      "epoch": 1.6022284122562676,
      "grad_norm": 1.4558301413633943,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8808500170707703,
      "num_tokens": 300121449.0,
      "step": 8628
    },
    {
      "epoch": 1.602414113277623,
      "grad_norm": 1.548324434929778,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8840115070343018,
      "num_tokens": 300153310.0,
      "step": 8629
    },
    {
      "epoch": 1.6025998142989786,
      "grad_norm": 1.5282339163202532,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8866393566131592,
      "num_tokens": 300189137.0,
      "step": 8630
    },
    {
      "epoch": 1.6027855153203343,
      "grad_norm": 1.6841102338722362,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8853882551193237,
      "num_tokens": 300216279.0,
      "step": 8631
    },
    {
      "epoch": 1.60297121634169,
      "grad_norm": 1.5970018853434276,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8784587383270264,
      "num_tokens": 300248389.0,
      "step": 8632
    },
    {
      "epoch": 1.6031569173630453,
      "grad_norm": 1.3717273484413763,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8868704438209534,
      "num_tokens": 300285726.0,
      "step": 8633
    },
    {
      "epoch": 1.603342618384401,
      "grad_norm": 1.5280440351447986,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8876028060913086,
      "num_tokens": 300319359.0,
      "step": 8634
    },
    {
      "epoch": 1.6035283194057568,
      "grad_norm": 1.740225774518531,
      "learning_rate": 1e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8550933003425598,
      "num_tokens": 300349508.0,
      "step": 8635
    },
    {
      "epoch": 1.6037140204271123,
      "grad_norm": 1.6149413860675232,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8823294639587402,
      "num_tokens": 300378309.0,
      "step": 8636
    },
    {
      "epoch": 1.6038997214484678,
      "grad_norm": 1.4152420910220007,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8690847754478455,
      "num_tokens": 300419248.0,
      "step": 8637
    },
    {
      "epoch": 1.6040854224698236,
      "grad_norm": 1.596897171043369,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8790702223777771,
      "num_tokens": 300456148.0,
      "step": 8638
    },
    {
      "epoch": 1.6042711234911793,
      "grad_norm": 1.597515244354887,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8788175582885742,
      "num_tokens": 300485065.0,
      "step": 8639
    },
    {
      "epoch": 1.6044568245125348,
      "grad_norm": 1.6040246594806271,
      "learning_rate": 1e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.8838694095611572,
      "num_tokens": 300513297.0,
      "step": 8640
    },
    {
      "epoch": 1.6046425255338903,
      "grad_norm": 1.5261801556515353,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8678220510482788,
      "num_tokens": 300547677.0,
      "step": 8641
    },
    {
      "epoch": 1.604828226555246,
      "grad_norm": 1.5433787095062892,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8800814747810364,
      "num_tokens": 300581543.0,
      "step": 8642
    },
    {
      "epoch": 1.6050139275766018,
      "grad_norm": 1.472812795518465,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8768438696861267,
      "num_tokens": 300619057.0,
      "step": 8643
    },
    {
      "epoch": 1.6051996285979573,
      "grad_norm": 1.4523395682043951,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8764564990997314,
      "num_tokens": 300658023.0,
      "step": 8644
    },
    {
      "epoch": 1.6053853296193128,
      "grad_norm": 1.510651972292994,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8760011196136475,
      "num_tokens": 300695315.0,
      "step": 8645
    },
    {
      "epoch": 1.6055710306406685,
      "grad_norm": 1.4862531769454295,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8811511397361755,
      "num_tokens": 300729078.0,
      "step": 8646
    },
    {
      "epoch": 1.6057567316620243,
      "grad_norm": 1.5483387907702746,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8626368045806885,
      "num_tokens": 300761077.0,
      "step": 8647
    },
    {
      "epoch": 1.6059424326833798,
      "grad_norm": 1.5154751869427252,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8835976123809814,
      "num_tokens": 300794459.0,
      "step": 8648
    },
    {
      "epoch": 1.6061281337047353,
      "grad_norm": 1.5629500587578111,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8657084703445435,
      "num_tokens": 300829055.0,
      "step": 8649
    },
    {
      "epoch": 1.606313834726091,
      "grad_norm": 1.4460789536530487,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8678922057151794,
      "num_tokens": 300868992.0,
      "step": 8650
    },
    {
      "epoch": 1.6064995357474467,
      "grad_norm": 1.5567375766073355,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8741583824157715,
      "num_tokens": 300900964.0,
      "step": 8651
    },
    {
      "epoch": 1.6066852367688023,
      "grad_norm": 1.5102444047237673,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8913148641586304,
      "num_tokens": 300929180.0,
      "step": 8652
    },
    {
      "epoch": 1.6068709377901578,
      "grad_norm": 1.4744285139813562,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8941341638565063,
      "num_tokens": 300959158.0,
      "step": 8653
    },
    {
      "epoch": 1.6070566388115135,
      "grad_norm": 1.5450952564196216,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8814629316329956,
      "num_tokens": 300990590.0,
      "step": 8654
    },
    {
      "epoch": 1.6072423398328692,
      "grad_norm": 1.696822071937771,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.860605001449585,
      "num_tokens": 301023257.0,
      "step": 8655
    },
    {
      "epoch": 1.6074280408542247,
      "grad_norm": 1.3385198015613153,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8931589722633362,
      "num_tokens": 301069030.0,
      "step": 8656
    },
    {
      "epoch": 1.6076137418755803,
      "grad_norm": 1.4040727057907825,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8810504674911499,
      "num_tokens": 301108289.0,
      "step": 8657
    },
    {
      "epoch": 1.607799442896936,
      "grad_norm": 1.6836450163369325,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8779902458190918,
      "num_tokens": 301135385.0,
      "step": 8658
    },
    {
      "epoch": 1.6079851439182915,
      "grad_norm": 1.4911966488120831,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8832199573516846,
      "num_tokens": 301172232.0,
      "step": 8659
    },
    {
      "epoch": 1.608170844939647,
      "grad_norm": 1.6120984658463264,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8684442639350891,
      "num_tokens": 301203883.0,
      "step": 8660
    },
    {
      "epoch": 1.6083565459610027,
      "grad_norm": 1.5500231438880945,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8867571353912354,
      "num_tokens": 301239053.0,
      "step": 8661
    },
    {
      "epoch": 1.6085422469823585,
      "grad_norm": 1.3742648107400381,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8754452466964722,
      "num_tokens": 301283798.0,
      "step": 8662
    },
    {
      "epoch": 1.608727948003714,
      "grad_norm": 1.68410926901126,
      "learning_rate": 1e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8579589128494263,
      "num_tokens": 301314118.0,
      "step": 8663
    },
    {
      "epoch": 1.6089136490250695,
      "grad_norm": 1.6462113507307852,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8802465200424194,
      "num_tokens": 301343528.0,
      "step": 8664
    },
    {
      "epoch": 1.6090993500464252,
      "grad_norm": 1.3751753523928978,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.8992264866828918,
      "num_tokens": 301379101.0,
      "step": 8665
    },
    {
      "epoch": 1.609285051067781,
      "grad_norm": 1.5768939272306208,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8815447092056274,
      "num_tokens": 301411071.0,
      "step": 8666
    },
    {
      "epoch": 1.6094707520891365,
      "grad_norm": 1.4253667192448167,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8751015663146973,
      "num_tokens": 301449164.0,
      "step": 8667
    },
    {
      "epoch": 1.609656453110492,
      "grad_norm": 1.449610289442679,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8807733654975891,
      "num_tokens": 301483944.0,
      "step": 8668
    },
    {
      "epoch": 1.6098421541318477,
      "grad_norm": 1.5467488107371261,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8794914484024048,
      "num_tokens": 301519147.0,
      "step": 8669
    },
    {
      "epoch": 1.6100278551532035,
      "grad_norm": 1.4428366928398482,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8825597167015076,
      "num_tokens": 301553174.0,
      "step": 8670
    },
    {
      "epoch": 1.610213556174559,
      "grad_norm": 1.5144491746134598,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8774990439414978,
      "num_tokens": 301587971.0,
      "step": 8671
    },
    {
      "epoch": 1.6103992571959145,
      "grad_norm": 1.4447636163770259,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8878340125083923,
      "num_tokens": 301622315.0,
      "step": 8672
    },
    {
      "epoch": 1.6105849582172702,
      "grad_norm": 1.5115672148533863,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8679511547088623,
      "num_tokens": 301655912.0,
      "step": 8673
    },
    {
      "epoch": 1.610770659238626,
      "grad_norm": 1.4650416871937684,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.87520831823349,
      "num_tokens": 301692661.0,
      "step": 8674
    },
    {
      "epoch": 1.6109563602599815,
      "grad_norm": 1.5002972679400732,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8740043640136719,
      "num_tokens": 301727949.0,
      "step": 8675
    },
    {
      "epoch": 1.611142061281337,
      "grad_norm": 1.6786991265802804,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8749703168869019,
      "num_tokens": 301759339.0,
      "step": 8676
    },
    {
      "epoch": 1.6113277623026927,
      "grad_norm": 1.405491502460055,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8811193108558655,
      "num_tokens": 301798152.0,
      "step": 8677
    },
    {
      "epoch": 1.6115134633240484,
      "grad_norm": 1.4732337805173674,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.874121904373169,
      "num_tokens": 301832786.0,
      "step": 8678
    },
    {
      "epoch": 1.611699164345404,
      "grad_norm": 1.4111813804377797,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8877185583114624,
      "num_tokens": 301873839.0,
      "step": 8679
    },
    {
      "epoch": 1.6118848653667595,
      "grad_norm": 1.6435323911708606,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8679229617118835,
      "num_tokens": 301903856.0,
      "step": 8680
    },
    {
      "epoch": 1.6120705663881152,
      "grad_norm": 1.6283493704768297,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8746228218078613,
      "num_tokens": 301936464.0,
      "step": 8681
    },
    {
      "epoch": 1.6122562674094707,
      "grad_norm": 1.3590655204641864,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8830298781394958,
      "num_tokens": 301975267.0,
      "step": 8682
    },
    {
      "epoch": 1.6124419684308262,
      "grad_norm": 1.5517132729241974,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8753889203071594,
      "num_tokens": 302008000.0,
      "step": 8683
    },
    {
      "epoch": 1.612627669452182,
      "grad_norm": 1.5286414495948124,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8733819723129272,
      "num_tokens": 302049277.0,
      "step": 8684
    },
    {
      "epoch": 1.6128133704735377,
      "grad_norm": 1.564015624040583,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8617169857025146,
      "num_tokens": 302085451.0,
      "step": 8685
    },
    {
      "epoch": 1.6129990714948932,
      "grad_norm": 1.5925005825025436,
      "learning_rate": 1e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8638848066329956,
      "num_tokens": 302120598.0,
      "step": 8686
    },
    {
      "epoch": 1.6131847725162487,
      "grad_norm": 1.405093660298623,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8857855796813965,
      "num_tokens": 302157953.0,
      "step": 8687
    },
    {
      "epoch": 1.6133704735376044,
      "grad_norm": 1.5882100462079456,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8794718980789185,
      "num_tokens": 302191123.0,
      "step": 8688
    },
    {
      "epoch": 1.6135561745589602,
      "grad_norm": 1.6444959791463372,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8774759769439697,
      "num_tokens": 302219799.0,
      "step": 8689
    },
    {
      "epoch": 1.6137418755803157,
      "grad_norm": 1.521432236679453,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8652248382568359,
      "num_tokens": 302255629.0,
      "step": 8690
    },
    {
      "epoch": 1.6139275766016712,
      "grad_norm": 1.5061070398588288,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8692050576210022,
      "num_tokens": 302290911.0,
      "step": 8691
    },
    {
      "epoch": 1.614113277623027,
      "grad_norm": 1.3812778606885843,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8843685388565063,
      "num_tokens": 302327497.0,
      "step": 8692
    },
    {
      "epoch": 1.6142989786443827,
      "grad_norm": 1.4482061009451401,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8695385456085205,
      "num_tokens": 302363528.0,
      "step": 8693
    },
    {
      "epoch": 1.6144846796657382,
      "grad_norm": 1.5655851991369283,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8816257119178772,
      "num_tokens": 302394971.0,
      "step": 8694
    },
    {
      "epoch": 1.6146703806870937,
      "grad_norm": 1.6531887871640438,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8753111362457275,
      "num_tokens": 302423931.0,
      "step": 8695
    },
    {
      "epoch": 1.6148560817084494,
      "grad_norm": 1.512043060382905,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8679007291793823,
      "num_tokens": 302459521.0,
      "step": 8696
    },
    {
      "epoch": 1.6150417827298051,
      "grad_norm": 1.5600273701913314,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.880959689617157,
      "num_tokens": 302490086.0,
      "step": 8697
    },
    {
      "epoch": 1.6152274837511607,
      "grad_norm": 1.403495986537715,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8763549327850342,
      "num_tokens": 302530389.0,
      "step": 8698
    },
    {
      "epoch": 1.6154131847725162,
      "grad_norm": 1.3936686112262036,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8873673677444458,
      "num_tokens": 302567046.0,
      "step": 8699
    },
    {
      "epoch": 1.615598885793872,
      "grad_norm": 1.4121530633570554,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.883301854133606,
      "num_tokens": 302603653.0,
      "step": 8700
    },
    {
      "epoch": 1.6157845868152276,
      "grad_norm": 1.415057170227604,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8782338500022888,
      "num_tokens": 302642365.0,
      "step": 8701
    },
    {
      "epoch": 1.6159702878365831,
      "grad_norm": 1.3592077034443348,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8908765912055969,
      "num_tokens": 302678646.0,
      "step": 8702
    },
    {
      "epoch": 1.6161559888579387,
      "grad_norm": 1.5616229697989503,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8754196166992188,
      "num_tokens": 302713300.0,
      "step": 8703
    },
    {
      "epoch": 1.6163416898792944,
      "grad_norm": 1.4888463270101633,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8933157920837402,
      "num_tokens": 302746115.0,
      "step": 8704
    },
    {
      "epoch": 1.6165273909006501,
      "grad_norm": 1.5287604634735483,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8800680637359619,
      "num_tokens": 302777510.0,
      "step": 8705
    },
    {
      "epoch": 1.6167130919220054,
      "grad_norm": 1.3082919586349775,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8819047212600708,
      "num_tokens": 302820066.0,
      "step": 8706
    },
    {
      "epoch": 1.6168987929433611,
      "grad_norm": 1.568154577258664,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8834981322288513,
      "num_tokens": 302851719.0,
      "step": 8707
    },
    {
      "epoch": 1.6170844939647169,
      "grad_norm": 1.645221468465393,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8722646236419678,
      "num_tokens": 302882150.0,
      "step": 8708
    },
    {
      "epoch": 1.6172701949860724,
      "grad_norm": 1.4070900960504245,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8814061880111694,
      "num_tokens": 302919635.0,
      "step": 8709
    },
    {
      "epoch": 1.617455896007428,
      "grad_norm": 1.4222799152649799,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8871045708656311,
      "num_tokens": 302957624.0,
      "step": 8710
    },
    {
      "epoch": 1.6176415970287836,
      "grad_norm": 1.6369341096738126,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8796159625053406,
      "num_tokens": 302988734.0,
      "step": 8711
    },
    {
      "epoch": 1.6178272980501394,
      "grad_norm": 1.4702123039071908,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8724517226219177,
      "num_tokens": 303027482.0,
      "step": 8712
    },
    {
      "epoch": 1.6180129990714949,
      "grad_norm": 1.389457353876049,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.869563102722168,
      "num_tokens": 303067788.0,
      "step": 8713
    },
    {
      "epoch": 1.6181987000928504,
      "grad_norm": 1.6554854424680032,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8687396049499512,
      "num_tokens": 303098270.0,
      "step": 8714
    },
    {
      "epoch": 1.6183844011142061,
      "grad_norm": 1.5781790237971818,
      "learning_rate": 1e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8689544796943665,
      "num_tokens": 303130676.0,
      "step": 8715
    },
    {
      "epoch": 1.6185701021355619,
      "grad_norm": 1.611075790155722,
      "learning_rate": 1e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.8698317408561707,
      "num_tokens": 303164294.0,
      "step": 8716
    },
    {
      "epoch": 1.6187558031569174,
      "grad_norm": 1.6951860752137742,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8653168678283691,
      "num_tokens": 303200073.0,
      "step": 8717
    },
    {
      "epoch": 1.6189415041782729,
      "grad_norm": 1.5535732567843756,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8768510222434998,
      "num_tokens": 303236350.0,
      "step": 8718
    },
    {
      "epoch": 1.6191272051996286,
      "grad_norm": 1.4687772749128927,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.881574809551239,
      "num_tokens": 303273986.0,
      "step": 8719
    },
    {
      "epoch": 1.6193129062209843,
      "grad_norm": 1.4405274415060914,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8750502467155457,
      "num_tokens": 303310374.0,
      "step": 8720
    },
    {
      "epoch": 1.6194986072423398,
      "grad_norm": 1.7265673725793111,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8769848346710205,
      "num_tokens": 303339729.0,
      "step": 8721
    },
    {
      "epoch": 1.6196843082636954,
      "grad_norm": 1.531895624234184,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.879514217376709,
      "num_tokens": 303377548.0,
      "step": 8722
    },
    {
      "epoch": 1.619870009285051,
      "grad_norm": 1.4754250961392452,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8778367042541504,
      "num_tokens": 303414659.0,
      "step": 8723
    },
    {
      "epoch": 1.6200557103064068,
      "grad_norm": 1.4625141179513343,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8799053430557251,
      "num_tokens": 303448250.0,
      "step": 8724
    },
    {
      "epoch": 1.6202414113277623,
      "grad_norm": 1.4037817127168435,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8777713775634766,
      "num_tokens": 303487607.0,
      "step": 8725
    },
    {
      "epoch": 1.6204271123491178,
      "grad_norm": 1.3873136960482375,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8732237815856934,
      "num_tokens": 303526863.0,
      "step": 8726
    },
    {
      "epoch": 1.6206128133704736,
      "grad_norm": 1.4959107393915154,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8808671236038208,
      "num_tokens": 303559173.0,
      "step": 8727
    },
    {
      "epoch": 1.6207985143918293,
      "grad_norm": 1.3703903129313024,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8650453090667725,
      "num_tokens": 303600304.0,
      "step": 8728
    },
    {
      "epoch": 1.6209842154131848,
      "grad_norm": 1.6040371350991656,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8782840967178345,
      "num_tokens": 303629930.0,
      "step": 8729
    },
    {
      "epoch": 1.6211699164345403,
      "grad_norm": 1.4964005458126055,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8767337799072266,
      "num_tokens": 303666342.0,
      "step": 8730
    },
    {
      "epoch": 1.621355617455896,
      "grad_norm": 1.3903299856445497,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8797159194946289,
      "num_tokens": 303706736.0,
      "step": 8731
    },
    {
      "epoch": 1.6215413184772516,
      "grad_norm": 1.6921377671960978,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8726211786270142,
      "num_tokens": 303735657.0,
      "step": 8732
    },
    {
      "epoch": 1.621727019498607,
      "grad_norm": 1.4224735538791506,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8826100826263428,
      "num_tokens": 303773209.0,
      "step": 8733
    },
    {
      "epoch": 1.6219127205199628,
      "grad_norm": 1.464219960128325,
      "learning_rate": 1e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.847977876663208,
      "num_tokens": 303813961.0,
      "step": 8734
    },
    {
      "epoch": 1.6220984215413186,
      "grad_norm": 1.510804238110556,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8759046792984009,
      "num_tokens": 303847203.0,
      "step": 8735
    },
    {
      "epoch": 1.622284122562674,
      "grad_norm": 1.6836032687966866,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8776038289070129,
      "num_tokens": 303875571.0,
      "step": 8736
    },
    {
      "epoch": 1.6224698235840296,
      "grad_norm": 1.4634861812930857,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.883371114730835,
      "num_tokens": 303917576.0,
      "step": 8737
    },
    {
      "epoch": 1.6226555246053853,
      "grad_norm": 1.5039066364355786,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.869945764541626,
      "num_tokens": 303954462.0,
      "step": 8738
    },
    {
      "epoch": 1.622841225626741,
      "grad_norm": 1.5898366345621049,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8760055899620056,
      "num_tokens": 303985864.0,
      "step": 8739
    },
    {
      "epoch": 1.6230269266480966,
      "grad_norm": 1.7394020955761547,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.866470456123352,
      "num_tokens": 304016935.0,
      "step": 8740
    },
    {
      "epoch": 1.623212627669452,
      "grad_norm": 1.4906870826332053,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.873603343963623,
      "num_tokens": 304052272.0,
      "step": 8741
    },
    {
      "epoch": 1.6233983286908078,
      "grad_norm": 1.532521673766072,
      "learning_rate": 1e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8591312170028687,
      "num_tokens": 304087278.0,
      "step": 8742
    },
    {
      "epoch": 1.6235840297121635,
      "grad_norm": 1.551453470061926,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8745604753494263,
      "num_tokens": 304123894.0,
      "step": 8743
    },
    {
      "epoch": 1.623769730733519,
      "grad_norm": 1.5177948016071876,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8771283626556396,
      "num_tokens": 304155115.0,
      "step": 8744
    },
    {
      "epoch": 1.6239554317548746,
      "grad_norm": 1.4406868752487094,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8791946768760681,
      "num_tokens": 304191028.0,
      "step": 8745
    },
    {
      "epoch": 1.6241411327762303,
      "grad_norm": 1.377565976978594,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8722472786903381,
      "num_tokens": 304230529.0,
      "step": 8746
    },
    {
      "epoch": 1.624326833797586,
      "grad_norm": 1.6682050373659876,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8773825764656067,
      "num_tokens": 304254646.0,
      "step": 8747
    },
    {
      "epoch": 1.6245125348189415,
      "grad_norm": 1.5516976944545493,
      "learning_rate": 1e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8659529685974121,
      "num_tokens": 304289660.0,
      "step": 8748
    },
    {
      "epoch": 1.624698235840297,
      "grad_norm": 1.4735431513230444,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8841961622238159,
      "num_tokens": 304324705.0,
      "step": 8749
    },
    {
      "epoch": 1.6248839368616528,
      "grad_norm": 1.521503857946085,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8753204345703125,
      "num_tokens": 304357121.0,
      "step": 8750
    },
    {
      "epoch": 1.6250696378830085,
      "grad_norm": 1.6352202565527358,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8837569355964661,
      "num_tokens": 304386234.0,
      "step": 8751
    },
    {
      "epoch": 1.625255338904364,
      "grad_norm": 1.4902247735871776,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8745694160461426,
      "num_tokens": 304422519.0,
      "step": 8752
    },
    {
      "epoch": 1.6254410399257195,
      "grad_norm": 1.4629257949751866,
      "learning_rate": 1e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8663623332977295,
      "num_tokens": 304461996.0,
      "step": 8753
    },
    {
      "epoch": 1.6256267409470753,
      "grad_norm": 1.5569071914016543,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8698890209197998,
      "num_tokens": 304496175.0,
      "step": 8754
    },
    {
      "epoch": 1.6258124419684308,
      "grad_norm": 1.5818548670280124,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.875687837600708,
      "num_tokens": 304529254.0,
      "step": 8755
    },
    {
      "epoch": 1.6259981429897863,
      "grad_norm": 1.3932149113949908,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8892014026641846,
      "num_tokens": 304567781.0,
      "step": 8756
    },
    {
      "epoch": 1.626183844011142,
      "grad_norm": 1.4644889924816218,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8828155994415283,
      "num_tokens": 304604174.0,
      "step": 8757
    },
    {
      "epoch": 1.6263695450324978,
      "grad_norm": 1.610414415700503,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8719377517700195,
      "num_tokens": 304638904.0,
      "step": 8758
    },
    {
      "epoch": 1.6265552460538533,
      "grad_norm": 1.5950560902876365,
      "learning_rate": 1e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.8706018924713135,
      "num_tokens": 304670713.0,
      "step": 8759
    },
    {
      "epoch": 1.6267409470752088,
      "grad_norm": 1.4579772603862375,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8805409669876099,
      "num_tokens": 304703884.0,
      "step": 8760
    },
    {
      "epoch": 1.6269266480965645,
      "grad_norm": 1.399896534026845,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8804748058319092,
      "num_tokens": 304741837.0,
      "step": 8761
    },
    {
      "epoch": 1.6271123491179202,
      "grad_norm": 1.602054774701516,
      "learning_rate": 1e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8664371967315674,
      "num_tokens": 304776340.0,
      "step": 8762
    },
    {
      "epoch": 1.6272980501392758,
      "grad_norm": 1.4775537238039824,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8796600699424744,
      "num_tokens": 304812175.0,
      "step": 8763
    },
    {
      "epoch": 1.6274837511606313,
      "grad_norm": 1.6284239800997524,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8745282888412476,
      "num_tokens": 304845900.0,
      "step": 8764
    },
    {
      "epoch": 1.627669452181987,
      "grad_norm": 1.6062083047889641,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8806370496749878,
      "num_tokens": 304873946.0,
      "step": 8765
    },
    {
      "epoch": 1.6278551532033427,
      "grad_norm": 1.4012542383110922,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8834882378578186,
      "num_tokens": 304909192.0,
      "step": 8766
    },
    {
      "epoch": 1.6280408542246982,
      "grad_norm": 1.362734686914435,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8910078406333923,
      "num_tokens": 304944527.0,
      "step": 8767
    },
    {
      "epoch": 1.6282265552460538,
      "grad_norm": 1.547500979771872,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8757567405700684,
      "num_tokens": 304976410.0,
      "step": 8768
    },
    {
      "epoch": 1.6284122562674095,
      "grad_norm": 1.4834290203256375,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8714749813079834,
      "num_tokens": 305011495.0,
      "step": 8769
    },
    {
      "epoch": 1.6285979572887652,
      "grad_norm": 1.4760585483054325,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8750534057617188,
      "num_tokens": 305049064.0,
      "step": 8770
    },
    {
      "epoch": 1.6287836583101207,
      "grad_norm": 1.479404423738712,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8785216212272644,
      "num_tokens": 305085302.0,
      "step": 8771
    },
    {
      "epoch": 1.6289693593314762,
      "grad_norm": 1.449937393364452,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8914671540260315,
      "num_tokens": 305115236.0,
      "step": 8772
    },
    {
      "epoch": 1.629155060352832,
      "grad_norm": 1.5609147194566761,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8752672672271729,
      "num_tokens": 305147646.0,
      "step": 8773
    },
    {
      "epoch": 1.6293407613741877,
      "grad_norm": 1.558195770692804,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8746821880340576,
      "num_tokens": 305183097.0,
      "step": 8774
    },
    {
      "epoch": 1.6295264623955432,
      "grad_norm": 1.4665310471327913,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8760616779327393,
      "num_tokens": 305219683.0,
      "step": 8775
    },
    {
      "epoch": 1.6297121634168987,
      "grad_norm": 1.4220735755415639,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8780069351196289,
      "num_tokens": 305258246.0,
      "step": 8776
    },
    {
      "epoch": 1.6298978644382545,
      "grad_norm": 1.6397400910289577,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8814645409584045,
      "num_tokens": 305286427.0,
      "step": 8777
    },
    {
      "epoch": 1.63008356545961,
      "grad_norm": 1.583175322489565,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8691001534461975,
      "num_tokens": 305318330.0,
      "step": 8778
    },
    {
      "epoch": 1.6302692664809655,
      "grad_norm": 1.422494463673698,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8820582628250122,
      "num_tokens": 305356897.0,
      "step": 8779
    },
    {
      "epoch": 1.6304549675023212,
      "grad_norm": 1.611796178667315,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8797715306282043,
      "num_tokens": 305388350.0,
      "step": 8780
    },
    {
      "epoch": 1.630640668523677,
      "grad_norm": 1.4814555473781719,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8831009864807129,
      "num_tokens": 305423374.0,
      "step": 8781
    },
    {
      "epoch": 1.6308263695450325,
      "grad_norm": 1.421653853554701,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8620602488517761,
      "num_tokens": 305463632.0,
      "step": 8782
    },
    {
      "epoch": 1.631012070566388,
      "grad_norm": 1.604722244538316,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.88853919506073,
      "num_tokens": 305493205.0,
      "step": 8783
    },
    {
      "epoch": 1.6311977715877437,
      "grad_norm": 1.4447067897893162,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8806803226470947,
      "num_tokens": 305530261.0,
      "step": 8784
    },
    {
      "epoch": 1.6313834726090994,
      "grad_norm": 1.4978029551462586,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8615771532058716,
      "num_tokens": 305566400.0,
      "step": 8785
    },
    {
      "epoch": 1.631569173630455,
      "grad_norm": 1.4880140579681758,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8767949342727661,
      "num_tokens": 305600411.0,
      "step": 8786
    },
    {
      "epoch": 1.6317548746518105,
      "grad_norm": 1.4566570397894454,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8874295949935913,
      "num_tokens": 305634458.0,
      "step": 8787
    },
    {
      "epoch": 1.6319405756731662,
      "grad_norm": 1.515070044849912,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8842350244522095,
      "num_tokens": 305668383.0,
      "step": 8788
    },
    {
      "epoch": 1.632126276694522,
      "grad_norm": 1.5952772116565201,
      "learning_rate": 1e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8678301572799683,
      "num_tokens": 305699803.0,
      "step": 8789
    },
    {
      "epoch": 1.6323119777158774,
      "grad_norm": 1.659707388559891,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8760956525802612,
      "num_tokens": 305726547.0,
      "step": 8790
    },
    {
      "epoch": 1.632497678737233,
      "grad_norm": 1.5040053838289957,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8875330090522766,
      "num_tokens": 305755912.0,
      "step": 8791
    },
    {
      "epoch": 1.6326833797585887,
      "grad_norm": 1.4396184833711658,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8776799440383911,
      "num_tokens": 305791081.0,
      "step": 8792
    },
    {
      "epoch": 1.6328690807799444,
      "grad_norm": 1.463177650739571,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8779340982437134,
      "num_tokens": 305827195.0,
      "step": 8793
    },
    {
      "epoch": 1.6330547818013,
      "grad_norm": 1.3949616622364749,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8854386806488037,
      "num_tokens": 305862644.0,
      "step": 8794
    },
    {
      "epoch": 1.6332404828226554,
      "grad_norm": 1.5202734173659362,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8718377351760864,
      "num_tokens": 305900356.0,
      "step": 8795
    },
    {
      "epoch": 1.6334261838440112,
      "grad_norm": 1.5025686183022304,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8784361481666565,
      "num_tokens": 305936309.0,
      "step": 8796
    },
    {
      "epoch": 1.633611884865367,
      "grad_norm": 1.3897295191106342,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8966543674468994,
      "num_tokens": 305969573.0,
      "step": 8797
    },
    {
      "epoch": 1.6337975858867224,
      "grad_norm": 1.5607118098613144,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8788034319877625,
      "num_tokens": 306001125.0,
      "step": 8798
    },
    {
      "epoch": 1.633983286908078,
      "grad_norm": 1.411173390098597,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8856967091560364,
      "num_tokens": 306040675.0,
      "step": 8799
    },
    {
      "epoch": 1.6341689879294337,
      "grad_norm": 1.3661425917521322,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8720123171806335,
      "num_tokens": 306083272.0,
      "step": 8800
    },
    {
      "epoch": 1.6343546889507894,
      "grad_norm": 1.5233340359318999,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8716599941253662,
      "num_tokens": 306118050.0,
      "step": 8801
    },
    {
      "epoch": 1.6345403899721447,
      "grad_norm": 1.5336399447546298,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8874423503875732,
      "num_tokens": 306149304.0,
      "step": 8802
    },
    {
      "epoch": 1.6347260909935004,
      "grad_norm": 1.5587364536027506,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8663743138313293,
      "num_tokens": 306183709.0,
      "step": 8803
    },
    {
      "epoch": 1.6349117920148561,
      "grad_norm": 1.462040879728883,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8790960311889648,
      "num_tokens": 306220741.0,
      "step": 8804
    },
    {
      "epoch": 1.6350974930362117,
      "grad_norm": 1.607122632152587,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8817150592803955,
      "num_tokens": 306251296.0,
      "step": 8805
    },
    {
      "epoch": 1.6352831940575672,
      "grad_norm": 1.408738465573957,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8787276744842529,
      "num_tokens": 306288204.0,
      "step": 8806
    },
    {
      "epoch": 1.635468895078923,
      "grad_norm": 1.5619992592167915,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8800385594367981,
      "num_tokens": 306319954.0,
      "step": 8807
    },
    {
      "epoch": 1.6356545961002786,
      "grad_norm": 1.380793716054217,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8856508731842041,
      "num_tokens": 306357847.0,
      "step": 8808
    },
    {
      "epoch": 1.6358402971216341,
      "grad_norm": 1.5655327336254725,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8877475261688232,
      "num_tokens": 306389613.0,
      "step": 8809
    },
    {
      "epoch": 1.6360259981429897,
      "grad_norm": 1.485729292109,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8780380487442017,
      "num_tokens": 306426560.0,
      "step": 8810
    },
    {
      "epoch": 1.6362116991643454,
      "grad_norm": 1.420528204189153,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8706541657447815,
      "num_tokens": 306465098.0,
      "step": 8811
    },
    {
      "epoch": 1.6363974001857011,
      "grad_norm": 1.59754830020217,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.865658700466156,
      "num_tokens": 306502122.0,
      "step": 8812
    },
    {
      "epoch": 1.6365831012070566,
      "grad_norm": 1.5391217425529522,
      "learning_rate": 1e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8694158792495728,
      "num_tokens": 306537202.0,
      "step": 8813
    },
    {
      "epoch": 1.6367688022284121,
      "grad_norm": 1.656846391861024,
      "learning_rate": 1e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8671607971191406,
      "num_tokens": 306569060.0,
      "step": 8814
    },
    {
      "epoch": 1.6369545032497679,
      "grad_norm": 1.5685519482496495,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8874919414520264,
      "num_tokens": 306598107.0,
      "step": 8815
    },
    {
      "epoch": 1.6371402042711236,
      "grad_norm": 1.5080177257284724,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.880783200263977,
      "num_tokens": 306630902.0,
      "step": 8816
    },
    {
      "epoch": 1.6373259052924791,
      "grad_norm": 1.4592702781303541,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8876080513000488,
      "num_tokens": 306662643.0,
      "step": 8817
    },
    {
      "epoch": 1.6375116063138346,
      "grad_norm": 1.4022972447490263,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8786983489990234,
      "num_tokens": 306699572.0,
      "step": 8818
    },
    {
      "epoch": 1.6376973073351904,
      "grad_norm": 1.4745456737875127,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8922173380851746,
      "num_tokens": 306732028.0,
      "step": 8819
    },
    {
      "epoch": 1.637883008356546,
      "grad_norm": 1.351247197453156,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8787336349487305,
      "num_tokens": 306775420.0,
      "step": 8820
    },
    {
      "epoch": 1.6380687093779016,
      "grad_norm": 1.6053125926591907,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8769306540489197,
      "num_tokens": 306804331.0,
      "step": 8821
    },
    {
      "epoch": 1.6382544103992571,
      "grad_norm": 1.442647216767296,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8882827162742615,
      "num_tokens": 306839890.0,
      "step": 8822
    },
    {
      "epoch": 1.6384401114206129,
      "grad_norm": 1.4191737649167273,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8837668299674988,
      "num_tokens": 306876656.0,
      "step": 8823
    },
    {
      "epoch": 1.6386258124419686,
      "grad_norm": 1.5117842129329404,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8770468831062317,
      "num_tokens": 306912151.0,
      "step": 8824
    },
    {
      "epoch": 1.638811513463324,
      "grad_norm": 1.4941579610264113,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8757843971252441,
      "num_tokens": 306945878.0,
      "step": 8825
    },
    {
      "epoch": 1.6389972144846796,
      "grad_norm": 1.5567748746831185,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8746240139007568,
      "num_tokens": 306982710.0,
      "step": 8826
    },
    {
      "epoch": 1.6391829155060353,
      "grad_norm": 1.4733867644851075,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8839465379714966,
      "num_tokens": 307015714.0,
      "step": 8827
    },
    {
      "epoch": 1.6393686165273909,
      "grad_norm": 1.2844645061293345,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.887247622013092,
      "num_tokens": 307060197.0,
      "step": 8828
    },
    {
      "epoch": 1.6395543175487464,
      "grad_norm": 1.3112267175968193,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8754113912582397,
      "num_tokens": 307102952.0,
      "step": 8829
    },
    {
      "epoch": 1.639740018570102,
      "grad_norm": 1.322615410820184,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8884193897247314,
      "num_tokens": 307144501.0,
      "step": 8830
    },
    {
      "epoch": 1.6399257195914578,
      "grad_norm": 1.4643181994141843,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8919781446456909,
      "num_tokens": 307177230.0,
      "step": 8831
    },
    {
      "epoch": 1.6401114206128133,
      "grad_norm": 1.4287971071563308,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8693432807922363,
      "num_tokens": 307218897.0,
      "step": 8832
    },
    {
      "epoch": 1.6402971216341689,
      "grad_norm": 1.4114068778266708,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8887033462524414,
      "num_tokens": 307256661.0,
      "step": 8833
    },
    {
      "epoch": 1.6404828226555246,
      "grad_norm": 1.413509878363364,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8795655965805054,
      "num_tokens": 307296008.0,
      "step": 8834
    },
    {
      "epoch": 1.6406685236768803,
      "grad_norm": 1.3795846183797917,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.890140175819397,
      "num_tokens": 307333209.0,
      "step": 8835
    },
    {
      "epoch": 1.6408542246982358,
      "grad_norm": 1.5907144656950492,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8768000602722168,
      "num_tokens": 307365152.0,
      "step": 8836
    },
    {
      "epoch": 1.6410399257195913,
      "grad_norm": 1.3514754256403445,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8930840492248535,
      "num_tokens": 307401762.0,
      "step": 8837
    },
    {
      "epoch": 1.641225626740947,
      "grad_norm": 1.5625432365543417,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8766415119171143,
      "num_tokens": 307432692.0,
      "step": 8838
    },
    {
      "epoch": 1.6414113277623028,
      "grad_norm": 1.3646638380330685,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8769011497497559,
      "num_tokens": 307474389.0,
      "step": 8839
    },
    {
      "epoch": 1.6415970287836583,
      "grad_norm": 1.665261699949827,
      "learning_rate": 1e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8650142550468445,
      "num_tokens": 307506979.0,
      "step": 8840
    },
    {
      "epoch": 1.6417827298050138,
      "grad_norm": 1.3032548978263045,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8889073729515076,
      "num_tokens": 307547260.0,
      "step": 8841
    },
    {
      "epoch": 1.6419684308263696,
      "grad_norm": 1.4761584339332858,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8822470307350159,
      "num_tokens": 307578892.0,
      "step": 8842
    },
    {
      "epoch": 1.6421541318477253,
      "grad_norm": 1.5315141941163908,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8722119331359863,
      "num_tokens": 307615057.0,
      "step": 8843
    },
    {
      "epoch": 1.6423398328690808,
      "grad_norm": 1.468535759197894,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8762943744659424,
      "num_tokens": 307648651.0,
      "step": 8844
    },
    {
      "epoch": 1.6425255338904363,
      "grad_norm": 1.3023561432886186,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8885621428489685,
      "num_tokens": 307690246.0,
      "step": 8845
    },
    {
      "epoch": 1.642711234911792,
      "grad_norm": 1.4985889313697138,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8657403588294983,
      "num_tokens": 307729867.0,
      "step": 8846
    },
    {
      "epoch": 1.6428969359331478,
      "grad_norm": 1.6329949647558475,
      "learning_rate": 1e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.8675332069396973,
      "num_tokens": 307761380.0,
      "step": 8847
    },
    {
      "epoch": 1.6430826369545033,
      "grad_norm": 1.4776290882267247,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8824465870857239,
      "num_tokens": 307798228.0,
      "step": 8848
    },
    {
      "epoch": 1.6432683379758588,
      "grad_norm": 1.543260826701291,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8734400868415833,
      "num_tokens": 307831821.0,
      "step": 8849
    },
    {
      "epoch": 1.6434540389972145,
      "grad_norm": 1.4665285111958724,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8793465495109558,
      "num_tokens": 307868091.0,
      "step": 8850
    },
    {
      "epoch": 1.64363974001857,
      "grad_norm": 1.5172643795119245,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8725706934928894,
      "num_tokens": 307902280.0,
      "step": 8851
    },
    {
      "epoch": 1.6438254410399256,
      "grad_norm": 1.59191690791718,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8744776248931885,
      "num_tokens": 307932308.0,
      "step": 8852
    },
    {
      "epoch": 1.6440111420612813,
      "grad_norm": 1.3489128052173789,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8896878361701965,
      "num_tokens": 307971401.0,
      "step": 8853
    },
    {
      "epoch": 1.644196843082637,
      "grad_norm": 1.5073644127651442,
      "learning_rate": 1e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8685684204101562,
      "num_tokens": 308007981.0,
      "step": 8854
    },
    {
      "epoch": 1.6443825441039925,
      "grad_norm": 1.6473888486525243,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8523460030555725,
      "num_tokens": 308035128.0,
      "step": 8855
    },
    {
      "epoch": 1.644568245125348,
      "grad_norm": 1.6134507617445988,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8740003108978271,
      "num_tokens": 308065694.0,
      "step": 8856
    },
    {
      "epoch": 1.6447539461467038,
      "grad_norm": 1.4012598864964634,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8794229030609131,
      "num_tokens": 308102846.0,
      "step": 8857
    },
    {
      "epoch": 1.6449396471680595,
      "grad_norm": 1.507974741469835,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8707268238067627,
      "num_tokens": 308137217.0,
      "step": 8858
    },
    {
      "epoch": 1.645125348189415,
      "grad_norm": 1.531539996152344,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8724504709243774,
      "num_tokens": 308172149.0,
      "step": 8859
    },
    {
      "epoch": 1.6453110492107705,
      "grad_norm": 1.5150440363427078,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8730588555335999,
      "num_tokens": 308210305.0,
      "step": 8860
    },
    {
      "epoch": 1.6454967502321263,
      "grad_norm": 1.4668069661906764,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.883619487285614,
      "num_tokens": 308245104.0,
      "step": 8861
    },
    {
      "epoch": 1.645682451253482,
      "grad_norm": 1.4902490281399607,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.8728062510490417,
      "num_tokens": 308278743.0,
      "step": 8862
    },
    {
      "epoch": 1.6458681522748375,
      "grad_norm": 1.398189679207567,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8819649815559387,
      "num_tokens": 308312874.0,
      "step": 8863
    },
    {
      "epoch": 1.646053853296193,
      "grad_norm": 1.5626598573452664,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8831011056900024,
      "num_tokens": 308345136.0,
      "step": 8864
    },
    {
      "epoch": 1.6462395543175488,
      "grad_norm": 1.5782460429923273,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8796231150627136,
      "num_tokens": 308376102.0,
      "step": 8865
    },
    {
      "epoch": 1.6464252553389045,
      "grad_norm": 1.6225244965321812,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8663557171821594,
      "num_tokens": 308407635.0,
      "step": 8866
    },
    {
      "epoch": 1.64661095636026,
      "grad_norm": 1.4341331345979293,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8785200119018555,
      "num_tokens": 308447353.0,
      "step": 8867
    },
    {
      "epoch": 1.6467966573816155,
      "grad_norm": 1.4598229146400652,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8840181231498718,
      "num_tokens": 308484332.0,
      "step": 8868
    },
    {
      "epoch": 1.6469823584029712,
      "grad_norm": 1.6541263265016408,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.8688287734985352,
      "num_tokens": 308514473.0,
      "step": 8869
    },
    {
      "epoch": 1.647168059424327,
      "grad_norm": 1.4214797465179512,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8823121786117554,
      "num_tokens": 308553403.0,
      "step": 8870
    },
    {
      "epoch": 1.6473537604456825,
      "grad_norm": 1.598625923686696,
      "learning_rate": 1e-06,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8585115671157837,
      "num_tokens": 308586344.0,
      "step": 8871
    },
    {
      "epoch": 1.647539461467038,
      "grad_norm": 1.4676373453042753,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8788236379623413,
      "num_tokens": 308623827.0,
      "step": 8872
    },
    {
      "epoch": 1.6477251624883937,
      "grad_norm": 1.5309235153191194,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8815544843673706,
      "num_tokens": 308654189.0,
      "step": 8873
    },
    {
      "epoch": 1.6479108635097495,
      "grad_norm": 1.485813599457797,
      "learning_rate": 1e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8685410022735596,
      "num_tokens": 308688023.0,
      "step": 8874
    },
    {
      "epoch": 1.6480965645311048,
      "grad_norm": 1.5454130788676543,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8705543875694275,
      "num_tokens": 308717599.0,
      "step": 8875
    },
    {
      "epoch": 1.6482822655524605,
      "grad_norm": 1.4769129135123167,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.8853212594985962,
      "num_tokens": 308754229.0,
      "step": 8876
    },
    {
      "epoch": 1.6484679665738162,
      "grad_norm": 1.4815373553973883,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8824113607406616,
      "num_tokens": 308789997.0,
      "step": 8877
    },
    {
      "epoch": 1.6486536675951717,
      "grad_norm": 1.5182188204189717,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8659359216690063,
      "num_tokens": 308826664.0,
      "step": 8878
    },
    {
      "epoch": 1.6488393686165272,
      "grad_norm": 1.5270305559284088,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8685194849967957,
      "num_tokens": 308859113.0,
      "step": 8879
    },
    {
      "epoch": 1.649025069637883,
      "grad_norm": 1.4764661448859207,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8887999653816223,
      "num_tokens": 308890459.0,
      "step": 8880
    },
    {
      "epoch": 1.6492107706592387,
      "grad_norm": 1.4817102236843664,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8749646544456482,
      "num_tokens": 308925794.0,
      "step": 8881
    },
    {
      "epoch": 1.6493964716805942,
      "grad_norm": 1.4244055459724816,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8839328289031982,
      "num_tokens": 308961414.0,
      "step": 8882
    },
    {
      "epoch": 1.6495821727019497,
      "grad_norm": 1.5929711086109046,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8777288198471069,
      "num_tokens": 308990881.0,
      "step": 8883
    },
    {
      "epoch": 1.6497678737233055,
      "grad_norm": 1.4384310047177582,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8852533102035522,
      "num_tokens": 309026457.0,
      "step": 8884
    },
    {
      "epoch": 1.6499535747446612,
      "grad_norm": 1.4552331116177217,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8783327341079712,
      "num_tokens": 309061399.0,
      "step": 8885
    },
    {
      "epoch": 1.6501392757660167,
      "grad_norm": 1.456172165009202,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8699765205383301,
      "num_tokens": 309101541.0,
      "step": 8886
    },
    {
      "epoch": 1.6503249767873722,
      "grad_norm": 1.615900954550799,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.875759482383728,
      "num_tokens": 309132548.0,
      "step": 8887
    },
    {
      "epoch": 1.650510677808728,
      "grad_norm": 1.4943510069802135,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8877487182617188,
      "num_tokens": 309165927.0,
      "step": 8888
    },
    {
      "epoch": 1.6506963788300837,
      "grad_norm": 1.4331333200964396,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8759523630142212,
      "num_tokens": 309207923.0,
      "step": 8889
    },
    {
      "epoch": 1.6508820798514392,
      "grad_norm": 1.5250150302894314,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8715355396270752,
      "num_tokens": 309242672.0,
      "step": 8890
    },
    {
      "epoch": 1.6510677808727947,
      "grad_norm": 1.555105144746696,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.885621964931488,
      "num_tokens": 309274475.0,
      "step": 8891
    },
    {
      "epoch": 1.6512534818941504,
      "grad_norm": 1.5697465187342547,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8671621084213257,
      "num_tokens": 309308786.0,
      "step": 8892
    },
    {
      "epoch": 1.6514391829155062,
      "grad_norm": 1.5462124258700751,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8815075159072876,
      "num_tokens": 309341713.0,
      "step": 8893
    },
    {
      "epoch": 1.6516248839368617,
      "grad_norm": 1.579597398640653,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8672091960906982,
      "num_tokens": 309376750.0,
      "step": 8894
    },
    {
      "epoch": 1.6518105849582172,
      "grad_norm": 1.4211427801536818,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.894173264503479,
      "num_tokens": 309413761.0,
      "step": 8895
    },
    {
      "epoch": 1.651996285979573,
      "grad_norm": 1.5874958308201181,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.877821683883667,
      "num_tokens": 309445308.0,
      "step": 8896
    },
    {
      "epoch": 1.6521819870009287,
      "grad_norm": 1.3989932311457973,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.887381374835968,
      "num_tokens": 309481212.0,
      "step": 8897
    },
    {
      "epoch": 1.6523676880222842,
      "grad_norm": 1.519666580051062,
      "learning_rate": 1e-06,
      "loss": 0.3933,
      "mean_token_accuracy": 0.8676146864891052,
      "num_tokens": 309518480.0,
      "step": 8898
    },
    {
      "epoch": 1.6525533890436397,
      "grad_norm": 1.4821547027701814,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.8707456588745117,
      "num_tokens": 309555739.0,
      "step": 8899
    },
    {
      "epoch": 1.6527390900649954,
      "grad_norm": 1.547022389262994,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8619939088821411,
      "num_tokens": 309590997.0,
      "step": 8900
    },
    {
      "epoch": 1.652924791086351,
      "grad_norm": 1.6570824417525019,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8659394383430481,
      "num_tokens": 309619772.0,
      "step": 8901
    },
    {
      "epoch": 1.6531104921077064,
      "grad_norm": 1.5092996901717683,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8751753568649292,
      "num_tokens": 309655321.0,
      "step": 8902
    },
    {
      "epoch": 1.6532961931290622,
      "grad_norm": 1.5228384682221192,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8776096105575562,
      "num_tokens": 309691042.0,
      "step": 8903
    },
    {
      "epoch": 1.653481894150418,
      "grad_norm": 1.5449260804070157,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8827527761459351,
      "num_tokens": 309722770.0,
      "step": 8904
    },
    {
      "epoch": 1.6536675951717734,
      "grad_norm": 1.5890932479573576,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8759263157844543,
      "num_tokens": 309755447.0,
      "step": 8905
    },
    {
      "epoch": 1.653853296193129,
      "grad_norm": 1.6103041861946927,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8821185827255249,
      "num_tokens": 309787131.0,
      "step": 8906
    },
    {
      "epoch": 1.6540389972144847,
      "grad_norm": 1.608755426599816,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8774022459983826,
      "num_tokens": 309817447.0,
      "step": 8907
    },
    {
      "epoch": 1.6542246982358404,
      "grad_norm": 1.590730894385848,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8789621591567993,
      "num_tokens": 309847201.0,
      "step": 8908
    },
    {
      "epoch": 1.654410399257196,
      "grad_norm": 1.4306877721928426,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8812447786331177,
      "num_tokens": 309882894.0,
      "step": 8909
    },
    {
      "epoch": 1.6545961002785514,
      "grad_norm": 1.5134228721946423,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8894914388656616,
      "num_tokens": 309915571.0,
      "step": 8910
    },
    {
      "epoch": 1.6547818012999072,
      "grad_norm": 1.5681185074023078,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8784915208816528,
      "num_tokens": 309948365.0,
      "step": 8911
    },
    {
      "epoch": 1.6549675023212629,
      "grad_norm": 1.4054076430765865,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8819240927696228,
      "num_tokens": 309985678.0,
      "step": 8912
    },
    {
      "epoch": 1.6551532033426184,
      "grad_norm": 1.5163351903671296,
      "learning_rate": 1e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8621166944503784,
      "num_tokens": 310023306.0,
      "step": 8913
    },
    {
      "epoch": 1.655338904363974,
      "grad_norm": 1.5486192224635462,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8718371391296387,
      "num_tokens": 310059521.0,
      "step": 8914
    },
    {
      "epoch": 1.6555246053853296,
      "grad_norm": 1.4546750742110244,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8731459379196167,
      "num_tokens": 310096863.0,
      "step": 8915
    },
    {
      "epoch": 1.6557103064066854,
      "grad_norm": 1.4367735053866044,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.871268093585968,
      "num_tokens": 310134912.0,
      "step": 8916
    },
    {
      "epoch": 1.6558960074280409,
      "grad_norm": 1.5111217534988666,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.870521068572998,
      "num_tokens": 310170415.0,
      "step": 8917
    },
    {
      "epoch": 1.6560817084493964,
      "grad_norm": 1.4750867703833654,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8917985558509827,
      "num_tokens": 310201935.0,
      "step": 8918
    },
    {
      "epoch": 1.6562674094707521,
      "grad_norm": 1.519551088042306,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8783621788024902,
      "num_tokens": 310242059.0,
      "step": 8919
    },
    {
      "epoch": 1.6564531104921079,
      "grad_norm": 1.4390763455063853,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8817216753959656,
      "num_tokens": 310275464.0,
      "step": 8920
    },
    {
      "epoch": 1.6566388115134634,
      "grad_norm": 1.4051215223781823,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.884662389755249,
      "num_tokens": 310314900.0,
      "step": 8921
    },
    {
      "epoch": 1.6568245125348189,
      "grad_norm": 1.4157990110563077,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.869542121887207,
      "num_tokens": 310355780.0,
      "step": 8922
    },
    {
      "epoch": 1.6570102135561746,
      "grad_norm": 1.457164007194647,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8830145597457886,
      "num_tokens": 310389485.0,
      "step": 8923
    },
    {
      "epoch": 1.6571959145775301,
      "grad_norm": 1.4016085000645553,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8869216442108154,
      "num_tokens": 310427829.0,
      "step": 8924
    },
    {
      "epoch": 1.6573816155988856,
      "grad_norm": 1.5989387781526652,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.876121997833252,
      "num_tokens": 310460213.0,
      "step": 8925
    },
    {
      "epoch": 1.6575673166202414,
      "grad_norm": 1.658140459884654,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8646245002746582,
      "num_tokens": 310493101.0,
      "step": 8926
    },
    {
      "epoch": 1.657753017641597,
      "grad_norm": 1.5148884458315113,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8875722885131836,
      "num_tokens": 310530783.0,
      "step": 8927
    },
    {
      "epoch": 1.6579387186629526,
      "grad_norm": 1.5315190510432872,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.8689975142478943,
      "num_tokens": 310570724.0,
      "step": 8928
    },
    {
      "epoch": 1.6581244196843081,
      "grad_norm": 1.4234542444652225,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8845909833908081,
      "num_tokens": 310603848.0,
      "step": 8929
    },
    {
      "epoch": 1.6583101207056639,
      "grad_norm": 1.4225297667049084,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8869341611862183,
      "num_tokens": 310641049.0,
      "step": 8930
    },
    {
      "epoch": 1.6584958217270196,
      "grad_norm": 1.563598515687674,
      "learning_rate": 1e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8681282997131348,
      "num_tokens": 310673714.0,
      "step": 8931
    },
    {
      "epoch": 1.658681522748375,
      "grad_norm": 1.6714232735652836,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8840497732162476,
      "num_tokens": 310702458.0,
      "step": 8932
    },
    {
      "epoch": 1.6588672237697306,
      "grad_norm": 1.6982433295170698,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8787602782249451,
      "num_tokens": 310732315.0,
      "step": 8933
    },
    {
      "epoch": 1.6590529247910863,
      "grad_norm": 1.5899899334968053,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8757187724113464,
      "num_tokens": 310765443.0,
      "step": 8934
    },
    {
      "epoch": 1.659238625812442,
      "grad_norm": 1.561635834804809,
      "learning_rate": 1e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8612126708030701,
      "num_tokens": 310800613.0,
      "step": 8935
    },
    {
      "epoch": 1.6594243268337976,
      "grad_norm": 1.32945965895561,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8848634958267212,
      "num_tokens": 310840432.0,
      "step": 8936
    },
    {
      "epoch": 1.659610027855153,
      "grad_norm": 1.4213681084808314,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8791253566741943,
      "num_tokens": 310879129.0,
      "step": 8937
    },
    {
      "epoch": 1.6597957288765088,
      "grad_norm": 1.5658668858363096,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8737202286720276,
      "num_tokens": 310910470.0,
      "step": 8938
    },
    {
      "epoch": 1.6599814298978646,
      "grad_norm": 1.4773675788570588,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.879501223564148,
      "num_tokens": 310942183.0,
      "step": 8939
    },
    {
      "epoch": 1.66016713091922,
      "grad_norm": 1.6461641639444065,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.868605375289917,
      "num_tokens": 310971988.0,
      "step": 8940
    },
    {
      "epoch": 1.6603528319405756,
      "grad_norm": 1.384537496634751,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8895632028579712,
      "num_tokens": 311009075.0,
      "step": 8941
    },
    {
      "epoch": 1.6605385329619313,
      "grad_norm": 1.5850089249292054,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8815478086471558,
      "num_tokens": 311040613.0,
      "step": 8942
    },
    {
      "epoch": 1.660724233983287,
      "grad_norm": 1.5857104098511194,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8938268423080444,
      "num_tokens": 311069398.0,
      "step": 8943
    },
    {
      "epoch": 1.6609099350046426,
      "grad_norm": 1.3901026963630265,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8680704832077026,
      "num_tokens": 311116822.0,
      "step": 8944
    },
    {
      "epoch": 1.661095636025998,
      "grad_norm": 1.6278287609310347,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.88481605052948,
      "num_tokens": 311145091.0,
      "step": 8945
    },
    {
      "epoch": 1.6612813370473538,
      "grad_norm": 1.4331055967802497,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8725263476371765,
      "num_tokens": 311182792.0,
      "step": 8946
    },
    {
      "epoch": 1.6614670380687093,
      "grad_norm": 1.6617619285585878,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8705700635910034,
      "num_tokens": 311209368.0,
      "step": 8947
    },
    {
      "epoch": 1.6616527390900648,
      "grad_norm": 1.4855660664810515,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8872267603874207,
      "num_tokens": 311246730.0,
      "step": 8948
    },
    {
      "epoch": 1.6618384401114206,
      "grad_norm": 1.736094639456131,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8799132704734802,
      "num_tokens": 311278115.0,
      "step": 8949
    },
    {
      "epoch": 1.6620241411327763,
      "grad_norm": 1.5131786089673267,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8784106373786926,
      "num_tokens": 311311583.0,
      "step": 8950
    },
    {
      "epoch": 1.6622098421541318,
      "grad_norm": 1.684185460198444,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.86683589220047,
      "num_tokens": 311339946.0,
      "step": 8951
    },
    {
      "epoch": 1.6623955431754873,
      "grad_norm": 1.5223842416477296,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.881654679775238,
      "num_tokens": 311372051.0,
      "step": 8952
    },
    {
      "epoch": 1.662581244196843,
      "grad_norm": 1.451531970728145,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.874025285243988,
      "num_tokens": 311411680.0,
      "step": 8953
    },
    {
      "epoch": 1.6627669452181988,
      "grad_norm": 1.436812065369227,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.882370114326477,
      "num_tokens": 311449894.0,
      "step": 8954
    },
    {
      "epoch": 1.6629526462395543,
      "grad_norm": 1.494564971260475,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.878726601600647,
      "num_tokens": 311491299.0,
      "step": 8955
    },
    {
      "epoch": 1.6631383472609098,
      "grad_norm": 1.42686444747068,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8743874430656433,
      "num_tokens": 311527961.0,
      "step": 8956
    },
    {
      "epoch": 1.6633240482822655,
      "grad_norm": 1.4059990601119017,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8848772048950195,
      "num_tokens": 311566230.0,
      "step": 8957
    },
    {
      "epoch": 1.6635097493036213,
      "grad_norm": 1.3084001478071567,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.88603675365448,
      "num_tokens": 311608172.0,
      "step": 8958
    },
    {
      "epoch": 1.6636954503249768,
      "grad_norm": 1.5223361507586495,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8726843595504761,
      "num_tokens": 311642375.0,
      "step": 8959
    },
    {
      "epoch": 1.6638811513463323,
      "grad_norm": 1.403350784752098,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8691723942756653,
      "num_tokens": 311684867.0,
      "step": 8960
    },
    {
      "epoch": 1.664066852367688,
      "grad_norm": 1.4059795960880113,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8839707374572754,
      "num_tokens": 311722154.0,
      "step": 8961
    },
    {
      "epoch": 1.6642525533890438,
      "grad_norm": 1.49779886450851,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8719563484191895,
      "num_tokens": 311760629.0,
      "step": 8962
    },
    {
      "epoch": 1.6644382544103993,
      "grad_norm": 1.4904238097976539,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8802313804626465,
      "num_tokens": 311794134.0,
      "step": 8963
    },
    {
      "epoch": 1.6646239554317548,
      "grad_norm": 1.591137507289585,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8799618482589722,
      "num_tokens": 311820738.0,
      "step": 8964
    },
    {
      "epoch": 1.6648096564531105,
      "grad_norm": 1.4978828190170013,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8785268068313599,
      "num_tokens": 311854407.0,
      "step": 8965
    },
    {
      "epoch": 1.6649953574744663,
      "grad_norm": 1.57713844369864,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8693655133247375,
      "num_tokens": 311887906.0,
      "step": 8966
    },
    {
      "epoch": 1.6651810584958218,
      "grad_norm": 1.5779749404125858,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8753222227096558,
      "num_tokens": 311918633.0,
      "step": 8967
    },
    {
      "epoch": 1.6653667595171773,
      "grad_norm": 1.4879597212731381,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8752986192703247,
      "num_tokens": 311952980.0,
      "step": 8968
    },
    {
      "epoch": 1.665552460538533,
      "grad_norm": 1.6020588475480582,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8767061233520508,
      "num_tokens": 311982665.0,
      "step": 8969
    },
    {
      "epoch": 1.6657381615598887,
      "grad_norm": 1.4021180221815042,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8768913745880127,
      "num_tokens": 312020602.0,
      "step": 8970
    },
    {
      "epoch": 1.665923862581244,
      "grad_norm": 1.3176869060044383,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8789077401161194,
      "num_tokens": 312065636.0,
      "step": 8971
    },
    {
      "epoch": 1.6661095636025998,
      "grad_norm": 1.5595284142023131,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.876282811164856,
      "num_tokens": 312095373.0,
      "step": 8972
    },
    {
      "epoch": 1.6662952646239555,
      "grad_norm": 1.4776946972597762,
      "learning_rate": 1e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8642117977142334,
      "num_tokens": 312132725.0,
      "step": 8973
    },
    {
      "epoch": 1.666480965645311,
      "grad_norm": 1.5005120532011769,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8822575807571411,
      "num_tokens": 312167697.0,
      "step": 8974
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.3491929894333532,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8842987418174744,
      "num_tokens": 312208370.0,
      "step": 8975
    },
    {
      "epoch": 1.6668523676880223,
      "grad_norm": 1.6324340537448132,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8714734315872192,
      "num_tokens": 312239375.0,
      "step": 8976
    },
    {
      "epoch": 1.667038068709378,
      "grad_norm": 1.4712395250678778,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.8795216679573059,
      "num_tokens": 312274638.0,
      "step": 8977
    },
    {
      "epoch": 1.6672237697307335,
      "grad_norm": 1.3857500183930773,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8745771646499634,
      "num_tokens": 312314700.0,
      "step": 8978
    },
    {
      "epoch": 1.667409470752089,
      "grad_norm": 1.4277141893727325,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8856548070907593,
      "num_tokens": 312349556.0,
      "step": 8979
    },
    {
      "epoch": 1.6675951717734447,
      "grad_norm": 1.4786443484381486,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8838991522789001,
      "num_tokens": 312384426.0,
      "step": 8980
    },
    {
      "epoch": 1.6677808727948005,
      "grad_norm": 1.5094695255474144,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8793473243713379,
      "num_tokens": 312426623.0,
      "step": 8981
    },
    {
      "epoch": 1.667966573816156,
      "grad_norm": 1.723717161321228,
      "learning_rate": 1e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8637627363204956,
      "num_tokens": 312455250.0,
      "step": 8982
    },
    {
      "epoch": 1.6681522748375115,
      "grad_norm": 1.3619493505963927,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.880572497844696,
      "num_tokens": 312497780.0,
      "step": 8983
    },
    {
      "epoch": 1.6683379758588672,
      "grad_norm": 1.4635747584924454,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8631186485290527,
      "num_tokens": 312535907.0,
      "step": 8984
    },
    {
      "epoch": 1.668523676880223,
      "grad_norm": 1.4419436273020243,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8769881129264832,
      "num_tokens": 312572418.0,
      "step": 8985
    },
    {
      "epoch": 1.6687093779015785,
      "grad_norm": 1.4685706614241143,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8748515844345093,
      "num_tokens": 312611628.0,
      "step": 8986
    },
    {
      "epoch": 1.668895078922934,
      "grad_norm": 1.4580845990690248,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8725725412368774,
      "num_tokens": 312652969.0,
      "step": 8987
    },
    {
      "epoch": 1.6690807799442897,
      "grad_norm": 1.6179770929585922,
      "learning_rate": 1e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8619449138641357,
      "num_tokens": 312690897.0,
      "step": 8988
    },
    {
      "epoch": 1.6692664809656454,
      "grad_norm": 1.4261764787233366,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8765476942062378,
      "num_tokens": 312728981.0,
      "step": 8989
    },
    {
      "epoch": 1.669452181987001,
      "grad_norm": 1.5239654761727228,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8779504299163818,
      "num_tokens": 312762437.0,
      "step": 8990
    },
    {
      "epoch": 1.6696378830083565,
      "grad_norm": 1.4918671764219944,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8807142972946167,
      "num_tokens": 312794435.0,
      "step": 8991
    },
    {
      "epoch": 1.6698235840297122,
      "grad_norm": 1.4652109229551622,
      "learning_rate": 1e-06,
      "loss": 0.3564,
      "mean_token_accuracy": 0.8720680475234985,
      "num_tokens": 312831121.0,
      "step": 8992
    },
    {
      "epoch": 1.670009285051068,
      "grad_norm": 1.540834313998575,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8891817331314087,
      "num_tokens": 312867488.0,
      "step": 8993
    },
    {
      "epoch": 1.6701949860724234,
      "grad_norm": 1.521127295227109,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.867286205291748,
      "num_tokens": 312901599.0,
      "step": 8994
    },
    {
      "epoch": 1.670380687093779,
      "grad_norm": 1.4443691909449017,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8852638006210327,
      "num_tokens": 312938486.0,
      "step": 8995
    },
    {
      "epoch": 1.6705663881151347,
      "grad_norm": 1.3363318850463661,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8868553042411804,
      "num_tokens": 312974729.0,
      "step": 8996
    },
    {
      "epoch": 1.6707520891364902,
      "grad_norm": 1.497885754118438,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8842171430587769,
      "num_tokens": 313003786.0,
      "step": 8997
    },
    {
      "epoch": 1.6709377901578457,
      "grad_norm": 1.3786826673344792,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8839305639266968,
      "num_tokens": 313045994.0,
      "step": 8998
    },
    {
      "epoch": 1.6711234911792014,
      "grad_norm": 1.4536764025002602,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8794804811477661,
      "num_tokens": 313081207.0,
      "step": 8999
    },
    {
      "epoch": 1.6713091922005572,
      "grad_norm": 1.459045522272049,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8884913921356201,
      "num_tokens": 313112763.0,
      "step": 9000
    },
    {
      "epoch": 1.6714948932219127,
      "grad_norm": 1.4644681888092022,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8673608303070068,
      "num_tokens": 313148861.0,
      "step": 9001
    },
    {
      "epoch": 1.6716805942432682,
      "grad_norm": 1.5099958382919345,
      "learning_rate": 1e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8667759299278259,
      "num_tokens": 313184831.0,
      "step": 9002
    },
    {
      "epoch": 1.671866295264624,
      "grad_norm": 1.6157349694600733,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8728718757629395,
      "num_tokens": 313214166.0,
      "step": 9003
    },
    {
      "epoch": 1.6720519962859797,
      "grad_norm": 1.424555904113373,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8755056858062744,
      "num_tokens": 313254152.0,
      "step": 9004
    },
    {
      "epoch": 1.6722376973073352,
      "grad_norm": 1.3216503510088424,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8883768916130066,
      "num_tokens": 313295148.0,
      "step": 9005
    },
    {
      "epoch": 1.6724233983286907,
      "grad_norm": 1.5387303349455568,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8650736808776855,
      "num_tokens": 313334650.0,
      "step": 9006
    },
    {
      "epoch": 1.6726090993500464,
      "grad_norm": 1.533083407302871,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8837295770645142,
      "num_tokens": 313365327.0,
      "step": 9007
    },
    {
      "epoch": 1.6727948003714022,
      "grad_norm": 1.7992216598176338,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8718832731246948,
      "num_tokens": 313395528.0,
      "step": 9008
    },
    {
      "epoch": 1.6729805013927577,
      "grad_norm": 1.4281587070491781,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8865624666213989,
      "num_tokens": 313430965.0,
      "step": 9009
    },
    {
      "epoch": 1.6731662024141132,
      "grad_norm": 1.6930551335732116,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8764479160308838,
      "num_tokens": 313459339.0,
      "step": 9010
    },
    {
      "epoch": 1.673351903435469,
      "grad_norm": 1.5309898383840874,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8810371160507202,
      "num_tokens": 313492804.0,
      "step": 9011
    },
    {
      "epoch": 1.6735376044568246,
      "grad_norm": 1.747929587969937,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8786517381668091,
      "num_tokens": 313521549.0,
      "step": 9012
    },
    {
      "epoch": 1.6737233054781802,
      "grad_norm": 1.502994126826737,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8826215863227844,
      "num_tokens": 313554468.0,
      "step": 9013
    },
    {
      "epoch": 1.6739090064995357,
      "grad_norm": 1.4987282187588662,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8745155334472656,
      "num_tokens": 313589104.0,
      "step": 9014
    },
    {
      "epoch": 1.6740947075208914,
      "grad_norm": 1.5499495473779175,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.874254584312439,
      "num_tokens": 313622943.0,
      "step": 9015
    },
    {
      "epoch": 1.6742804085422471,
      "grad_norm": 1.5450158310730038,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8907462358474731,
      "num_tokens": 313654150.0,
      "step": 9016
    },
    {
      "epoch": 1.6744661095636026,
      "grad_norm": 1.7914641500113109,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8688293099403381,
      "num_tokens": 313686476.0,
      "step": 9017
    },
    {
      "epoch": 1.6746518105849582,
      "grad_norm": 1.4520252055797265,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8709030151367188,
      "num_tokens": 313723235.0,
      "step": 9018
    },
    {
      "epoch": 1.674837511606314,
      "grad_norm": 1.5406940306789856,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.881529688835144,
      "num_tokens": 313753423.0,
      "step": 9019
    },
    {
      "epoch": 1.6750232126276694,
      "grad_norm": 1.556484538389134,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8912442922592163,
      "num_tokens": 313782227.0,
      "step": 9020
    },
    {
      "epoch": 1.675208913649025,
      "grad_norm": 1.490584140373382,
      "learning_rate": 1e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8675290942192078,
      "num_tokens": 313819249.0,
      "step": 9021
    },
    {
      "epoch": 1.6753946146703806,
      "grad_norm": 1.389741152152576,
      "learning_rate": 1e-06,
      "loss": 0.2784,
      "mean_token_accuracy": 0.9025817513465881,
      "num_tokens": 313852474.0,
      "step": 9022
    },
    {
      "epoch": 1.6755803156917364,
      "grad_norm": 1.3814448280120442,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8766646981239319,
      "num_tokens": 313892259.0,
      "step": 9023
    },
    {
      "epoch": 1.6757660167130919,
      "grad_norm": 1.435442013891892,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8892632722854614,
      "num_tokens": 313931330.0,
      "step": 9024
    },
    {
      "epoch": 1.6759517177344474,
      "grad_norm": 1.4832339507782564,
      "learning_rate": 1e-06,
      "loss": 0.3045,
      "mean_token_accuracy": 0.8949369192123413,
      "num_tokens": 313961587.0,
      "step": 9025
    },
    {
      "epoch": 1.6761374187558031,
      "grad_norm": 1.6706559748356722,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8706931471824646,
      "num_tokens": 313993959.0,
      "step": 9026
    },
    {
      "epoch": 1.6763231197771589,
      "grad_norm": 1.4625035221850988,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8814652562141418,
      "num_tokens": 314029356.0,
      "step": 9027
    },
    {
      "epoch": 1.6765088207985144,
      "grad_norm": 1.322554507441506,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8939546346664429,
      "num_tokens": 314066797.0,
      "step": 9028
    },
    {
      "epoch": 1.6766945218198699,
      "grad_norm": 1.6423554733296475,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8663142919540405,
      "num_tokens": 314098622.0,
      "step": 9029
    },
    {
      "epoch": 1.6768802228412256,
      "grad_norm": 1.6842166516607766,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8783758282661438,
      "num_tokens": 314126875.0,
      "step": 9030
    },
    {
      "epoch": 1.6770659238625814,
      "grad_norm": 1.4429191974602504,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8800384998321533,
      "num_tokens": 314162901.0,
      "step": 9031
    },
    {
      "epoch": 1.6772516248839369,
      "grad_norm": 1.4860937966956564,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8786236047744751,
      "num_tokens": 314200398.0,
      "step": 9032
    },
    {
      "epoch": 1.6774373259052924,
      "grad_norm": 1.5307444525953948,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.8799088001251221,
      "num_tokens": 314232322.0,
      "step": 9033
    },
    {
      "epoch": 1.677623026926648,
      "grad_norm": 1.6159243087943025,
      "learning_rate": 1e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8607777953147888,
      "num_tokens": 314272041.0,
      "step": 9034
    },
    {
      "epoch": 1.6778087279480038,
      "grad_norm": 1.5193054291416257,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8739738464355469,
      "num_tokens": 314304827.0,
      "step": 9035
    },
    {
      "epoch": 1.6779944289693594,
      "grad_norm": 1.591894177742128,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.869469165802002,
      "num_tokens": 314335855.0,
      "step": 9036
    },
    {
      "epoch": 1.6781801299907149,
      "grad_norm": 1.6705421658054758,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8660976886749268,
      "num_tokens": 314369892.0,
      "step": 9037
    },
    {
      "epoch": 1.6783658310120706,
      "grad_norm": 1.3794067508992696,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8826488852500916,
      "num_tokens": 314407602.0,
      "step": 9038
    },
    {
      "epoch": 1.6785515320334263,
      "grad_norm": 1.5136235329163297,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.870453953742981,
      "num_tokens": 314441421.0,
      "step": 9039
    },
    {
      "epoch": 1.6787372330547818,
      "grad_norm": 1.4563466416680508,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8871423006057739,
      "num_tokens": 314475932.0,
      "step": 9040
    },
    {
      "epoch": 1.6789229340761374,
      "grad_norm": 1.564923351999391,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8657349944114685,
      "num_tokens": 314506339.0,
      "step": 9041
    },
    {
      "epoch": 1.679108635097493,
      "grad_norm": 1.582376409456413,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.884416401386261,
      "num_tokens": 314540719.0,
      "step": 9042
    },
    {
      "epoch": 1.6792943361188488,
      "grad_norm": 1.5192108417409853,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8741352558135986,
      "num_tokens": 314576383.0,
      "step": 9043
    },
    {
      "epoch": 1.679480037140204,
      "grad_norm": 1.4590799824824414,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.881974995136261,
      "num_tokens": 314614327.0,
      "step": 9044
    },
    {
      "epoch": 1.6796657381615598,
      "grad_norm": 1.4663060754663007,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8813827633857727,
      "num_tokens": 314649613.0,
      "step": 9045
    },
    {
      "epoch": 1.6798514391829156,
      "grad_norm": 1.4439729659875284,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8837214112281799,
      "num_tokens": 314685958.0,
      "step": 9046
    },
    {
      "epoch": 1.680037140204271,
      "grad_norm": 1.5755870428165721,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8742232918739319,
      "num_tokens": 314718338.0,
      "step": 9047
    },
    {
      "epoch": 1.6802228412256266,
      "grad_norm": 1.5891135132696312,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8738695979118347,
      "num_tokens": 314751414.0,
      "step": 9048
    },
    {
      "epoch": 1.6804085422469823,
      "grad_norm": 1.5586018434920328,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8820539712905884,
      "num_tokens": 314785614.0,
      "step": 9049
    },
    {
      "epoch": 1.680594243268338,
      "grad_norm": 1.5320694466521116,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8726269602775574,
      "num_tokens": 314818417.0,
      "step": 9050
    },
    {
      "epoch": 1.6807799442896936,
      "grad_norm": 1.4317213202222236,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8812142014503479,
      "num_tokens": 314854451.0,
      "step": 9051
    },
    {
      "epoch": 1.680965645311049,
      "grad_norm": 1.4428013592347715,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8791850209236145,
      "num_tokens": 314893352.0,
      "step": 9052
    },
    {
      "epoch": 1.6811513463324048,
      "grad_norm": 1.318692754665186,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8908970355987549,
      "num_tokens": 314932430.0,
      "step": 9053
    },
    {
      "epoch": 1.6813370473537605,
      "grad_norm": 1.5990135252885358,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8754137754440308,
      "num_tokens": 314963290.0,
      "step": 9054
    },
    {
      "epoch": 1.681522748375116,
      "grad_norm": 1.3716914036719026,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8905670642852783,
      "num_tokens": 314999282.0,
      "step": 9055
    },
    {
      "epoch": 1.6817084493964716,
      "grad_norm": 1.5768265380442115,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8850277066230774,
      "num_tokens": 315028685.0,
      "step": 9056
    },
    {
      "epoch": 1.6818941504178273,
      "grad_norm": 1.364382769617669,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8726337552070618,
      "num_tokens": 315068091.0,
      "step": 9057
    },
    {
      "epoch": 1.682079851439183,
      "grad_norm": 1.6662137441188287,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8822875022888184,
      "num_tokens": 315096134.0,
      "step": 9058
    },
    {
      "epoch": 1.6822655524605385,
      "grad_norm": 1.6211136845012666,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8823522329330444,
      "num_tokens": 315130492.0,
      "step": 9059
    },
    {
      "epoch": 1.682451253481894,
      "grad_norm": 1.435979455777625,
      "learning_rate": 1e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8645882606506348,
      "num_tokens": 315170377.0,
      "step": 9060
    },
    {
      "epoch": 1.6826369545032498,
      "grad_norm": 1.4165630235290887,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8796191215515137,
      "num_tokens": 315210300.0,
      "step": 9061
    },
    {
      "epoch": 1.6828226555246055,
      "grad_norm": 1.4956212794066315,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8711153268814087,
      "num_tokens": 315242136.0,
      "step": 9062
    },
    {
      "epoch": 1.683008356545961,
      "grad_norm": 1.5218424169155458,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.873684823513031,
      "num_tokens": 315275342.0,
      "step": 9063
    },
    {
      "epoch": 1.6831940575673165,
      "grad_norm": 1.5826877969233168,
      "learning_rate": 1e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8695127367973328,
      "num_tokens": 315311841.0,
      "step": 9064
    },
    {
      "epoch": 1.6833797585886723,
      "grad_norm": 1.4302029596172858,
      "learning_rate": 1e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8711429834365845,
      "num_tokens": 315351730.0,
      "step": 9065
    },
    {
      "epoch": 1.683565459610028,
      "grad_norm": 1.4728477889111813,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8774985074996948,
      "num_tokens": 315387968.0,
      "step": 9066
    },
    {
      "epoch": 1.6837511606313835,
      "grad_norm": 1.3399081427399528,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8920609951019287,
      "num_tokens": 315425909.0,
      "step": 9067
    },
    {
      "epoch": 1.683936861652739,
      "grad_norm": 1.5019739542818251,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8803349137306213,
      "num_tokens": 315459020.0,
      "step": 9068
    },
    {
      "epoch": 1.6841225626740948,
      "grad_norm": 1.4743554086938035,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8910577297210693,
      "num_tokens": 315490886.0,
      "step": 9069
    },
    {
      "epoch": 1.6843082636954503,
      "grad_norm": 1.4764579425089337,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8779071569442749,
      "num_tokens": 315530068.0,
      "step": 9070
    },
    {
      "epoch": 1.6844939647168058,
      "grad_norm": 1.412960898253757,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8777649402618408,
      "num_tokens": 315569715.0,
      "step": 9071
    },
    {
      "epoch": 1.6846796657381615,
      "grad_norm": 1.4261465350705789,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.885971188545227,
      "num_tokens": 315603128.0,
      "step": 9072
    },
    {
      "epoch": 1.6848653667595173,
      "grad_norm": 1.5155473044624583,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8655214905738831,
      "num_tokens": 315638122.0,
      "step": 9073
    },
    {
      "epoch": 1.6850510677808728,
      "grad_norm": 1.625914317397132,
      "learning_rate": 1e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8650915622711182,
      "num_tokens": 315668957.0,
      "step": 9074
    },
    {
      "epoch": 1.6852367688022283,
      "grad_norm": 1.514802113866681,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8797249794006348,
      "num_tokens": 315704218.0,
      "step": 9075
    },
    {
      "epoch": 1.685422469823584,
      "grad_norm": 1.6534211312975677,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8769550323486328,
      "num_tokens": 315736091.0,
      "step": 9076
    },
    {
      "epoch": 1.6856081708449397,
      "grad_norm": 1.533626620903343,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8805968761444092,
      "num_tokens": 315770248.0,
      "step": 9077
    },
    {
      "epoch": 1.6857938718662953,
      "grad_norm": 1.484256273793592,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8753677606582642,
      "num_tokens": 315806215.0,
      "step": 9078
    },
    {
      "epoch": 1.6859795728876508,
      "grad_norm": 1.4735960910514467,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8807933330535889,
      "num_tokens": 315847568.0,
      "step": 9079
    },
    {
      "epoch": 1.6861652739090065,
      "grad_norm": 1.5619478385282877,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8752211332321167,
      "num_tokens": 315880089.0,
      "step": 9080
    },
    {
      "epoch": 1.6863509749303622,
      "grad_norm": 1.4746760580273954,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8709630370140076,
      "num_tokens": 315915058.0,
      "step": 9081
    },
    {
      "epoch": 1.6865366759517177,
      "grad_norm": 1.6401355027304434,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8821135759353638,
      "num_tokens": 315943122.0,
      "step": 9082
    },
    {
      "epoch": 1.6867223769730733,
      "grad_norm": 1.399185560497335,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8827760219573975,
      "num_tokens": 315982023.0,
      "step": 9083
    },
    {
      "epoch": 1.686908077994429,
      "grad_norm": 1.40165758588921,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8782665729522705,
      "num_tokens": 316019391.0,
      "step": 9084
    },
    {
      "epoch": 1.6870937790157847,
      "grad_norm": 1.5353406376718486,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8768132328987122,
      "num_tokens": 316053100.0,
      "step": 9085
    },
    {
      "epoch": 1.6872794800371402,
      "grad_norm": 1.475699699330823,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.87348872423172,
      "num_tokens": 316087258.0,
      "step": 9086
    },
    {
      "epoch": 1.6874651810584957,
      "grad_norm": 1.5129169660825277,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8774078488349915,
      "num_tokens": 316120033.0,
      "step": 9087
    },
    {
      "epoch": 1.6876508820798515,
      "grad_norm": 1.3223735546303896,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8887854218482971,
      "num_tokens": 316162738.0,
      "step": 9088
    },
    {
      "epoch": 1.6878365831012072,
      "grad_norm": 1.4176862662584118,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.877601146697998,
      "num_tokens": 316203268.0,
      "step": 9089
    },
    {
      "epoch": 1.6880222841225627,
      "grad_norm": 1.5099150713565739,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8686661720275879,
      "num_tokens": 316239873.0,
      "step": 9090
    },
    {
      "epoch": 1.6882079851439182,
      "grad_norm": 1.4041202206784937,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8844203352928162,
      "num_tokens": 316277393.0,
      "step": 9091
    },
    {
      "epoch": 1.688393686165274,
      "grad_norm": 1.6883253222378918,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8862903118133545,
      "num_tokens": 316312341.0,
      "step": 9092
    },
    {
      "epoch": 1.6885793871866295,
      "grad_norm": 1.49708914787014,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8739765286445618,
      "num_tokens": 316345373.0,
      "step": 9093
    },
    {
      "epoch": 1.688765088207985,
      "grad_norm": 1.3472699776606314,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8907887935638428,
      "num_tokens": 316384681.0,
      "step": 9094
    },
    {
      "epoch": 1.6889507892293407,
      "grad_norm": 1.4807955156602253,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.875190258026123,
      "num_tokens": 316423537.0,
      "step": 9095
    },
    {
      "epoch": 1.6891364902506965,
      "grad_norm": 1.4475172978465984,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8846814036369324,
      "num_tokens": 316460899.0,
      "step": 9096
    },
    {
      "epoch": 1.689322191272052,
      "grad_norm": 1.5084664434815103,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8839378356933594,
      "num_tokens": 316497297.0,
      "step": 9097
    },
    {
      "epoch": 1.6895078922934075,
      "grad_norm": 1.6267944527504334,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8697783946990967,
      "num_tokens": 316527627.0,
      "step": 9098
    },
    {
      "epoch": 1.6896935933147632,
      "grad_norm": 1.6218215679510855,
      "learning_rate": 1e-06,
      "loss": 0.3734,
      "mean_token_accuracy": 0.8746762275695801,
      "num_tokens": 316562516.0,
      "step": 9099
    },
    {
      "epoch": 1.689879294336119,
      "grad_norm": 1.6134436468891054,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.8719758987426758,
      "num_tokens": 316597577.0,
      "step": 9100
    },
    {
      "epoch": 1.6900649953574745,
      "grad_norm": 1.4626227744939924,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8814453482627869,
      "num_tokens": 316635749.0,
      "step": 9101
    },
    {
      "epoch": 1.69025069637883,
      "grad_norm": 1.577271176010885,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8696544170379639,
      "num_tokens": 316667319.0,
      "step": 9102
    },
    {
      "epoch": 1.6904363974001857,
      "grad_norm": 1.638949017843003,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.879413366317749,
      "num_tokens": 316696925.0,
      "step": 9103
    },
    {
      "epoch": 1.6906220984215414,
      "grad_norm": 1.4739792717622628,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8753722906112671,
      "num_tokens": 316729750.0,
      "step": 9104
    },
    {
      "epoch": 1.690807799442897,
      "grad_norm": 1.5276381625079718,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8744890093803406,
      "num_tokens": 316762336.0,
      "step": 9105
    },
    {
      "epoch": 1.6909935004642525,
      "grad_norm": 1.7270439151994919,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8932034969329834,
      "num_tokens": 316786526.0,
      "step": 9106
    },
    {
      "epoch": 1.6911792014856082,
      "grad_norm": 1.5727613590185117,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8837951421737671,
      "num_tokens": 316815063.0,
      "step": 9107
    },
    {
      "epoch": 1.691364902506964,
      "grad_norm": 1.5106531473304852,
      "learning_rate": 1e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8481818437576294,
      "num_tokens": 316857203.0,
      "step": 9108
    },
    {
      "epoch": 1.6915506035283194,
      "grad_norm": 1.4042116486118497,
      "learning_rate": 1e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.8787126541137695,
      "num_tokens": 316896134.0,
      "step": 9109
    },
    {
      "epoch": 1.691736304549675,
      "grad_norm": 1.4195729736247877,
      "learning_rate": 1e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8682400584220886,
      "num_tokens": 316938054.0,
      "step": 9110
    },
    {
      "epoch": 1.6919220055710307,
      "grad_norm": 1.3705297393420663,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8839815258979797,
      "num_tokens": 316980575.0,
      "step": 9111
    },
    {
      "epoch": 1.6921077065923864,
      "grad_norm": 1.4545658328598605,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8805768489837646,
      "num_tokens": 317016073.0,
      "step": 9112
    },
    {
      "epoch": 1.692293407613742,
      "grad_norm": 1.4229990625217026,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8648128509521484,
      "num_tokens": 317052617.0,
      "step": 9113
    },
    {
      "epoch": 1.6924791086350974,
      "grad_norm": 1.69231368431796,
      "learning_rate": 1e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8491772413253784,
      "num_tokens": 317084916.0,
      "step": 9114
    },
    {
      "epoch": 1.6926648096564532,
      "grad_norm": 1.5839239201658197,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8705248832702637,
      "num_tokens": 317121140.0,
      "step": 9115
    },
    {
      "epoch": 1.6928505106778087,
      "grad_norm": 1.6086459295238258,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.866025984287262,
      "num_tokens": 317151725.0,
      "step": 9116
    },
    {
      "epoch": 1.6930362116991642,
      "grad_norm": 1.3575918306261985,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8896083235740662,
      "num_tokens": 317187057.0,
      "step": 9117
    },
    {
      "epoch": 1.69322191272052,
      "grad_norm": 1.5842023806828622,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8761539459228516,
      "num_tokens": 317223046.0,
      "step": 9118
    },
    {
      "epoch": 1.6934076137418757,
      "grad_norm": 1.5953341699007115,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8800176382064819,
      "num_tokens": 317250633.0,
      "step": 9119
    },
    {
      "epoch": 1.6935933147632312,
      "grad_norm": 1.3432194498247048,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8839882612228394,
      "num_tokens": 317288505.0,
      "step": 9120
    },
    {
      "epoch": 1.6937790157845867,
      "grad_norm": 1.7911304266481511,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8811787366867065,
      "num_tokens": 317314567.0,
      "step": 9121
    },
    {
      "epoch": 1.6939647168059424,
      "grad_norm": 1.4178594280358543,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8833925127983093,
      "num_tokens": 317351839.0,
      "step": 9122
    },
    {
      "epoch": 1.6941504178272981,
      "grad_norm": 1.4635580940658368,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8688235282897949,
      "num_tokens": 317388384.0,
      "step": 9123
    },
    {
      "epoch": 1.6943361188486536,
      "grad_norm": 1.6158620456192887,
      "learning_rate": 1e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8653472661972046,
      "num_tokens": 317421804.0,
      "step": 9124
    },
    {
      "epoch": 1.6945218198700092,
      "grad_norm": 1.4616972336304375,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8772280216217041,
      "num_tokens": 317454368.0,
      "step": 9125
    },
    {
      "epoch": 1.694707520891365,
      "grad_norm": 1.5693016734387852,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8659900426864624,
      "num_tokens": 317491512.0,
      "step": 9126
    },
    {
      "epoch": 1.6948932219127206,
      "grad_norm": 1.5930471091692304,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8796037435531616,
      "num_tokens": 317522665.0,
      "step": 9127
    },
    {
      "epoch": 1.6950789229340761,
      "grad_norm": 1.521108112700505,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.883173406124115,
      "num_tokens": 317555472.0,
      "step": 9128
    },
    {
      "epoch": 1.6952646239554316,
      "grad_norm": 1.4883052475149259,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8755044937133789,
      "num_tokens": 317593371.0,
      "step": 9129
    },
    {
      "epoch": 1.6954503249767874,
      "grad_norm": 1.4046922921799814,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8762474656105042,
      "num_tokens": 317630556.0,
      "step": 9130
    },
    {
      "epoch": 1.6956360259981431,
      "grad_norm": 1.4981470637813665,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8588240146636963,
      "num_tokens": 317668909.0,
      "step": 9131
    },
    {
      "epoch": 1.6958217270194986,
      "grad_norm": 1.4822504969198467,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.881104588508606,
      "num_tokens": 317701995.0,
      "step": 9132
    },
    {
      "epoch": 1.6960074280408541,
      "grad_norm": 1.5702891971687432,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8705923557281494,
      "num_tokens": 317734222.0,
      "step": 9133
    },
    {
      "epoch": 1.6961931290622099,
      "grad_norm": 1.475664124660769,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8889592885971069,
      "num_tokens": 317771903.0,
      "step": 9134
    },
    {
      "epoch": 1.6963788300835656,
      "grad_norm": 1.6071626425067522,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8743113279342651,
      "num_tokens": 317806954.0,
      "step": 9135
    },
    {
      "epoch": 1.6965645311049211,
      "grad_norm": 1.393901256973303,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8781628608703613,
      "num_tokens": 317850881.0,
      "step": 9136
    },
    {
      "epoch": 1.6967502321262766,
      "grad_norm": 1.4127175075649263,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8716946840286255,
      "num_tokens": 317890154.0,
      "step": 9137
    },
    {
      "epoch": 1.6969359331476324,
      "grad_norm": 1.4606898821493886,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8724721670150757,
      "num_tokens": 317925247.0,
      "step": 9138
    },
    {
      "epoch": 1.697121634168988,
      "grad_norm": 1.5306984891341262,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.884040355682373,
      "num_tokens": 317960385.0,
      "step": 9139
    },
    {
      "epoch": 1.6973073351903436,
      "grad_norm": 1.6497322613467047,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8732983469963074,
      "num_tokens": 317991564.0,
      "step": 9140
    },
    {
      "epoch": 1.6974930362116991,
      "grad_norm": 1.549862065421979,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8950746059417725,
      "num_tokens": 318023811.0,
      "step": 9141
    },
    {
      "epoch": 1.6976787372330548,
      "grad_norm": 1.431902262129642,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8820434212684631,
      "num_tokens": 318059940.0,
      "step": 9142
    },
    {
      "epoch": 1.6978644382544104,
      "grad_norm": 1.5405075812710827,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.884479820728302,
      "num_tokens": 318095330.0,
      "step": 9143
    },
    {
      "epoch": 1.6980501392757659,
      "grad_norm": 1.721744970430196,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8793468475341797,
      "num_tokens": 318120612.0,
      "step": 9144
    },
    {
      "epoch": 1.6982358402971216,
      "grad_norm": 1.5870734952534526,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8770256042480469,
      "num_tokens": 318152691.0,
      "step": 9145
    },
    {
      "epoch": 1.6984215413184773,
      "grad_norm": 1.5804063879288515,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8750264644622803,
      "num_tokens": 318186062.0,
      "step": 9146
    },
    {
      "epoch": 1.6986072423398328,
      "grad_norm": 1.7126667267764162,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8688642978668213,
      "num_tokens": 318215223.0,
      "step": 9147
    },
    {
      "epoch": 1.6987929433611884,
      "grad_norm": 1.4804089441167576,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8783430457115173,
      "num_tokens": 318253653.0,
      "step": 9148
    },
    {
      "epoch": 1.698978644382544,
      "grad_norm": 1.5569266548081035,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.879167914390564,
      "num_tokens": 318290000.0,
      "step": 9149
    },
    {
      "epoch": 1.6991643454038998,
      "grad_norm": 1.5479487316293359,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8784052133560181,
      "num_tokens": 318326581.0,
      "step": 9150
    },
    {
      "epoch": 1.6993500464252553,
      "grad_norm": 1.5568081863045087,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8709625005722046,
      "num_tokens": 318363698.0,
      "step": 9151
    },
    {
      "epoch": 1.6995357474466108,
      "grad_norm": 1.5465301643790685,
      "learning_rate": 1e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8676971197128296,
      "num_tokens": 318402114.0,
      "step": 9152
    },
    {
      "epoch": 1.6997214484679666,
      "grad_norm": 1.5984221610924658,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8698145151138306,
      "num_tokens": 318433952.0,
      "step": 9153
    },
    {
      "epoch": 1.6999071494893223,
      "grad_norm": 1.4146299899728296,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8822463750839233,
      "num_tokens": 318471702.0,
      "step": 9154
    },
    {
      "epoch": 1.7000928505106778,
      "grad_norm": 1.4260421411813609,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8916767835617065,
      "num_tokens": 318509541.0,
      "step": 9155
    },
    {
      "epoch": 1.7002785515320333,
      "grad_norm": 1.5049275590287972,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8813741207122803,
      "num_tokens": 318544362.0,
      "step": 9156
    },
    {
      "epoch": 1.700464252553389,
      "grad_norm": 1.4410659921799542,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8727811574935913,
      "num_tokens": 318585474.0,
      "step": 9157
    },
    {
      "epoch": 1.7006499535747448,
      "grad_norm": 1.4226993125747898,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8777634501457214,
      "num_tokens": 318622353.0,
      "step": 9158
    },
    {
      "epoch": 1.7008356545961003,
      "grad_norm": 1.3495957787711095,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8866747617721558,
      "num_tokens": 318665163.0,
      "step": 9159
    },
    {
      "epoch": 1.7010213556174558,
      "grad_norm": 1.5127109871084161,
      "learning_rate": 1e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8685032725334167,
      "num_tokens": 318701512.0,
      "step": 9160
    },
    {
      "epoch": 1.7012070566388116,
      "grad_norm": 1.7074150096630594,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.864827036857605,
      "num_tokens": 318731132.0,
      "step": 9161
    },
    {
      "epoch": 1.7013927576601673,
      "grad_norm": 1.5353229444689602,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.881668210029602,
      "num_tokens": 318764510.0,
      "step": 9162
    },
    {
      "epoch": 1.7015784586815228,
      "grad_norm": 1.5836682047330972,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8680355548858643,
      "num_tokens": 318795738.0,
      "step": 9163
    },
    {
      "epoch": 1.7017641597028783,
      "grad_norm": 1.590558165552058,
      "learning_rate": 1e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8630878329277039,
      "num_tokens": 318830203.0,
      "step": 9164
    },
    {
      "epoch": 1.701949860724234,
      "grad_norm": 1.4679973263830906,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8724070191383362,
      "num_tokens": 318865100.0,
      "step": 9165
    },
    {
      "epoch": 1.7021355617455896,
      "grad_norm": 1.6177713261043114,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8741257190704346,
      "num_tokens": 318896160.0,
      "step": 9166
    },
    {
      "epoch": 1.702321262766945,
      "grad_norm": 1.3823582443341207,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8844457864761353,
      "num_tokens": 318934109.0,
      "step": 9167
    },
    {
      "epoch": 1.7025069637883008,
      "grad_norm": 1.509063031816856,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8729476928710938,
      "num_tokens": 318970602.0,
      "step": 9168
    },
    {
      "epoch": 1.7026926648096565,
      "grad_norm": 1.444012230957511,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8721351623535156,
      "num_tokens": 319010845.0,
      "step": 9169
    },
    {
      "epoch": 1.702878365831012,
      "grad_norm": 1.4889725398532725,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8671321868896484,
      "num_tokens": 319047800.0,
      "step": 9170
    },
    {
      "epoch": 1.7030640668523676,
      "grad_norm": 1.541533461531441,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8856172561645508,
      "num_tokens": 319080531.0,
      "step": 9171
    },
    {
      "epoch": 1.7032497678737233,
      "grad_norm": 1.5530341029248182,
      "learning_rate": 1e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8644058704376221,
      "num_tokens": 319113980.0,
      "step": 9172
    },
    {
      "epoch": 1.703435468895079,
      "grad_norm": 1.4161518372101927,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8749094605445862,
      "num_tokens": 319151293.0,
      "step": 9173
    },
    {
      "epoch": 1.7036211699164345,
      "grad_norm": 1.510395169842534,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8902931213378906,
      "num_tokens": 319181573.0,
      "step": 9174
    },
    {
      "epoch": 1.70380687093779,
      "grad_norm": 1.4720221254927808,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8809957504272461,
      "num_tokens": 319213163.0,
      "step": 9175
    },
    {
      "epoch": 1.7039925719591458,
      "grad_norm": 1.4411271770224885,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8829764723777771,
      "num_tokens": 319248282.0,
      "step": 9176
    },
    {
      "epoch": 1.7041782729805015,
      "grad_norm": 1.5738913580665108,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8710318207740784,
      "num_tokens": 319280499.0,
      "step": 9177
    },
    {
      "epoch": 1.704363974001857,
      "grad_norm": 1.5058678286359939,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8709328770637512,
      "num_tokens": 319317785.0,
      "step": 9178
    },
    {
      "epoch": 1.7045496750232125,
      "grad_norm": 1.5215774750924087,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8792577981948853,
      "num_tokens": 319349220.0,
      "step": 9179
    },
    {
      "epoch": 1.7047353760445683,
      "grad_norm": 1.5795290316389454,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8763086795806885,
      "num_tokens": 319377543.0,
      "step": 9180
    },
    {
      "epoch": 1.704921077065924,
      "grad_norm": 1.5355539364738335,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8711096048355103,
      "num_tokens": 319411611.0,
      "step": 9181
    },
    {
      "epoch": 1.7051067780872795,
      "grad_norm": 1.3664562573196486,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8856159448623657,
      "num_tokens": 319448286.0,
      "step": 9182
    },
    {
      "epoch": 1.705292479108635,
      "grad_norm": 1.6349304916265037,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8784404993057251,
      "num_tokens": 319477594.0,
      "step": 9183
    },
    {
      "epoch": 1.7054781801299908,
      "grad_norm": 1.5615136410812949,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8737026453018188,
      "num_tokens": 319510387.0,
      "step": 9184
    },
    {
      "epoch": 1.7056638811513465,
      "grad_norm": 1.4411538402493804,
      "learning_rate": 1e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.8797703981399536,
      "num_tokens": 319548398.0,
      "step": 9185
    },
    {
      "epoch": 1.705849582172702,
      "grad_norm": 1.4359514520053165,
      "learning_rate": 1e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.864662766456604,
      "num_tokens": 319589762.0,
      "step": 9186
    },
    {
      "epoch": 1.7060352831940575,
      "grad_norm": 1.49509287862055,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8772166967391968,
      "num_tokens": 319622577.0,
      "step": 9187
    },
    {
      "epoch": 1.7062209842154132,
      "grad_norm": 1.2894651198167513,
      "learning_rate": 1e-06,
      "loss": 0.3023,
      "mean_token_accuracy": 0.8920114040374756,
      "num_tokens": 319661804.0,
      "step": 9188
    },
    {
      "epoch": 1.7064066852367687,
      "grad_norm": 1.5682996506942652,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8742803335189819,
      "num_tokens": 319695787.0,
      "step": 9189
    },
    {
      "epoch": 1.7065923862581243,
      "grad_norm": 1.484360424172678,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8817033767700195,
      "num_tokens": 319729705.0,
      "step": 9190
    },
    {
      "epoch": 1.70677808727948,
      "grad_norm": 1.618321974787918,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8708504438400269,
      "num_tokens": 319762135.0,
      "step": 9191
    },
    {
      "epoch": 1.7069637883008357,
      "grad_norm": 1.497798532956076,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8859786987304688,
      "num_tokens": 319796868.0,
      "step": 9192
    },
    {
      "epoch": 1.7071494893221912,
      "grad_norm": 1.5511571735652936,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8716668486595154,
      "num_tokens": 319825263.0,
      "step": 9193
    },
    {
      "epoch": 1.7073351903435467,
      "grad_norm": 1.5620651991043344,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.877862274646759,
      "num_tokens": 319855007.0,
      "step": 9194
    },
    {
      "epoch": 1.7075208913649025,
      "grad_norm": 1.4101225371813326,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.876234769821167,
      "num_tokens": 319891182.0,
      "step": 9195
    },
    {
      "epoch": 1.7077065923862582,
      "grad_norm": 1.3476791240987662,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.8958996534347534,
      "num_tokens": 319925851.0,
      "step": 9196
    },
    {
      "epoch": 1.7078922934076137,
      "grad_norm": 1.3854003112414488,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8756665587425232,
      "num_tokens": 319963195.0,
      "step": 9197
    },
    {
      "epoch": 1.7080779944289692,
      "grad_norm": 1.2857521099409677,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8793660402297974,
      "num_tokens": 320005010.0,
      "step": 9198
    },
    {
      "epoch": 1.708263695450325,
      "grad_norm": 1.4140845857744817,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8708169460296631,
      "num_tokens": 320045966.0,
      "step": 9199
    },
    {
      "epoch": 1.7084493964716807,
      "grad_norm": 1.3839445058057236,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8817340135574341,
      "num_tokens": 320086708.0,
      "step": 9200
    },
    {
      "epoch": 1.7086350974930362,
      "grad_norm": 1.4530998619607471,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8843569755554199,
      "num_tokens": 320122092.0,
      "step": 9201
    },
    {
      "epoch": 1.7088207985143917,
      "grad_norm": 1.5991348884725947,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.870240330696106,
      "num_tokens": 320157293.0,
      "step": 9202
    },
    {
      "epoch": 1.7090064995357475,
      "grad_norm": 1.4898078825036427,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8879125714302063,
      "num_tokens": 320191217.0,
      "step": 9203
    },
    {
      "epoch": 1.7091922005571032,
      "grad_norm": 1.7087885518272563,
      "learning_rate": 1e-06,
      "loss": 0.3743,
      "mean_token_accuracy": 0.8765100240707397,
      "num_tokens": 320219913.0,
      "step": 9204
    },
    {
      "epoch": 1.7093779015784587,
      "grad_norm": 1.4456421957659802,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8844398260116577,
      "num_tokens": 320252917.0,
      "step": 9205
    },
    {
      "epoch": 1.7095636025998142,
      "grad_norm": 1.414166974977971,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8733670711517334,
      "num_tokens": 320292640.0,
      "step": 9206
    },
    {
      "epoch": 1.70974930362117,
      "grad_norm": 1.6425508365496482,
      "learning_rate": 1e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.860686182975769,
      "num_tokens": 320322415.0,
      "step": 9207
    },
    {
      "epoch": 1.7099350046425257,
      "grad_norm": 1.5771267802347029,
      "learning_rate": 1e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8523866534233093,
      "num_tokens": 320363560.0,
      "step": 9208
    },
    {
      "epoch": 1.7101207056638812,
      "grad_norm": 1.480645105283844,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8709259033203125,
      "num_tokens": 320400055.0,
      "step": 9209
    },
    {
      "epoch": 1.7103064066852367,
      "grad_norm": 1.4382649510300305,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8910915851593018,
      "num_tokens": 320435134.0,
      "step": 9210
    },
    {
      "epoch": 1.7104921077065924,
      "grad_norm": 1.4530335766297884,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8730996251106262,
      "num_tokens": 320470380.0,
      "step": 9211
    },
    {
      "epoch": 1.7106778087279482,
      "grad_norm": 1.6332913938550015,
      "learning_rate": 1e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8614720702171326,
      "num_tokens": 320501109.0,
      "step": 9212
    },
    {
      "epoch": 1.7108635097493035,
      "grad_norm": 1.4014546866374853,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.8947314023971558,
      "num_tokens": 320533386.0,
      "step": 9213
    },
    {
      "epoch": 1.7110492107706592,
      "grad_norm": 1.4098628715957917,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8860374689102173,
      "num_tokens": 320567506.0,
      "step": 9214
    },
    {
      "epoch": 1.711234911792015,
      "grad_norm": 1.413282683240423,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8697399497032166,
      "num_tokens": 320606346.0,
      "step": 9215
    },
    {
      "epoch": 1.7114206128133704,
      "grad_norm": 1.5694702191289123,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.870203971862793,
      "num_tokens": 320639817.0,
      "step": 9216
    },
    {
      "epoch": 1.711606313834726,
      "grad_norm": 1.6964454576929626,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8866877555847168,
      "num_tokens": 320664813.0,
      "step": 9217
    },
    {
      "epoch": 1.7117920148560817,
      "grad_norm": 1.5864579678056778,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8698363304138184,
      "num_tokens": 320701503.0,
      "step": 9218
    },
    {
      "epoch": 1.7119777158774374,
      "grad_norm": 1.5598008957954852,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8668067455291748,
      "num_tokens": 320737463.0,
      "step": 9219
    },
    {
      "epoch": 1.712163416898793,
      "grad_norm": 1.4083417739215591,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8717148303985596,
      "num_tokens": 320777252.0,
      "step": 9220
    },
    {
      "epoch": 1.7123491179201484,
      "grad_norm": 1.4724460835569773,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8632686138153076,
      "num_tokens": 320817706.0,
      "step": 9221
    },
    {
      "epoch": 1.7125348189415042,
      "grad_norm": 1.4024704540897228,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8784531950950623,
      "num_tokens": 320854805.0,
      "step": 9222
    },
    {
      "epoch": 1.71272051996286,
      "grad_norm": 1.6758654327423639,
      "learning_rate": 1e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8622367978096008,
      "num_tokens": 320883720.0,
      "step": 9223
    },
    {
      "epoch": 1.7129062209842154,
      "grad_norm": 1.5899226037359393,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8786484003067017,
      "num_tokens": 320919571.0,
      "step": 9224
    },
    {
      "epoch": 1.713091922005571,
      "grad_norm": 1.6166626091527876,
      "learning_rate": 1e-06,
      "loss": 0.3888,
      "mean_token_accuracy": 0.8658797144889832,
      "num_tokens": 320955995.0,
      "step": 9225
    },
    {
      "epoch": 1.7132776230269267,
      "grad_norm": 1.5039845814996784,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8760440349578857,
      "num_tokens": 320991430.0,
      "step": 9226
    },
    {
      "epoch": 1.7134633240482824,
      "grad_norm": 1.5524588025457755,
      "learning_rate": 1e-06,
      "loss": 0.3769,
      "mean_token_accuracy": 0.8697694540023804,
      "num_tokens": 321023305.0,
      "step": 9227
    },
    {
      "epoch": 1.713649025069638,
      "grad_norm": 1.5927970863902707,
      "learning_rate": 1e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8527776598930359,
      "num_tokens": 321058972.0,
      "step": 9228
    },
    {
      "epoch": 1.7138347260909934,
      "grad_norm": 1.5379574686539867,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8808944225311279,
      "num_tokens": 321090148.0,
      "step": 9229
    },
    {
      "epoch": 1.7140204271123491,
      "grad_norm": 1.436796718493825,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8819810748100281,
      "num_tokens": 321129638.0,
      "step": 9230
    },
    {
      "epoch": 1.7142061281337049,
      "grad_norm": 1.5405789042531055,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8743667006492615,
      "num_tokens": 321164097.0,
      "step": 9231
    },
    {
      "epoch": 1.7143918291550604,
      "grad_norm": 1.6939527349981587,
      "learning_rate": 1e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8408393263816833,
      "num_tokens": 321195533.0,
      "step": 9232
    },
    {
      "epoch": 1.714577530176416,
      "grad_norm": 1.5913812068618103,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8858497738838196,
      "num_tokens": 321226480.0,
      "step": 9233
    },
    {
      "epoch": 1.7147632311977716,
      "grad_norm": 1.5917568780810476,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8768696784973145,
      "num_tokens": 321258858.0,
      "step": 9234
    },
    {
      "epoch": 1.7149489322191274,
      "grad_norm": 1.4726917980580363,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8804707527160645,
      "num_tokens": 321290955.0,
      "step": 9235
    },
    {
      "epoch": 1.7151346332404829,
      "grad_norm": 1.4791818597747621,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8785414695739746,
      "num_tokens": 321323376.0,
      "step": 9236
    },
    {
      "epoch": 1.7153203342618384,
      "grad_norm": 1.3997191310269315,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8812200427055359,
      "num_tokens": 321360602.0,
      "step": 9237
    },
    {
      "epoch": 1.7155060352831941,
      "grad_norm": 1.5328570956066638,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8882343769073486,
      "num_tokens": 321397362.0,
      "step": 9238
    },
    {
      "epoch": 1.7156917363045496,
      "grad_norm": 1.6416382665279006,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8641453981399536,
      "num_tokens": 321427556.0,
      "step": 9239
    },
    {
      "epoch": 1.7158774373259051,
      "grad_norm": 1.4993351150546035,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8664119243621826,
      "num_tokens": 321466948.0,
      "step": 9240
    },
    {
      "epoch": 1.7160631383472609,
      "grad_norm": 1.5213584493857415,
      "learning_rate": 1e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8603780269622803,
      "num_tokens": 321505549.0,
      "step": 9241
    },
    {
      "epoch": 1.7162488393686166,
      "grad_norm": 1.434667026731135,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8729315400123596,
      "num_tokens": 321544583.0,
      "step": 9242
    },
    {
      "epoch": 1.7164345403899721,
      "grad_norm": 1.5037160591847374,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8697474002838135,
      "num_tokens": 321579580.0,
      "step": 9243
    },
    {
      "epoch": 1.7166202414113276,
      "grad_norm": 1.5267221053955158,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8841907382011414,
      "num_tokens": 321613231.0,
      "step": 9244
    },
    {
      "epoch": 1.7168059424326834,
      "grad_norm": 1.4618693661722093,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8806547522544861,
      "num_tokens": 321649935.0,
      "step": 9245
    },
    {
      "epoch": 1.716991643454039,
      "grad_norm": 1.3208650298417177,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8960568904876709,
      "num_tokens": 321688869.0,
      "step": 9246
    },
    {
      "epoch": 1.7171773444753946,
      "grad_norm": 1.4794098027860367,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8801361918449402,
      "num_tokens": 321723625.0,
      "step": 9247
    },
    {
      "epoch": 1.7173630454967501,
      "grad_norm": 1.6148115390676203,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8757774829864502,
      "num_tokens": 321750903.0,
      "step": 9248
    },
    {
      "epoch": 1.7175487465181059,
      "grad_norm": 1.5516549290903396,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8876379728317261,
      "num_tokens": 321782528.0,
      "step": 9249
    },
    {
      "epoch": 1.7177344475394616,
      "grad_norm": 1.6327325055853026,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8877235651016235,
      "num_tokens": 321809134.0,
      "step": 9250
    },
    {
      "epoch": 1.717920148560817,
      "grad_norm": 1.4715076242601257,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8724966645240784,
      "num_tokens": 321846099.0,
      "step": 9251
    },
    {
      "epoch": 1.7181058495821726,
      "grad_norm": 1.5389643172328182,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8790229558944702,
      "num_tokens": 321885687.0,
      "step": 9252
    },
    {
      "epoch": 1.7182915506035283,
      "grad_norm": 1.4587423697744977,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8712884187698364,
      "num_tokens": 321922329.0,
      "step": 9253
    },
    {
      "epoch": 1.718477251624884,
      "grad_norm": 1.3898420542003347,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8831026554107666,
      "num_tokens": 321962367.0,
      "step": 9254
    },
    {
      "epoch": 1.7186629526462396,
      "grad_norm": 1.5428633457029017,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8815603256225586,
      "num_tokens": 321993811.0,
      "step": 9255
    },
    {
      "epoch": 1.718848653667595,
      "grad_norm": 1.448477845889884,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8719515800476074,
      "num_tokens": 322032283.0,
      "step": 9256
    },
    {
      "epoch": 1.7190343546889508,
      "grad_norm": 1.425289913750209,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8778018951416016,
      "num_tokens": 322072195.0,
      "step": 9257
    },
    {
      "epoch": 1.7192200557103066,
      "grad_norm": 1.3717961825331282,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8894806504249573,
      "num_tokens": 322110921.0,
      "step": 9258
    },
    {
      "epoch": 1.719405756731662,
      "grad_norm": 1.3715408717784978,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8787742257118225,
      "num_tokens": 322150102.0,
      "step": 9259
    },
    {
      "epoch": 1.7195914577530176,
      "grad_norm": 1.508934061415677,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.878533124923706,
      "num_tokens": 322181860.0,
      "step": 9260
    },
    {
      "epoch": 1.7197771587743733,
      "grad_norm": 1.44068552992105,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8864719271659851,
      "num_tokens": 322218956.0,
      "step": 9261
    },
    {
      "epoch": 1.7199628597957288,
      "grad_norm": 1.5451259758472575,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8819096088409424,
      "num_tokens": 322251424.0,
      "step": 9262
    },
    {
      "epoch": 1.7201485608170843,
      "grad_norm": 1.453528652979066,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.8927775025367737,
      "num_tokens": 322285105.0,
      "step": 9263
    },
    {
      "epoch": 1.72033426183844,
      "grad_norm": 1.4408765753746813,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8933334946632385,
      "num_tokens": 322321285.0,
      "step": 9264
    },
    {
      "epoch": 1.7205199628597958,
      "grad_norm": 1.4302619713312659,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8671084046363831,
      "num_tokens": 322362512.0,
      "step": 9265
    },
    {
      "epoch": 1.7207056638811513,
      "grad_norm": 1.4478084116768106,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8788471221923828,
      "num_tokens": 322399264.0,
      "step": 9266
    },
    {
      "epoch": 1.7208913649025068,
      "grad_norm": 1.4911370664064916,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8719202876091003,
      "num_tokens": 322435792.0,
      "step": 9267
    },
    {
      "epoch": 1.7210770659238626,
      "grad_norm": 1.7573993304715119,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8905770778656006,
      "num_tokens": 322460138.0,
      "step": 9268
    },
    {
      "epoch": 1.7212627669452183,
      "grad_norm": 1.6172374100136668,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.874740481376648,
      "num_tokens": 322490445.0,
      "step": 9269
    },
    {
      "epoch": 1.7214484679665738,
      "grad_norm": 1.431443746403622,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8811755180358887,
      "num_tokens": 322524842.0,
      "step": 9270
    },
    {
      "epoch": 1.7216341689879293,
      "grad_norm": 1.4838715175068355,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8764978647232056,
      "num_tokens": 322558931.0,
      "step": 9271
    },
    {
      "epoch": 1.721819870009285,
      "grad_norm": 1.4265168175819036,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8754435777664185,
      "num_tokens": 322599008.0,
      "step": 9272
    },
    {
      "epoch": 1.7220055710306408,
      "grad_norm": 1.502140297565929,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8864214420318604,
      "num_tokens": 322633139.0,
      "step": 9273
    },
    {
      "epoch": 1.7221912720519963,
      "grad_norm": 1.3620508891173042,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8814578056335449,
      "num_tokens": 322673541.0,
      "step": 9274
    },
    {
      "epoch": 1.7223769730733518,
      "grad_norm": 1.6823242219703862,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8804145455360413,
      "num_tokens": 322706379.0,
      "step": 9275
    },
    {
      "epoch": 1.7225626740947075,
      "grad_norm": 1.3629258106208029,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8873007893562317,
      "num_tokens": 322744162.0,
      "step": 9276
    },
    {
      "epoch": 1.7227483751160633,
      "grad_norm": 1.4389363479904798,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8854119777679443,
      "num_tokens": 322779368.0,
      "step": 9277
    },
    {
      "epoch": 1.7229340761374188,
      "grad_norm": 1.498824302127877,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8830243349075317,
      "num_tokens": 322814760.0,
      "step": 9278
    },
    {
      "epoch": 1.7231197771587743,
      "grad_norm": 1.4912667540523559,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8920657634735107,
      "num_tokens": 322850785.0,
      "step": 9279
    },
    {
      "epoch": 1.72330547818013,
      "grad_norm": 1.4748897591286005,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8773949146270752,
      "num_tokens": 322887439.0,
      "step": 9280
    },
    {
      "epoch": 1.7234911792014858,
      "grad_norm": 1.4317260440248678,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8721534013748169,
      "num_tokens": 322925015.0,
      "step": 9281
    },
    {
      "epoch": 1.7236768802228413,
      "grad_norm": 1.5354855691599216,
      "learning_rate": 1e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8649821281433105,
      "num_tokens": 322963215.0,
      "step": 9282
    },
    {
      "epoch": 1.7238625812441968,
      "grad_norm": 1.4756264772691627,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8805298209190369,
      "num_tokens": 322997723.0,
      "step": 9283
    },
    {
      "epoch": 1.7240482822655525,
      "grad_norm": 1.4028415779754195,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8682218790054321,
      "num_tokens": 323037960.0,
      "step": 9284
    },
    {
      "epoch": 1.724233983286908,
      "grad_norm": 1.487203335137173,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8753691911697388,
      "num_tokens": 323074424.0,
      "step": 9285
    },
    {
      "epoch": 1.7244196843082635,
      "grad_norm": 1.3555509672066086,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8843562602996826,
      "num_tokens": 323111754.0,
      "step": 9286
    },
    {
      "epoch": 1.7246053853296193,
      "grad_norm": 1.4059411335953815,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8916438817977905,
      "num_tokens": 323146601.0,
      "step": 9287
    },
    {
      "epoch": 1.724791086350975,
      "grad_norm": 1.6321400717852452,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8766916990280151,
      "num_tokens": 323177627.0,
      "step": 9288
    },
    {
      "epoch": 1.7249767873723305,
      "grad_norm": 1.339529786282849,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8840019702911377,
      "num_tokens": 323218980.0,
      "step": 9289
    },
    {
      "epoch": 1.725162488393686,
      "grad_norm": 1.4789148219618797,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8629105091094971,
      "num_tokens": 323257952.0,
      "step": 9290
    },
    {
      "epoch": 1.7253481894150418,
      "grad_norm": 1.5205178233829952,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8806968331336975,
      "num_tokens": 323290897.0,
      "step": 9291
    },
    {
      "epoch": 1.7255338904363975,
      "grad_norm": 1.552889722911531,
      "learning_rate": 1e-06,
      "loss": 0.2977,
      "mean_token_accuracy": 0.8958077430725098,
      "num_tokens": 323319112.0,
      "step": 9292
    },
    {
      "epoch": 1.725719591457753,
      "grad_norm": 1.5493371649723322,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8899978399276733,
      "num_tokens": 323349967.0,
      "step": 9293
    },
    {
      "epoch": 1.7259052924791085,
      "grad_norm": 1.4768645039658606,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.890178918838501,
      "num_tokens": 323382800.0,
      "step": 9294
    },
    {
      "epoch": 1.7260909935004642,
      "grad_norm": 1.5138855214741869,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8695183992385864,
      "num_tokens": 323422766.0,
      "step": 9295
    },
    {
      "epoch": 1.72627669452182,
      "grad_norm": 1.6479732431094924,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8723999857902527,
      "num_tokens": 323453710.0,
      "step": 9296
    },
    {
      "epoch": 1.7264623955431755,
      "grad_norm": 1.4699297485440415,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.889137864112854,
      "num_tokens": 323483134.0,
      "step": 9297
    },
    {
      "epoch": 1.726648096564531,
      "grad_norm": 1.4705639986751555,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8809368014335632,
      "num_tokens": 323519121.0,
      "step": 9298
    },
    {
      "epoch": 1.7268337975858867,
      "grad_norm": 1.490124150388511,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8741607666015625,
      "num_tokens": 323557371.0,
      "step": 9299
    },
    {
      "epoch": 1.7270194986072425,
      "grad_norm": 1.4836087706572807,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8823449015617371,
      "num_tokens": 323592734.0,
      "step": 9300
    },
    {
      "epoch": 1.727205199628598,
      "grad_norm": 1.5716250528405238,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8778855800628662,
      "num_tokens": 323620104.0,
      "step": 9301
    },
    {
      "epoch": 1.7273909006499535,
      "grad_norm": 1.5250729512059353,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8827405571937561,
      "num_tokens": 323650470.0,
      "step": 9302
    },
    {
      "epoch": 1.7275766016713092,
      "grad_norm": 1.3752682813249961,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8819625973701477,
      "num_tokens": 323687535.0,
      "step": 9303
    },
    {
      "epoch": 1.727762302692665,
      "grad_norm": 1.5216693185445842,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8827347755432129,
      "num_tokens": 323717896.0,
      "step": 9304
    },
    {
      "epoch": 1.7279480037140205,
      "grad_norm": 1.5462419120174964,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8762795925140381,
      "num_tokens": 323750567.0,
      "step": 9305
    },
    {
      "epoch": 1.728133704735376,
      "grad_norm": 1.5671499375747964,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8791218996047974,
      "num_tokens": 323780297.0,
      "step": 9306
    },
    {
      "epoch": 1.7283194057567317,
      "grad_norm": 1.4570517556872353,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8859540224075317,
      "num_tokens": 323816766.0,
      "step": 9307
    },
    {
      "epoch": 1.7285051067780874,
      "grad_norm": 1.4450766894787053,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8913779258728027,
      "num_tokens": 323847222.0,
      "step": 9308
    },
    {
      "epoch": 1.728690807799443,
      "grad_norm": 1.5011225384109224,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8918368220329285,
      "num_tokens": 323883768.0,
      "step": 9309
    },
    {
      "epoch": 1.7288765088207985,
      "grad_norm": 1.6051494303213185,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8763856291770935,
      "num_tokens": 323914931.0,
      "step": 9310
    },
    {
      "epoch": 1.7290622098421542,
      "grad_norm": 1.5367457687714114,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8713200092315674,
      "num_tokens": 323949194.0,
      "step": 9311
    },
    {
      "epoch": 1.7292479108635097,
      "grad_norm": 1.5037164100383318,
      "learning_rate": 1e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8563700914382935,
      "num_tokens": 323987137.0,
      "step": 9312
    },
    {
      "epoch": 1.7294336118848652,
      "grad_norm": 1.4897370456146541,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.871924877166748,
      "num_tokens": 324026805.0,
      "step": 9313
    },
    {
      "epoch": 1.729619312906221,
      "grad_norm": 1.4415975209134286,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8802365064620972,
      "num_tokens": 324064335.0,
      "step": 9314
    },
    {
      "epoch": 1.7298050139275767,
      "grad_norm": 1.674816809707568,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8775337338447571,
      "num_tokens": 324098343.0,
      "step": 9315
    },
    {
      "epoch": 1.7299907149489322,
      "grad_norm": 1.4014917029900433,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8771127462387085,
      "num_tokens": 324138684.0,
      "step": 9316
    },
    {
      "epoch": 1.7301764159702877,
      "grad_norm": 1.5799255393053513,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8827152252197266,
      "num_tokens": 324166896.0,
      "step": 9317
    },
    {
      "epoch": 1.7303621169916434,
      "grad_norm": 1.467622313565037,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8778675198554993,
      "num_tokens": 324199648.0,
      "step": 9318
    },
    {
      "epoch": 1.7305478180129992,
      "grad_norm": 1.4886510116789484,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.877945065498352,
      "num_tokens": 324235411.0,
      "step": 9319
    },
    {
      "epoch": 1.7307335190343547,
      "grad_norm": 1.6250998751000307,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.874657392501831,
      "num_tokens": 324267534.0,
      "step": 9320
    },
    {
      "epoch": 1.7309192200557102,
      "grad_norm": 1.6140888605249921,
      "learning_rate": 1e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.867974042892456,
      "num_tokens": 324298544.0,
      "step": 9321
    },
    {
      "epoch": 1.731104921077066,
      "grad_norm": 1.535707650716669,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8789083957672119,
      "num_tokens": 324331208.0,
      "step": 9322
    },
    {
      "epoch": 1.7312906220984217,
      "grad_norm": 1.4742707887572915,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8778945803642273,
      "num_tokens": 324364417.0,
      "step": 9323
    },
    {
      "epoch": 1.7314763231197772,
      "grad_norm": 1.6534315072451347,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8713275194168091,
      "num_tokens": 324394506.0,
      "step": 9324
    },
    {
      "epoch": 1.7316620241411327,
      "grad_norm": 1.4830191470723548,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8733953237533569,
      "num_tokens": 324429349.0,
      "step": 9325
    },
    {
      "epoch": 1.7318477251624884,
      "grad_norm": 1.5202279824806784,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8874554634094238,
      "num_tokens": 324462333.0,
      "step": 9326
    },
    {
      "epoch": 1.7320334261838441,
      "grad_norm": 1.6075343230496204,
      "learning_rate": 1e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8618778586387634,
      "num_tokens": 324494651.0,
      "step": 9327
    },
    {
      "epoch": 1.7322191272051997,
      "grad_norm": 1.5121086575840932,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8830623030662537,
      "num_tokens": 324528676.0,
      "step": 9328
    },
    {
      "epoch": 1.7324048282265552,
      "grad_norm": 1.6208742911242036,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8760637044906616,
      "num_tokens": 324557905.0,
      "step": 9329
    },
    {
      "epoch": 1.732590529247911,
      "grad_norm": 1.64102751900077,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.8943480849266052,
      "num_tokens": 324586226.0,
      "step": 9330
    },
    {
      "epoch": 1.7327762302692666,
      "grad_norm": 1.4876779069449406,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8810231685638428,
      "num_tokens": 324620634.0,
      "step": 9331
    },
    {
      "epoch": 1.7329619312906221,
      "grad_norm": 1.5460643771191451,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8731727600097656,
      "num_tokens": 324653113.0,
      "step": 9332
    },
    {
      "epoch": 1.7331476323119777,
      "grad_norm": 1.4338351625527797,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8951380252838135,
      "num_tokens": 324687811.0,
      "step": 9333
    },
    {
      "epoch": 1.7333333333333334,
      "grad_norm": 1.4020690107786786,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8852030038833618,
      "num_tokens": 324723258.0,
      "step": 9334
    },
    {
      "epoch": 1.733519034354689,
      "grad_norm": 1.5732368450558556,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8736870884895325,
      "num_tokens": 324756195.0,
      "step": 9335
    },
    {
      "epoch": 1.7337047353760444,
      "grad_norm": 1.4844741061582565,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8702960014343262,
      "num_tokens": 324796114.0,
      "step": 9336
    },
    {
      "epoch": 1.7338904363974001,
      "grad_norm": 1.4426231761299908,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8826122283935547,
      "num_tokens": 324834432.0,
      "step": 9337
    },
    {
      "epoch": 1.7340761374187559,
      "grad_norm": 1.4509602158000368,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8849481344223022,
      "num_tokens": 324869141.0,
      "step": 9338
    },
    {
      "epoch": 1.7342618384401114,
      "grad_norm": 1.4877328824466496,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8680229783058167,
      "num_tokens": 324909057.0,
      "step": 9339
    },
    {
      "epoch": 1.734447539461467,
      "grad_norm": 1.4428797299348455,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8898208141326904,
      "num_tokens": 324945991.0,
      "step": 9340
    },
    {
      "epoch": 1.7346332404828226,
      "grad_norm": 1.3992459005987075,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8856749534606934,
      "num_tokens": 324983406.0,
      "step": 9341
    },
    {
      "epoch": 1.7348189415041784,
      "grad_norm": 1.4814837942457462,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8761197328567505,
      "num_tokens": 325018075.0,
      "step": 9342
    },
    {
      "epoch": 1.7350046425255339,
      "grad_norm": 1.5149585839423299,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8903575539588928,
      "num_tokens": 325050313.0,
      "step": 9343
    },
    {
      "epoch": 1.7351903435468894,
      "grad_norm": 1.4322438056160665,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8739712238311768,
      "num_tokens": 325088328.0,
      "step": 9344
    },
    {
      "epoch": 1.7353760445682451,
      "grad_norm": 1.5296208546233558,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8736701011657715,
      "num_tokens": 325120911.0,
      "step": 9345
    },
    {
      "epoch": 1.7355617455896009,
      "grad_norm": 1.507340614335204,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8878254890441895,
      "num_tokens": 325152317.0,
      "step": 9346
    },
    {
      "epoch": 1.7357474466109564,
      "grad_norm": 1.455703955637507,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8764934539794922,
      "num_tokens": 325189335.0,
      "step": 9347
    },
    {
      "epoch": 1.7359331476323119,
      "grad_norm": 1.3465633655989682,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8729060888290405,
      "num_tokens": 325232544.0,
      "step": 9348
    },
    {
      "epoch": 1.7361188486536676,
      "grad_norm": 1.5065958703621332,
      "learning_rate": 1e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8569843769073486,
      "num_tokens": 325270174.0,
      "step": 9349
    },
    {
      "epoch": 1.7363045496750233,
      "grad_norm": 1.5335214518312337,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8841575384140015,
      "num_tokens": 325302811.0,
      "step": 9350
    },
    {
      "epoch": 1.7364902506963789,
      "grad_norm": 1.4687442247003264,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8790417909622192,
      "num_tokens": 325341467.0,
      "step": 9351
    },
    {
      "epoch": 1.7366759517177344,
      "grad_norm": 1.4507154136051863,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8777531981468201,
      "num_tokens": 325379420.0,
      "step": 9352
    },
    {
      "epoch": 1.73686165273909,
      "grad_norm": 2.3166967990873317,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.872778058052063,
      "num_tokens": 325412553.0,
      "step": 9353
    },
    {
      "epoch": 1.7370473537604458,
      "grad_norm": 1.6631208327536866,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8742367029190063,
      "num_tokens": 325439261.0,
      "step": 9354
    },
    {
      "epoch": 1.7372330547818013,
      "grad_norm": 1.4484844503525258,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8800661563873291,
      "num_tokens": 325475397.0,
      "step": 9355
    },
    {
      "epoch": 1.7374187558031569,
      "grad_norm": 1.478234314834368,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8809502124786377,
      "num_tokens": 325512237.0,
      "step": 9356
    },
    {
      "epoch": 1.7376044568245126,
      "grad_norm": 1.4045426994527075,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8773113489151001,
      "num_tokens": 325554199.0,
      "step": 9357
    },
    {
      "epoch": 1.737790157845868,
      "grad_norm": 1.7990607370307143,
      "learning_rate": 1e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8675116300582886,
      "num_tokens": 325584090.0,
      "step": 9358
    },
    {
      "epoch": 1.7379758588672236,
      "grad_norm": 1.3950916875544155,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8665336966514587,
      "num_tokens": 325624279.0,
      "step": 9359
    },
    {
      "epoch": 1.7381615598885793,
      "grad_norm": 1.325454708743366,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8865669965744019,
      "num_tokens": 325665188.0,
      "step": 9360
    },
    {
      "epoch": 1.738347260909935,
      "grad_norm": 1.4093872012875008,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.8893199563026428,
      "num_tokens": 325699573.0,
      "step": 9361
    },
    {
      "epoch": 1.7385329619312906,
      "grad_norm": 1.31072236578006,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8875565528869629,
      "num_tokens": 325738596.0,
      "step": 9362
    },
    {
      "epoch": 1.738718662952646,
      "grad_norm": 1.390814654563029,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8738239407539368,
      "num_tokens": 325779321.0,
      "step": 9363
    },
    {
      "epoch": 1.7389043639740018,
      "grad_norm": 1.4181202533682955,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.872066080570221,
      "num_tokens": 325818070.0,
      "step": 9364
    },
    {
      "epoch": 1.7390900649953576,
      "grad_norm": 1.3831374062476034,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8899108171463013,
      "num_tokens": 325855717.0,
      "step": 9365
    },
    {
      "epoch": 1.739275766016713,
      "grad_norm": 1.498406859038996,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8727348446846008,
      "num_tokens": 325892710.0,
      "step": 9366
    },
    {
      "epoch": 1.7394614670380686,
      "grad_norm": 1.5253395832341252,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8825880289077759,
      "num_tokens": 325926715.0,
      "step": 9367
    },
    {
      "epoch": 1.7396471680594243,
      "grad_norm": 1.4063690471101868,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.885048508644104,
      "num_tokens": 325963082.0,
      "step": 9368
    },
    {
      "epoch": 1.73983286908078,
      "grad_norm": 1.5094964490097837,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.873197078704834,
      "num_tokens": 326000537.0,
      "step": 9369
    },
    {
      "epoch": 1.7400185701021356,
      "grad_norm": 1.4017348223510744,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.880390465259552,
      "num_tokens": 326040238.0,
      "step": 9370
    },
    {
      "epoch": 1.740204271123491,
      "grad_norm": 1.4825817437080382,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8745905160903931,
      "num_tokens": 326076046.0,
      "step": 9371
    },
    {
      "epoch": 1.7403899721448468,
      "grad_norm": 1.559267403699871,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8788290023803711,
      "num_tokens": 326108708.0,
      "step": 9372
    },
    {
      "epoch": 1.7405756731662025,
      "grad_norm": 1.4855503000849515,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8836239576339722,
      "num_tokens": 326142054.0,
      "step": 9373
    },
    {
      "epoch": 1.740761374187558,
      "grad_norm": 1.4921831850771972,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8720396757125854,
      "num_tokens": 326177025.0,
      "step": 9374
    },
    {
      "epoch": 1.7409470752089136,
      "grad_norm": 1.4004741943089243,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8727685809135437,
      "num_tokens": 326217362.0,
      "step": 9375
    },
    {
      "epoch": 1.7411327762302693,
      "grad_norm": 1.5766290729871508,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8737872838973999,
      "num_tokens": 326249187.0,
      "step": 9376
    },
    {
      "epoch": 1.741318477251625,
      "grad_norm": 1.4819776961473934,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8711075782775879,
      "num_tokens": 326286950.0,
      "step": 9377
    },
    {
      "epoch": 1.7415041782729805,
      "grad_norm": 1.5272869709070664,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8747824430465698,
      "num_tokens": 326325024.0,
      "step": 9378
    },
    {
      "epoch": 1.741689879294336,
      "grad_norm": 1.5815551819263847,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8818877935409546,
      "num_tokens": 326356309.0,
      "step": 9379
    },
    {
      "epoch": 1.7418755803156918,
      "grad_norm": 1.5105766217808514,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.877578854560852,
      "num_tokens": 326389251.0,
      "step": 9380
    },
    {
      "epoch": 1.7420612813370475,
      "grad_norm": 1.5210740682761201,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8775146007537842,
      "num_tokens": 326422129.0,
      "step": 9381
    },
    {
      "epoch": 1.7422469823584028,
      "grad_norm": 1.457004423566716,
      "learning_rate": 1e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8652670383453369,
      "num_tokens": 326460609.0,
      "step": 9382
    },
    {
      "epoch": 1.7424326833797585,
      "grad_norm": 1.6432488906475722,
      "learning_rate": 1e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8669880628585815,
      "num_tokens": 326492793.0,
      "step": 9383
    },
    {
      "epoch": 1.7426183844011143,
      "grad_norm": 1.455923160154421,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8737404346466064,
      "num_tokens": 326531339.0,
      "step": 9384
    },
    {
      "epoch": 1.7428040854224698,
      "grad_norm": 1.5994714372249286,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8764144778251648,
      "num_tokens": 326560763.0,
      "step": 9385
    },
    {
      "epoch": 1.7429897864438253,
      "grad_norm": 1.4290962552340256,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8675613403320312,
      "num_tokens": 326597349.0,
      "step": 9386
    },
    {
      "epoch": 1.743175487465181,
      "grad_norm": 1.4941917675920675,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8860316276550293,
      "num_tokens": 326630894.0,
      "step": 9387
    },
    {
      "epoch": 1.7433611884865368,
      "grad_norm": 1.5433560240935762,
      "learning_rate": 1e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8641381859779358,
      "num_tokens": 326664894.0,
      "step": 9388
    },
    {
      "epoch": 1.7435468895078923,
      "grad_norm": 1.2967251340629145,
      "learning_rate": 1e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.8707526922225952,
      "num_tokens": 326711086.0,
      "step": 9389
    },
    {
      "epoch": 1.7437325905292478,
      "grad_norm": 1.7405362587392046,
      "learning_rate": 1e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8691021800041199,
      "num_tokens": 326736929.0,
      "step": 9390
    },
    {
      "epoch": 1.7439182915506035,
      "grad_norm": 1.5451356052748442,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.87510746717453,
      "num_tokens": 326767223.0,
      "step": 9391
    },
    {
      "epoch": 1.7441039925719592,
      "grad_norm": 1.6018958743749283,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8764774799346924,
      "num_tokens": 326801381.0,
      "step": 9392
    },
    {
      "epoch": 1.7442896935933148,
      "grad_norm": 1.5977187728858107,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8713398575782776,
      "num_tokens": 326834542.0,
      "step": 9393
    },
    {
      "epoch": 1.7444753946146703,
      "grad_norm": 1.5433791681782072,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8788133859634399,
      "num_tokens": 326868029.0,
      "step": 9394
    },
    {
      "epoch": 1.744661095636026,
      "grad_norm": 1.5357541157358663,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8732101917266846,
      "num_tokens": 326906981.0,
      "step": 9395
    },
    {
      "epoch": 1.7448467966573817,
      "grad_norm": 1.5122924149662536,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8782302737236023,
      "num_tokens": 326941122.0,
      "step": 9396
    },
    {
      "epoch": 1.7450324976787372,
      "grad_norm": 1.5678645131348863,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8757435083389282,
      "num_tokens": 326977910.0,
      "step": 9397
    },
    {
      "epoch": 1.7452181987000928,
      "grad_norm": 1.6278869062419719,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8712272047996521,
      "num_tokens": 327011098.0,
      "step": 9398
    },
    {
      "epoch": 1.7454038997214485,
      "grad_norm": 1.4623791930638126,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8759413361549377,
      "num_tokens": 327048074.0,
      "step": 9399
    },
    {
      "epoch": 1.7455896007428042,
      "grad_norm": 1.4788132695911507,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8823674321174622,
      "num_tokens": 327086455.0,
      "step": 9400
    },
    {
      "epoch": 1.7457753017641597,
      "grad_norm": 1.5102214290643463,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8864195942878723,
      "num_tokens": 327120799.0,
      "step": 9401
    },
    {
      "epoch": 1.7459610027855152,
      "grad_norm": 1.5568232212617503,
      "learning_rate": 1e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.8744673728942871,
      "num_tokens": 327154326.0,
      "step": 9402
    },
    {
      "epoch": 1.746146703806871,
      "grad_norm": 1.6333274234716948,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8792216777801514,
      "num_tokens": 327185162.0,
      "step": 9403
    },
    {
      "epoch": 1.7463324048282267,
      "grad_norm": 1.5603149173962527,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8738595843315125,
      "num_tokens": 327217822.0,
      "step": 9404
    },
    {
      "epoch": 1.7465181058495822,
      "grad_norm": 1.3759107368213972,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8849555253982544,
      "num_tokens": 327252971.0,
      "step": 9405
    },
    {
      "epoch": 1.7467038068709377,
      "grad_norm": 1.408905906139336,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8860297203063965,
      "num_tokens": 327289183.0,
      "step": 9406
    },
    {
      "epoch": 1.7468895078922935,
      "grad_norm": 1.7249269040246855,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8703187704086304,
      "num_tokens": 327317065.0,
      "step": 9407
    },
    {
      "epoch": 1.747075208913649,
      "grad_norm": 1.527668140165602,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8847379088401794,
      "num_tokens": 327351167.0,
      "step": 9408
    },
    {
      "epoch": 1.7472609099350045,
      "grad_norm": 1.553209997517776,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8770522475242615,
      "num_tokens": 327383336.0,
      "step": 9409
    },
    {
      "epoch": 1.7474466109563602,
      "grad_norm": 1.6334132075796541,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8767578601837158,
      "num_tokens": 327411782.0,
      "step": 9410
    },
    {
      "epoch": 1.747632311977716,
      "grad_norm": 1.3975957939507813,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.8961137533187866,
      "num_tokens": 327447669.0,
      "step": 9411
    },
    {
      "epoch": 1.7478180129990715,
      "grad_norm": 1.7780189421797277,
      "learning_rate": 1e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.8662590980529785,
      "num_tokens": 327472792.0,
      "step": 9412
    },
    {
      "epoch": 1.748003714020427,
      "grad_norm": 1.6338642910140653,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8728185892105103,
      "num_tokens": 327506450.0,
      "step": 9413
    },
    {
      "epoch": 1.7481894150417827,
      "grad_norm": 1.3862791901650093,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8969114422798157,
      "num_tokens": 327542124.0,
      "step": 9414
    },
    {
      "epoch": 1.7483751160631384,
      "grad_norm": 1.5212394575749595,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8812711238861084,
      "num_tokens": 327576322.0,
      "step": 9415
    },
    {
      "epoch": 1.748560817084494,
      "grad_norm": 1.3900446141145215,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8848588466644287,
      "num_tokens": 327615239.0,
      "step": 9416
    },
    {
      "epoch": 1.7487465181058495,
      "grad_norm": 1.4755851025284985,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8682392239570618,
      "num_tokens": 327651790.0,
      "step": 9417
    },
    {
      "epoch": 1.7489322191272052,
      "grad_norm": 1.6784999553092754,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8719164133071899,
      "num_tokens": 327681258.0,
      "step": 9418
    },
    {
      "epoch": 1.749117920148561,
      "grad_norm": 1.5971096519606525,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8796074390411377,
      "num_tokens": 327713779.0,
      "step": 9419
    },
    {
      "epoch": 1.7493036211699164,
      "grad_norm": 1.418455107421457,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8840523958206177,
      "num_tokens": 327749663.0,
      "step": 9420
    },
    {
      "epoch": 1.749489322191272,
      "grad_norm": 1.4204096568232467,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8700384497642517,
      "num_tokens": 327789522.0,
      "step": 9421
    },
    {
      "epoch": 1.7496750232126277,
      "grad_norm": 1.5744815139644588,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8709619641304016,
      "num_tokens": 327820256.0,
      "step": 9422
    },
    {
      "epoch": 1.7498607242339834,
      "grad_norm": 1.5635903049867075,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8763171434402466,
      "num_tokens": 327853753.0,
      "step": 9423
    },
    {
      "epoch": 1.750046425255339,
      "grad_norm": 1.4983286309582808,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8828531503677368,
      "num_tokens": 327886623.0,
      "step": 9424
    },
    {
      "epoch": 1.7502321262766944,
      "grad_norm": 1.4997529429358214,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8698300123214722,
      "num_tokens": 327926119.0,
      "step": 9425
    },
    {
      "epoch": 1.7504178272980502,
      "grad_norm": 1.3500394512638372,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8742896318435669,
      "num_tokens": 327969123.0,
      "step": 9426
    },
    {
      "epoch": 1.750603528319406,
      "grad_norm": 1.4370395645758947,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8929271697998047,
      "num_tokens": 328001660.0,
      "step": 9427
    },
    {
      "epoch": 1.7507892293407614,
      "grad_norm": 1.5095039507855228,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8744362592697144,
      "num_tokens": 328033465.0,
      "step": 9428
    },
    {
      "epoch": 1.750974930362117,
      "grad_norm": 1.4531854956967964,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8871748447418213,
      "num_tokens": 328071563.0,
      "step": 9429
    },
    {
      "epoch": 1.7511606313834727,
      "grad_norm": 1.4243077341014057,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8880411386489868,
      "num_tokens": 328108662.0,
      "step": 9430
    },
    {
      "epoch": 1.7513463324048282,
      "grad_norm": 1.5475775005774466,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8570194244384766,
      "num_tokens": 328144028.0,
      "step": 9431
    },
    {
      "epoch": 1.7515320334261837,
      "grad_norm": 1.3779386012937567,
      "learning_rate": 1e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8738758563995361,
      "num_tokens": 328185288.0,
      "step": 9432
    },
    {
      "epoch": 1.7517177344475394,
      "grad_norm": 1.4243595822186295,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8766288757324219,
      "num_tokens": 328222492.0,
      "step": 9433
    },
    {
      "epoch": 1.7519034354688952,
      "grad_norm": 1.4411225759823525,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8879002332687378,
      "num_tokens": 328261512.0,
      "step": 9434
    },
    {
      "epoch": 1.7520891364902507,
      "grad_norm": 1.3887216763303833,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8838137984275818,
      "num_tokens": 328300973.0,
      "step": 9435
    },
    {
      "epoch": 1.7522748375116062,
      "grad_norm": 1.5032001332011875,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8788565397262573,
      "num_tokens": 328335555.0,
      "step": 9436
    },
    {
      "epoch": 1.752460538532962,
      "grad_norm": 1.4189147217959326,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8970963358879089,
      "num_tokens": 328370653.0,
      "step": 9437
    },
    {
      "epoch": 1.7526462395543176,
      "grad_norm": 1.5455709826022381,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8870279788970947,
      "num_tokens": 328401039.0,
      "step": 9438
    },
    {
      "epoch": 1.7528319405756732,
      "grad_norm": 1.656936096270452,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8777087330818176,
      "num_tokens": 328427783.0,
      "step": 9439
    },
    {
      "epoch": 1.7530176415970287,
      "grad_norm": 1.6117546950556203,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8816052675247192,
      "num_tokens": 328460638.0,
      "step": 9440
    },
    {
      "epoch": 1.7532033426183844,
      "grad_norm": 1.6766876223881868,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8734538555145264,
      "num_tokens": 328489841.0,
      "step": 9441
    },
    {
      "epoch": 1.7533890436397401,
      "grad_norm": 1.5125318058467592,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8862206935882568,
      "num_tokens": 328525025.0,
      "step": 9442
    },
    {
      "epoch": 1.7535747446610956,
      "grad_norm": 1.5221773693095817,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8728183507919312,
      "num_tokens": 328558583.0,
      "step": 9443
    },
    {
      "epoch": 1.7537604456824512,
      "grad_norm": 1.5021313753521919,
      "learning_rate": 1e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8598533272743225,
      "num_tokens": 328597502.0,
      "step": 9444
    },
    {
      "epoch": 1.7539461467038069,
      "grad_norm": 1.4356100791854134,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8813208341598511,
      "num_tokens": 328636255.0,
      "step": 9445
    },
    {
      "epoch": 1.7541318477251626,
      "grad_norm": 1.4615647883993748,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8748593926429749,
      "num_tokens": 328672539.0,
      "step": 9446
    },
    {
      "epoch": 1.7543175487465181,
      "grad_norm": 1.4696903912252475,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8738486766815186,
      "num_tokens": 328710663.0,
      "step": 9447
    },
    {
      "epoch": 1.7545032497678736,
      "grad_norm": 1.6033093605840179,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8581308722496033,
      "num_tokens": 328742903.0,
      "step": 9448
    },
    {
      "epoch": 1.7546889507892294,
      "grad_norm": 1.5111773697977888,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8764526844024658,
      "num_tokens": 328776280.0,
      "step": 9449
    },
    {
      "epoch": 1.754874651810585,
      "grad_norm": 1.4339666942538885,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.87404465675354,
      "num_tokens": 328814625.0,
      "step": 9450
    },
    {
      "epoch": 1.7550603528319406,
      "grad_norm": 1.700103359334923,
      "learning_rate": 1e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8705859184265137,
      "num_tokens": 328842155.0,
      "step": 9451
    },
    {
      "epoch": 1.7552460538532961,
      "grad_norm": 1.392531102894173,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8786939382553101,
      "num_tokens": 328880010.0,
      "step": 9452
    },
    {
      "epoch": 1.7554317548746519,
      "grad_norm": 1.4348220902837252,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8825087547302246,
      "num_tokens": 328915431.0,
      "step": 9453
    },
    {
      "epoch": 1.7556174558960074,
      "grad_norm": 1.831316391755313,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8737994432449341,
      "num_tokens": 328943311.0,
      "step": 9454
    },
    {
      "epoch": 1.7558031569173629,
      "grad_norm": 1.3529838098441032,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8930467367172241,
      "num_tokens": 328979544.0,
      "step": 9455
    },
    {
      "epoch": 1.7559888579387186,
      "grad_norm": 1.5120475692492708,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.878345251083374,
      "num_tokens": 329012264.0,
      "step": 9456
    },
    {
      "epoch": 1.7561745589600744,
      "grad_norm": 1.5800735479797765,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8834781646728516,
      "num_tokens": 329041977.0,
      "step": 9457
    },
    {
      "epoch": 1.7563602599814299,
      "grad_norm": 1.597736939166367,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8713024854660034,
      "num_tokens": 329076837.0,
      "step": 9458
    },
    {
      "epoch": 1.7565459610027854,
      "grad_norm": 1.3062519022249914,
      "learning_rate": 1e-06,
      "loss": 0.2904,
      "mean_token_accuracy": 0.8969587087631226,
      "num_tokens": 329114195.0,
      "step": 9459
    },
    {
      "epoch": 1.756731662024141,
      "grad_norm": 1.4926776235244505,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8767771124839783,
      "num_tokens": 329153553.0,
      "step": 9460
    },
    {
      "epoch": 1.7569173630454968,
      "grad_norm": 1.4466943376548715,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8717650175094604,
      "num_tokens": 329191390.0,
      "step": 9461
    },
    {
      "epoch": 1.7571030640668523,
      "grad_norm": 1.5120001173777058,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8833770751953125,
      "num_tokens": 329223380.0,
      "step": 9462
    },
    {
      "epoch": 1.7572887650882079,
      "grad_norm": 1.3983156584027867,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8871006965637207,
      "num_tokens": 329260348.0,
      "step": 9463
    },
    {
      "epoch": 1.7574744661095636,
      "grad_norm": 1.5478164828552563,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8759405612945557,
      "num_tokens": 329292446.0,
      "step": 9464
    },
    {
      "epoch": 1.7576601671309193,
      "grad_norm": 1.5240407821006714,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.869994044303894,
      "num_tokens": 329327172.0,
      "step": 9465
    },
    {
      "epoch": 1.7578458681522748,
      "grad_norm": 1.5077987100121364,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.877946138381958,
      "num_tokens": 329361299.0,
      "step": 9466
    },
    {
      "epoch": 1.7580315691736303,
      "grad_norm": 1.585159906545765,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8817299604415894,
      "num_tokens": 329394554.0,
      "step": 9467
    },
    {
      "epoch": 1.758217270194986,
      "grad_norm": 1.4671494421986244,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8815288543701172,
      "num_tokens": 329429602.0,
      "step": 9468
    },
    {
      "epoch": 1.7584029712163418,
      "grad_norm": 1.5208066109883291,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8830695152282715,
      "num_tokens": 329462704.0,
      "step": 9469
    },
    {
      "epoch": 1.7585886722376973,
      "grad_norm": 1.4975865450330992,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8731793165206909,
      "num_tokens": 329498756.0,
      "step": 9470
    },
    {
      "epoch": 1.7587743732590528,
      "grad_norm": 1.445646720971985,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8851717114448547,
      "num_tokens": 329536607.0,
      "step": 9471
    },
    {
      "epoch": 1.7589600742804086,
      "grad_norm": 1.483207885020412,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8698012828826904,
      "num_tokens": 329577403.0,
      "step": 9472
    },
    {
      "epoch": 1.7591457753017643,
      "grad_norm": 1.5837521868552897,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8733041286468506,
      "num_tokens": 329610230.0,
      "step": 9473
    },
    {
      "epoch": 1.7593314763231198,
      "grad_norm": 1.5257220780030278,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8718329668045044,
      "num_tokens": 329643676.0,
      "step": 9474
    },
    {
      "epoch": 1.7595171773444753,
      "grad_norm": 1.5233823339344925,
      "learning_rate": 1e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8765820264816284,
      "num_tokens": 329677603.0,
      "step": 9475
    },
    {
      "epoch": 1.759702878365831,
      "grad_norm": 1.495234670384333,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8764181733131409,
      "num_tokens": 329712898.0,
      "step": 9476
    },
    {
      "epoch": 1.7598885793871868,
      "grad_norm": 1.568638570391089,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8722012042999268,
      "num_tokens": 329745528.0,
      "step": 9477
    },
    {
      "epoch": 1.7600742804085423,
      "grad_norm": 1.5306358670185753,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8647609353065491,
      "num_tokens": 329782817.0,
      "step": 9478
    },
    {
      "epoch": 1.7602599814298978,
      "grad_norm": 1.71901776904134,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8821103572845459,
      "num_tokens": 329808878.0,
      "step": 9479
    },
    {
      "epoch": 1.7604456824512535,
      "grad_norm": 1.3523222132955832,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8778905868530273,
      "num_tokens": 329849840.0,
      "step": 9480
    },
    {
      "epoch": 1.760631383472609,
      "grad_norm": 1.5199775049933253,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8716977834701538,
      "num_tokens": 329883491.0,
      "step": 9481
    },
    {
      "epoch": 1.7608170844939646,
      "grad_norm": 1.4481382141846864,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8826169371604919,
      "num_tokens": 329919938.0,
      "step": 9482
    },
    {
      "epoch": 1.7610027855153203,
      "grad_norm": 1.4344719096407632,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.87562495470047,
      "num_tokens": 329955087.0,
      "step": 9483
    },
    {
      "epoch": 1.761188486536676,
      "grad_norm": 1.5843930214313848,
      "learning_rate": 1e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8648428916931152,
      "num_tokens": 329992765.0,
      "step": 9484
    },
    {
      "epoch": 1.7613741875580315,
      "grad_norm": 1.4107322497212613,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8838366270065308,
      "num_tokens": 330029091.0,
      "step": 9485
    },
    {
      "epoch": 1.761559888579387,
      "grad_norm": 1.4304127648982063,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8755171298980713,
      "num_tokens": 330071099.0,
      "step": 9486
    },
    {
      "epoch": 1.7617455896007428,
      "grad_norm": 1.4846619251219988,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8689168095588684,
      "num_tokens": 330106866.0,
      "step": 9487
    },
    {
      "epoch": 1.7619312906220985,
      "grad_norm": 1.500560709436622,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8831236362457275,
      "num_tokens": 330140069.0,
      "step": 9488
    },
    {
      "epoch": 1.762116991643454,
      "grad_norm": 1.3364628069274729,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8924346566200256,
      "num_tokens": 330179920.0,
      "step": 9489
    },
    {
      "epoch": 1.7623026926648095,
      "grad_norm": 1.3969596858196054,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8886209726333618,
      "num_tokens": 330218294.0,
      "step": 9490
    },
    {
      "epoch": 1.7624883936861653,
      "grad_norm": 1.4315469985771163,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8764090538024902,
      "num_tokens": 330255069.0,
      "step": 9491
    },
    {
      "epoch": 1.762674094707521,
      "grad_norm": 1.4881555826079444,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8712793588638306,
      "num_tokens": 330290724.0,
      "step": 9492
    },
    {
      "epoch": 1.7628597957288765,
      "grad_norm": 1.3923346258682014,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8793637156486511,
      "num_tokens": 330329563.0,
      "step": 9493
    },
    {
      "epoch": 1.763045496750232,
      "grad_norm": 1.4094994770226454,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8934468030929565,
      "num_tokens": 330370986.0,
      "step": 9494
    },
    {
      "epoch": 1.7632311977715878,
      "grad_norm": 1.5712054759909269,
      "learning_rate": 1e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8722844123840332,
      "num_tokens": 330406159.0,
      "step": 9495
    },
    {
      "epoch": 1.7634168987929435,
      "grad_norm": 1.491748980265393,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8826545476913452,
      "num_tokens": 330443612.0,
      "step": 9496
    },
    {
      "epoch": 1.763602599814299,
      "grad_norm": 1.4777678590975096,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8802752494812012,
      "num_tokens": 330481858.0,
      "step": 9497
    },
    {
      "epoch": 1.7637883008356545,
      "grad_norm": 1.4578824022596892,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8819069862365723,
      "num_tokens": 330515732.0,
      "step": 9498
    },
    {
      "epoch": 1.7639740018570103,
      "grad_norm": 1.5735155313099185,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8644545078277588,
      "num_tokens": 330547023.0,
      "step": 9499
    },
    {
      "epoch": 1.764159702878366,
      "grad_norm": 1.5145274309498125,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8856722116470337,
      "num_tokens": 330578778.0,
      "step": 9500
    },
    {
      "epoch": 1.7643454038997215,
      "grad_norm": 1.5538906310733214,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8766594529151917,
      "num_tokens": 330614075.0,
      "step": 9501
    },
    {
      "epoch": 1.764531104921077,
      "grad_norm": 1.411191866417447,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.86632239818573,
      "num_tokens": 330651533.0,
      "step": 9502
    },
    {
      "epoch": 1.7647168059424327,
      "grad_norm": 1.4021683638617315,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.873652994632721,
      "num_tokens": 330693032.0,
      "step": 9503
    },
    {
      "epoch": 1.7649025069637883,
      "grad_norm": 1.4073383389659402,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8934919834136963,
      "num_tokens": 330730186.0,
      "step": 9504
    },
    {
      "epoch": 1.7650882079851438,
      "grad_norm": 1.5559095396723313,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8817906379699707,
      "num_tokens": 330762051.0,
      "step": 9505
    },
    {
      "epoch": 1.7652739090064995,
      "grad_norm": 1.3869954766312238,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8807839155197144,
      "num_tokens": 330801703.0,
      "step": 9506
    },
    {
      "epoch": 1.7654596100278552,
      "grad_norm": 1.4387387436717423,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8725379109382629,
      "num_tokens": 330837585.0,
      "step": 9507
    },
    {
      "epoch": 1.7656453110492107,
      "grad_norm": 1.5514539568831978,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8664520978927612,
      "num_tokens": 330873507.0,
      "step": 9508
    },
    {
      "epoch": 1.7658310120705663,
      "grad_norm": 1.4295841424173594,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8838362693786621,
      "num_tokens": 330907695.0,
      "step": 9509
    },
    {
      "epoch": 1.766016713091922,
      "grad_norm": 1.4949548796880934,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8786328434944153,
      "num_tokens": 330941871.0,
      "step": 9510
    },
    {
      "epoch": 1.7662024141132777,
      "grad_norm": 1.5261743587108707,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8697563409805298,
      "num_tokens": 330974164.0,
      "step": 9511
    },
    {
      "epoch": 1.7663881151346332,
      "grad_norm": 1.5114160723486922,
      "learning_rate": 1e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8546291589736938,
      "num_tokens": 331012446.0,
      "step": 9512
    },
    {
      "epoch": 1.7665738161559887,
      "grad_norm": 1.7559169388099036,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8712642788887024,
      "num_tokens": 331040680.0,
      "step": 9513
    },
    {
      "epoch": 1.7667595171773445,
      "grad_norm": 1.560006956625608,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8746370673179626,
      "num_tokens": 331074650.0,
      "step": 9514
    },
    {
      "epoch": 1.7669452181987002,
      "grad_norm": 1.4127358599879358,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8742948174476624,
      "num_tokens": 331110685.0,
      "step": 9515
    },
    {
      "epoch": 1.7671309192200557,
      "grad_norm": 1.549340773399163,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8761446475982666,
      "num_tokens": 331143061.0,
      "step": 9516
    },
    {
      "epoch": 1.7673166202414112,
      "grad_norm": 1.5416891226513811,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.8765161037445068,
      "num_tokens": 331175815.0,
      "step": 9517
    },
    {
      "epoch": 1.767502321262767,
      "grad_norm": 1.4449207054229964,
      "learning_rate": 1e-06,
      "loss": 0.3675,
      "mean_token_accuracy": 0.8703007698059082,
      "num_tokens": 331215000.0,
      "step": 9518
    },
    {
      "epoch": 1.7676880222841227,
      "grad_norm": 1.4957860407525516,
      "learning_rate": 1e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8642318248748779,
      "num_tokens": 331256814.0,
      "step": 9519
    },
    {
      "epoch": 1.7678737233054782,
      "grad_norm": 1.5158004936666505,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8920162916183472,
      "num_tokens": 331287439.0,
      "step": 9520
    },
    {
      "epoch": 1.7680594243268337,
      "grad_norm": 1.4717832758645661,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8753833174705505,
      "num_tokens": 331323541.0,
      "step": 9521
    },
    {
      "epoch": 1.7682451253481895,
      "grad_norm": 1.5305612905223653,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8726222515106201,
      "num_tokens": 331360126.0,
      "step": 9522
    },
    {
      "epoch": 1.7684308263695452,
      "grad_norm": 1.4805416992156912,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8668256998062134,
      "num_tokens": 331396950.0,
      "step": 9523
    },
    {
      "epoch": 1.7686165273909007,
      "grad_norm": 1.4668805927229733,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8746928572654724,
      "num_tokens": 331436078.0,
      "step": 9524
    },
    {
      "epoch": 1.7688022284122562,
      "grad_norm": 1.4877557510965504,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8838797211647034,
      "num_tokens": 331472319.0,
      "step": 9525
    },
    {
      "epoch": 1.768987929433612,
      "grad_norm": 1.5942042880817813,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8697253465652466,
      "num_tokens": 331506881.0,
      "step": 9526
    },
    {
      "epoch": 1.7691736304549674,
      "grad_norm": 1.4716701301915307,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8798586130142212,
      "num_tokens": 331543338.0,
      "step": 9527
    },
    {
      "epoch": 1.769359331476323,
      "grad_norm": 1.4507070961156014,
      "learning_rate": 1e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8581433296203613,
      "num_tokens": 331583671.0,
      "step": 9528
    },
    {
      "epoch": 1.7695450324976787,
      "grad_norm": 1.5722407811377461,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8884331583976746,
      "num_tokens": 331617536.0,
      "step": 9529
    },
    {
      "epoch": 1.7697307335190344,
      "grad_norm": 1.381826312918722,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8827839493751526,
      "num_tokens": 331657359.0,
      "step": 9530
    },
    {
      "epoch": 1.76991643454039,
      "grad_norm": 1.5608445696188769,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.875579833984375,
      "num_tokens": 331689792.0,
      "step": 9531
    },
    {
      "epoch": 1.7701021355617454,
      "grad_norm": 1.5648010155721501,
      "learning_rate": 1e-06,
      "loss": 0.3868,
      "mean_token_accuracy": 0.8709491491317749,
      "num_tokens": 331723243.0,
      "step": 9532
    },
    {
      "epoch": 1.7702878365831012,
      "grad_norm": 1.4211981874131605,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8865642547607422,
      "num_tokens": 331758220.0,
      "step": 9533
    },
    {
      "epoch": 1.770473537604457,
      "grad_norm": 1.622488193606328,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8749323487281799,
      "num_tokens": 331788740.0,
      "step": 9534
    },
    {
      "epoch": 1.7706592386258124,
      "grad_norm": 1.5956529416810314,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8901358842849731,
      "num_tokens": 331817885.0,
      "step": 9535
    },
    {
      "epoch": 1.770844939647168,
      "grad_norm": 1.4436084100279973,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.885688841342926,
      "num_tokens": 331852886.0,
      "step": 9536
    },
    {
      "epoch": 1.7710306406685237,
      "grad_norm": 1.4992275643334045,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8826295733451843,
      "num_tokens": 331883746.0,
      "step": 9537
    },
    {
      "epoch": 1.7712163416898794,
      "grad_norm": 1.448779807715992,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.875970721244812,
      "num_tokens": 331922347.0,
      "step": 9538
    },
    {
      "epoch": 1.771402042711235,
      "grad_norm": 1.3628398055783906,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8828136324882507,
      "num_tokens": 331959638.0,
      "step": 9539
    },
    {
      "epoch": 1.7715877437325904,
      "grad_norm": 1.6691843320929167,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8777199983596802,
      "num_tokens": 331987881.0,
      "step": 9540
    },
    {
      "epoch": 1.7717734447539462,
      "grad_norm": 1.5887628721963145,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8741506934165955,
      "num_tokens": 332019231.0,
      "step": 9541
    },
    {
      "epoch": 1.771959145775302,
      "grad_norm": 1.4186107042202063,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8812446594238281,
      "num_tokens": 332056207.0,
      "step": 9542
    },
    {
      "epoch": 1.7721448467966574,
      "grad_norm": 1.3611516183981893,
      "learning_rate": 1e-06,
      "loss": 0.3023,
      "mean_token_accuracy": 0.8947011232376099,
      "num_tokens": 332092562.0,
      "step": 9543
    },
    {
      "epoch": 1.772330547818013,
      "grad_norm": 1.6370732403546917,
      "learning_rate": 1e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8559334874153137,
      "num_tokens": 332126907.0,
      "step": 9544
    },
    {
      "epoch": 1.7725162488393686,
      "grad_norm": 1.412007497063605,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8736567497253418,
      "num_tokens": 332165473.0,
      "step": 9545
    },
    {
      "epoch": 1.7727019498607244,
      "grad_norm": 1.4510172509012043,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8824424743652344,
      "num_tokens": 332202618.0,
      "step": 9546
    },
    {
      "epoch": 1.77288765088208,
      "grad_norm": 1.522473624262834,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8840323686599731,
      "num_tokens": 332237560.0,
      "step": 9547
    },
    {
      "epoch": 1.7730733519034354,
      "grad_norm": 1.6053208441264752,
      "learning_rate": 1e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.8715922236442566,
      "num_tokens": 332268934.0,
      "step": 9548
    },
    {
      "epoch": 1.7732590529247911,
      "grad_norm": 1.5096752770509252,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8737543821334839,
      "num_tokens": 332302416.0,
      "step": 9549
    },
    {
      "epoch": 1.7734447539461469,
      "grad_norm": 1.327447358275429,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8908365368843079,
      "num_tokens": 332343696.0,
      "step": 9550
    },
    {
      "epoch": 1.7736304549675022,
      "grad_norm": 1.6279227966297574,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8631239533424377,
      "num_tokens": 332376009.0,
      "step": 9551
    },
    {
      "epoch": 1.773816155988858,
      "grad_norm": 1.4414053442664478,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8761738538742065,
      "num_tokens": 332413631.0,
      "step": 9552
    },
    {
      "epoch": 1.7740018570102136,
      "grad_norm": 1.3791014422232748,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.89195716381073,
      "num_tokens": 332446076.0,
      "step": 9553
    },
    {
      "epoch": 1.7741875580315691,
      "grad_norm": 1.498978589965303,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8756313323974609,
      "num_tokens": 332477096.0,
      "step": 9554
    },
    {
      "epoch": 1.7743732590529246,
      "grad_norm": 1.3778913739580239,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8820191621780396,
      "num_tokens": 332514415.0,
      "step": 9555
    },
    {
      "epoch": 1.7745589600742804,
      "grad_norm": 1.539892077060889,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8766018748283386,
      "num_tokens": 332546610.0,
      "step": 9556
    },
    {
      "epoch": 1.7747446610956361,
      "grad_norm": 1.5478469169896985,
      "learning_rate": 1e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.869482159614563,
      "num_tokens": 332580113.0,
      "step": 9557
    },
    {
      "epoch": 1.7749303621169916,
      "grad_norm": 1.5126350941916065,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8851025104522705,
      "num_tokens": 332612527.0,
      "step": 9558
    },
    {
      "epoch": 1.7751160631383471,
      "grad_norm": 1.396938645769215,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8728928565979004,
      "num_tokens": 332652468.0,
      "step": 9559
    },
    {
      "epoch": 1.7753017641597029,
      "grad_norm": 1.4799002047473961,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8806650042533875,
      "num_tokens": 332687785.0,
      "step": 9560
    },
    {
      "epoch": 1.7754874651810586,
      "grad_norm": 1.5182784656487855,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8874616622924805,
      "num_tokens": 332722298.0,
      "step": 9561
    },
    {
      "epoch": 1.775673166202414,
      "grad_norm": 1.5808304917098408,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8838069438934326,
      "num_tokens": 332753384.0,
      "step": 9562
    },
    {
      "epoch": 1.7758588672237696,
      "grad_norm": 1.5708958322435387,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8621938228607178,
      "num_tokens": 332785177.0,
      "step": 9563
    },
    {
      "epoch": 1.7760445682451254,
      "grad_norm": 1.5051623678451413,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8827488422393799,
      "num_tokens": 332817976.0,
      "step": 9564
    },
    {
      "epoch": 1.776230269266481,
      "grad_norm": 1.571661344632462,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8738819360733032,
      "num_tokens": 332851446.0,
      "step": 9565
    },
    {
      "epoch": 1.7764159702878366,
      "grad_norm": 1.4080517331237767,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8852699995040894,
      "num_tokens": 332887491.0,
      "step": 9566
    },
    {
      "epoch": 1.776601671309192,
      "grad_norm": 1.4903123941578316,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8732228875160217,
      "num_tokens": 332922580.0,
      "step": 9567
    },
    {
      "epoch": 1.7767873723305478,
      "grad_norm": 1.329321922430677,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8851044178009033,
      "num_tokens": 332961129.0,
      "step": 9568
    },
    {
      "epoch": 1.7769730733519036,
      "grad_norm": 1.423001122248975,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8892411589622498,
      "num_tokens": 332995974.0,
      "step": 9569
    },
    {
      "epoch": 1.777158774373259,
      "grad_norm": 1.506349861632492,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8758533596992493,
      "num_tokens": 333031380.0,
      "step": 9570
    },
    {
      "epoch": 1.7773444753946146,
      "grad_norm": 1.5224893964427886,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8791964054107666,
      "num_tokens": 333062028.0,
      "step": 9571
    },
    {
      "epoch": 1.7775301764159703,
      "grad_norm": 1.494019750423264,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.875007152557373,
      "num_tokens": 333096848.0,
      "step": 9572
    },
    {
      "epoch": 1.777715877437326,
      "grad_norm": 1.4488622570249554,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8691425323486328,
      "num_tokens": 333135145.0,
      "step": 9573
    },
    {
      "epoch": 1.7779015784586816,
      "grad_norm": 1.5190168670001798,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8623557686805725,
      "num_tokens": 333171441.0,
      "step": 9574
    },
    {
      "epoch": 1.778087279480037,
      "grad_norm": 1.611418477747962,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8871928453445435,
      "num_tokens": 333198241.0,
      "step": 9575
    },
    {
      "epoch": 1.7782729805013928,
      "grad_norm": 1.560074098652991,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8647684454917908,
      "num_tokens": 333231750.0,
      "step": 9576
    },
    {
      "epoch": 1.7784586815227483,
      "grad_norm": 1.6001341840677523,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.870376706123352,
      "num_tokens": 333266084.0,
      "step": 9577
    },
    {
      "epoch": 1.7786443825441038,
      "grad_norm": 1.5745894548556614,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8799435496330261,
      "num_tokens": 333297464.0,
      "step": 9578
    },
    {
      "epoch": 1.7788300835654596,
      "grad_norm": 1.3538734271570791,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8783429265022278,
      "num_tokens": 333337882.0,
      "step": 9579
    },
    {
      "epoch": 1.7790157845868153,
      "grad_norm": 1.4601033238911492,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8795233964920044,
      "num_tokens": 333374964.0,
      "step": 9580
    },
    {
      "epoch": 1.7792014856081708,
      "grad_norm": 1.43623312688988,
      "learning_rate": 1e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.869368851184845,
      "num_tokens": 333411646.0,
      "step": 9581
    },
    {
      "epoch": 1.7793871866295263,
      "grad_norm": 1.4372809365577304,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8808941841125488,
      "num_tokens": 333449746.0,
      "step": 9582
    },
    {
      "epoch": 1.779572887650882,
      "grad_norm": 1.5547318712068787,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8781658411026001,
      "num_tokens": 333481856.0,
      "step": 9583
    },
    {
      "epoch": 1.7797585886722378,
      "grad_norm": 1.5083753020769841,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8768443465232849,
      "num_tokens": 333516728.0,
      "step": 9584
    },
    {
      "epoch": 1.7799442896935933,
      "grad_norm": 1.4984874597618878,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.881486177444458,
      "num_tokens": 333550400.0,
      "step": 9585
    },
    {
      "epoch": 1.7801299907149488,
      "grad_norm": 1.3978317138223522,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8929013013839722,
      "num_tokens": 333586278.0,
      "step": 9586
    },
    {
      "epoch": 1.7803156917363046,
      "grad_norm": 1.4338888678229367,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.886101484298706,
      "num_tokens": 333620553.0,
      "step": 9587
    },
    {
      "epoch": 1.7805013927576603,
      "grad_norm": 1.369401612454397,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8885422945022583,
      "num_tokens": 333659630.0,
      "step": 9588
    },
    {
      "epoch": 1.7806870937790158,
      "grad_norm": 1.421808314484137,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8765647411346436,
      "num_tokens": 333698131.0,
      "step": 9589
    },
    {
      "epoch": 1.7808727948003713,
      "grad_norm": 1.5494054498389433,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8743111491203308,
      "num_tokens": 333730718.0,
      "step": 9590
    },
    {
      "epoch": 1.781058495821727,
      "grad_norm": 1.5565753340398463,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8751556277275085,
      "num_tokens": 333764379.0,
      "step": 9591
    },
    {
      "epoch": 1.7812441968430828,
      "grad_norm": 1.4480692896165421,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8760358691215515,
      "num_tokens": 333801420.0,
      "step": 9592
    },
    {
      "epoch": 1.7814298978644383,
      "grad_norm": 1.6041186491575854,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8735288381576538,
      "num_tokens": 333832457.0,
      "step": 9593
    },
    {
      "epoch": 1.7816155988857938,
      "grad_norm": 1.4397423649881893,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.87788325548172,
      "num_tokens": 333865944.0,
      "step": 9594
    },
    {
      "epoch": 1.7818012999071495,
      "grad_norm": 1.5127809998745512,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8713981509208679,
      "num_tokens": 333900742.0,
      "step": 9595
    },
    {
      "epoch": 1.7819870009285053,
      "grad_norm": 1.4694829817722863,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8939704298973083,
      "num_tokens": 333935352.0,
      "step": 9596
    },
    {
      "epoch": 1.7821727019498608,
      "grad_norm": 1.6323747133156656,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8653614521026611,
      "num_tokens": 333970500.0,
      "step": 9597
    },
    {
      "epoch": 1.7823584029712163,
      "grad_norm": 1.7017985749864935,
      "learning_rate": 1e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8619546294212341,
      "num_tokens": 334000571.0,
      "step": 9598
    },
    {
      "epoch": 1.782544103992572,
      "grad_norm": 1.470427409243893,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8838894367218018,
      "num_tokens": 334038695.0,
      "step": 9599
    },
    {
      "epoch": 1.7827298050139275,
      "grad_norm": 1.4962553414578383,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8742306232452393,
      "num_tokens": 334074294.0,
      "step": 9600
    },
    {
      "epoch": 1.782915506035283,
      "grad_norm": 1.5419772980880202,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8775047063827515,
      "num_tokens": 334110774.0,
      "step": 9601
    },
    {
      "epoch": 1.7831012070566388,
      "grad_norm": 1.4065795640596064,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8823339939117432,
      "num_tokens": 334151200.0,
      "step": 9602
    },
    {
      "epoch": 1.7832869080779945,
      "grad_norm": 1.523494341854037,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8760607242584229,
      "num_tokens": 334185647.0,
      "step": 9603
    },
    {
      "epoch": 1.78347260909935,
      "grad_norm": 1.5476101578202208,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.877607524394989,
      "num_tokens": 334218796.0,
      "step": 9604
    },
    {
      "epoch": 1.7836583101207055,
      "grad_norm": 1.7170161035843796,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8862828612327576,
      "num_tokens": 334242763.0,
      "step": 9605
    },
    {
      "epoch": 1.7838440111420613,
      "grad_norm": 1.5514890593976822,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8694842457771301,
      "num_tokens": 334276515.0,
      "step": 9606
    },
    {
      "epoch": 1.784029712163417,
      "grad_norm": 1.367425795660869,
      "learning_rate": 1e-06,
      "loss": 0.3564,
      "mean_token_accuracy": 0.8804474472999573,
      "num_tokens": 334316390.0,
      "step": 9607
    },
    {
      "epoch": 1.7842154131847725,
      "grad_norm": 1.4801907776169323,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8952798247337341,
      "num_tokens": 334349839.0,
      "step": 9608
    },
    {
      "epoch": 1.784401114206128,
      "grad_norm": 1.586991226685328,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8755232095718384,
      "num_tokens": 334382392.0,
      "step": 9609
    },
    {
      "epoch": 1.7845868152274837,
      "grad_norm": 1.3562299915317575,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.884816586971283,
      "num_tokens": 334422640.0,
      "step": 9610
    },
    {
      "epoch": 1.7847725162488395,
      "grad_norm": 1.6504468537411747,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.883715033531189,
      "num_tokens": 334448851.0,
      "step": 9611
    },
    {
      "epoch": 1.784958217270195,
      "grad_norm": 1.5301560072847953,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8897948861122131,
      "num_tokens": 334478439.0,
      "step": 9612
    },
    {
      "epoch": 1.7851439182915505,
      "grad_norm": 1.533377903852234,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8847265243530273,
      "num_tokens": 334509595.0,
      "step": 9613
    },
    {
      "epoch": 1.7853296193129062,
      "grad_norm": 1.703331981776649,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8706626296043396,
      "num_tokens": 334538567.0,
      "step": 9614
    },
    {
      "epoch": 1.785515320334262,
      "grad_norm": 1.4781743216120913,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.88688063621521,
      "num_tokens": 334576749.0,
      "step": 9615
    },
    {
      "epoch": 1.7857010213556175,
      "grad_norm": 1.5377981460653283,
      "learning_rate": 1e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.8664101362228394,
      "num_tokens": 334611078.0,
      "step": 9616
    },
    {
      "epoch": 1.785886722376973,
      "grad_norm": 1.6392566094274896,
      "learning_rate": 1e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8629566431045532,
      "num_tokens": 334644103.0,
      "step": 9617
    },
    {
      "epoch": 1.7860724233983287,
      "grad_norm": 1.5315535030691316,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8678716421127319,
      "num_tokens": 334678662.0,
      "step": 9618
    },
    {
      "epoch": 1.7862581244196845,
      "grad_norm": 1.4370957180030226,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8833505511283875,
      "num_tokens": 334713711.0,
      "step": 9619
    },
    {
      "epoch": 1.78644382544104,
      "grad_norm": 1.52445769544512,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8834197521209717,
      "num_tokens": 334746572.0,
      "step": 9620
    },
    {
      "epoch": 1.7866295264623955,
      "grad_norm": 1.574291359068326,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8735908269882202,
      "num_tokens": 334779580.0,
      "step": 9621
    },
    {
      "epoch": 1.7868152274837512,
      "grad_norm": 1.8894336311369513,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8750894069671631,
      "num_tokens": 334802311.0,
      "step": 9622
    },
    {
      "epoch": 1.7870009285051067,
      "grad_norm": 1.5310962412933187,
      "learning_rate": 1e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8675202131271362,
      "num_tokens": 334836228.0,
      "step": 9623
    },
    {
      "epoch": 1.7871866295264622,
      "grad_norm": 1.5363008915602678,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8729997873306274,
      "num_tokens": 334868660.0,
      "step": 9624
    },
    {
      "epoch": 1.787372330547818,
      "grad_norm": 1.5131215882339386,
      "learning_rate": 1e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8624353408813477,
      "num_tokens": 334906637.0,
      "step": 9625
    },
    {
      "epoch": 1.7875580315691737,
      "grad_norm": 1.4200904728148114,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8877757787704468,
      "num_tokens": 334944567.0,
      "step": 9626
    },
    {
      "epoch": 1.7877437325905292,
      "grad_norm": 1.7002651889931548,
      "learning_rate": 1e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8648489713668823,
      "num_tokens": 334974213.0,
      "step": 9627
    },
    {
      "epoch": 1.7879294336118847,
      "grad_norm": 1.4646011616879406,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8791202306747437,
      "num_tokens": 335008974.0,
      "step": 9628
    },
    {
      "epoch": 1.7881151346332405,
      "grad_norm": 1.4797498491670544,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8691658973693848,
      "num_tokens": 335047259.0,
      "step": 9629
    },
    {
      "epoch": 1.7883008356545962,
      "grad_norm": 1.446276369690307,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.890394926071167,
      "num_tokens": 335088076.0,
      "step": 9630
    },
    {
      "epoch": 1.7884865366759517,
      "grad_norm": 1.531281866105619,
      "learning_rate": 1e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8667460680007935,
      "num_tokens": 335125706.0,
      "step": 9631
    },
    {
      "epoch": 1.7886722376973072,
      "grad_norm": 1.4821195108958647,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8767945766448975,
      "num_tokens": 335159850.0,
      "step": 9632
    },
    {
      "epoch": 1.788857938718663,
      "grad_norm": 1.4943990635415272,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8798721432685852,
      "num_tokens": 335192392.0,
      "step": 9633
    },
    {
      "epoch": 1.7890436397400187,
      "grad_norm": 1.366202161554277,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8803521990776062,
      "num_tokens": 335233729.0,
      "step": 9634
    },
    {
      "epoch": 1.7892293407613742,
      "grad_norm": 1.4679180725785883,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8717117309570312,
      "num_tokens": 335272078.0,
      "step": 9635
    },
    {
      "epoch": 1.7894150417827297,
      "grad_norm": 1.4893342214650402,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8666938543319702,
      "num_tokens": 335310681.0,
      "step": 9636
    },
    {
      "epoch": 1.7896007428040854,
      "grad_norm": 1.6661469266031486,
      "learning_rate": 1e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.863013744354248,
      "num_tokens": 335341529.0,
      "step": 9637
    },
    {
      "epoch": 1.7897864438254412,
      "grad_norm": 1.590285107949946,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8814475536346436,
      "num_tokens": 335376068.0,
      "step": 9638
    },
    {
      "epoch": 1.7899721448467967,
      "grad_norm": 1.5799263164049737,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8826369047164917,
      "num_tokens": 335408437.0,
      "step": 9639
    },
    {
      "epoch": 1.7901578458681522,
      "grad_norm": 1.3413606604514214,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8769180774688721,
      "num_tokens": 335450247.0,
      "step": 9640
    },
    {
      "epoch": 1.790343546889508,
      "grad_norm": 1.5786222337577867,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8618605136871338,
      "num_tokens": 335486081.0,
      "step": 9641
    },
    {
      "epoch": 1.7905292479108637,
      "grad_norm": 1.4235252079963447,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8918924331665039,
      "num_tokens": 335522100.0,
      "step": 9642
    },
    {
      "epoch": 1.7907149489322192,
      "grad_norm": 1.4819613956412008,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8731754422187805,
      "num_tokens": 335554254.0,
      "step": 9643
    },
    {
      "epoch": 1.7909006499535747,
      "grad_norm": 1.3434675610480369,
      "learning_rate": 1e-06,
      "loss": 0.2752,
      "mean_token_accuracy": 0.9036270380020142,
      "num_tokens": 335587118.0,
      "step": 9644
    },
    {
      "epoch": 1.7910863509749304,
      "grad_norm": 1.400404897537266,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8863075375556946,
      "num_tokens": 335620569.0,
      "step": 9645
    },
    {
      "epoch": 1.7912720519962861,
      "grad_norm": 1.610128856901238,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8783711194992065,
      "num_tokens": 335648878.0,
      "step": 9646
    },
    {
      "epoch": 1.7914577530176417,
      "grad_norm": 1.5204185928235137,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8824925422668457,
      "num_tokens": 335682352.0,
      "step": 9647
    },
    {
      "epoch": 1.7916434540389972,
      "grad_norm": 1.4866402206531457,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8916637897491455,
      "num_tokens": 335715658.0,
      "step": 9648
    },
    {
      "epoch": 1.791829155060353,
      "grad_norm": 1.4826686915156433,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8785771131515503,
      "num_tokens": 335749247.0,
      "step": 9649
    },
    {
      "epoch": 1.7920148560817084,
      "grad_norm": 1.4733065245589652,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8732658624649048,
      "num_tokens": 335791379.0,
      "step": 9650
    },
    {
      "epoch": 1.792200557103064,
      "grad_norm": 1.6055340521301023,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8821951746940613,
      "num_tokens": 335822903.0,
      "step": 9651
    },
    {
      "epoch": 1.7923862581244197,
      "grad_norm": 1.5752613562395725,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8735488057136536,
      "num_tokens": 335856955.0,
      "step": 9652
    },
    {
      "epoch": 1.7925719591457754,
      "grad_norm": 1.6081697350355393,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8696172833442688,
      "num_tokens": 335890583.0,
      "step": 9653
    },
    {
      "epoch": 1.792757660167131,
      "grad_norm": 1.7778639054964744,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.865699052810669,
      "num_tokens": 335920781.0,
      "step": 9654
    },
    {
      "epoch": 1.7929433611884864,
      "grad_norm": 1.5889367970964279,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8811970949172974,
      "num_tokens": 335949741.0,
      "step": 9655
    },
    {
      "epoch": 1.7931290622098421,
      "grad_norm": 1.3858803681694247,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8911113739013672,
      "num_tokens": 335987478.0,
      "step": 9656
    },
    {
      "epoch": 1.7933147632311979,
      "grad_norm": 1.513418038249935,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8814449310302734,
      "num_tokens": 336023816.0,
      "step": 9657
    },
    {
      "epoch": 1.7935004642525534,
      "grad_norm": 1.5637229128754495,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8745584487915039,
      "num_tokens": 336058027.0,
      "step": 9658
    },
    {
      "epoch": 1.793686165273909,
      "grad_norm": 1.396801457874952,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8859848976135254,
      "num_tokens": 336093022.0,
      "step": 9659
    },
    {
      "epoch": 1.7938718662952646,
      "grad_norm": 1.4500482566528052,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8856260776519775,
      "num_tokens": 336128457.0,
      "step": 9660
    },
    {
      "epoch": 1.7940575673166204,
      "grad_norm": 1.5334455710021633,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.879928708076477,
      "num_tokens": 336164691.0,
      "step": 9661
    },
    {
      "epoch": 1.7942432683379759,
      "grad_norm": 1.488632733215279,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8823573589324951,
      "num_tokens": 336196474.0,
      "step": 9662
    },
    {
      "epoch": 1.7944289693593314,
      "grad_norm": 1.339355613415506,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8735005855560303,
      "num_tokens": 336240822.0,
      "step": 9663
    },
    {
      "epoch": 1.7946146703806871,
      "grad_norm": 1.4222030928914466,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8793340921401978,
      "num_tokens": 336276071.0,
      "step": 9664
    },
    {
      "epoch": 1.7948003714020428,
      "grad_norm": 1.5697083999715677,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8687930107116699,
      "num_tokens": 336306370.0,
      "step": 9665
    },
    {
      "epoch": 1.7949860724233984,
      "grad_norm": 1.509722024383623,
      "learning_rate": 1e-06,
      "loss": 0.3727,
      "mean_token_accuracy": 0.8730128407478333,
      "num_tokens": 336341736.0,
      "step": 9666
    },
    {
      "epoch": 1.7951717734447539,
      "grad_norm": 1.4464945990880484,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8792003393173218,
      "num_tokens": 336378560.0,
      "step": 9667
    },
    {
      "epoch": 1.7953574744661096,
      "grad_norm": 1.5461024203673666,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8709026575088501,
      "num_tokens": 336410804.0,
      "step": 9668
    },
    {
      "epoch": 1.7955431754874653,
      "grad_norm": 1.443829131149468,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8825927972793579,
      "num_tokens": 336443549.0,
      "step": 9669
    },
    {
      "epoch": 1.7957288765088208,
      "grad_norm": 1.4801058635341635,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8750383257865906,
      "num_tokens": 336479774.0,
      "step": 9670
    },
    {
      "epoch": 1.7959145775301764,
      "grad_norm": 1.3502873264672124,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8840858340263367,
      "num_tokens": 336518119.0,
      "step": 9671
    },
    {
      "epoch": 1.796100278551532,
      "grad_norm": 1.6941253227201898,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8643277287483215,
      "num_tokens": 336549385.0,
      "step": 9672
    },
    {
      "epoch": 1.7962859795728876,
      "grad_norm": 1.4450912087728172,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8767744302749634,
      "num_tokens": 336587425.0,
      "step": 9673
    },
    {
      "epoch": 1.7964716805942431,
      "grad_norm": 1.4247235605463948,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8799009919166565,
      "num_tokens": 336625636.0,
      "step": 9674
    },
    {
      "epoch": 1.7966573816155988,
      "grad_norm": 1.3346988439363199,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8837718367576599,
      "num_tokens": 336668605.0,
      "step": 9675
    },
    {
      "epoch": 1.7968430826369546,
      "grad_norm": 1.4474016698552432,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.8813353776931763,
      "num_tokens": 336707200.0,
      "step": 9676
    },
    {
      "epoch": 1.79702878365831,
      "grad_norm": 1.4652538243369568,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8867911100387573,
      "num_tokens": 336742263.0,
      "step": 9677
    },
    {
      "epoch": 1.7972144846796656,
      "grad_norm": 1.4981751103653755,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8704237937927246,
      "num_tokens": 336777148.0,
      "step": 9678
    },
    {
      "epoch": 1.7974001857010213,
      "grad_norm": 1.3340024242057988,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8835791945457458,
      "num_tokens": 336817281.0,
      "step": 9679
    },
    {
      "epoch": 1.797585886722377,
      "grad_norm": 1.4605264953922161,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8956804275512695,
      "num_tokens": 336849028.0,
      "step": 9680
    },
    {
      "epoch": 1.7977715877437326,
      "grad_norm": 1.4917475836422855,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8853979110717773,
      "num_tokens": 336884093.0,
      "step": 9681
    },
    {
      "epoch": 1.797957288765088,
      "grad_norm": 1.5973461379507836,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8897061347961426,
      "num_tokens": 336915994.0,
      "step": 9682
    },
    {
      "epoch": 1.7981429897864438,
      "grad_norm": 1.4368583054681456,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8806095719337463,
      "num_tokens": 336952157.0,
      "step": 9683
    },
    {
      "epoch": 1.7983286908077996,
      "grad_norm": 1.484132966668584,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8726102113723755,
      "num_tokens": 336986969.0,
      "step": 9684
    },
    {
      "epoch": 1.798514391829155,
      "grad_norm": 1.485892717823668,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.868585467338562,
      "num_tokens": 337028174.0,
      "step": 9685
    },
    {
      "epoch": 1.7987000928505106,
      "grad_norm": 1.5337260988350665,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8856082558631897,
      "num_tokens": 337059662.0,
      "step": 9686
    },
    {
      "epoch": 1.7988857938718663,
      "grad_norm": 1.664995201121616,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8763622045516968,
      "num_tokens": 337085962.0,
      "step": 9687
    },
    {
      "epoch": 1.799071494893222,
      "grad_norm": 1.5550034074070402,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8694018125534058,
      "num_tokens": 337120645.0,
      "step": 9688
    },
    {
      "epoch": 1.7992571959145776,
      "grad_norm": 1.6078242194424817,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8816481828689575,
      "num_tokens": 337150824.0,
      "step": 9689
    },
    {
      "epoch": 1.799442896935933,
      "grad_norm": 1.622065984855058,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8859692811965942,
      "num_tokens": 337181353.0,
      "step": 9690
    },
    {
      "epoch": 1.7996285979572888,
      "grad_norm": 1.4145590801647738,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8823814392089844,
      "num_tokens": 337215521.0,
      "step": 9691
    },
    {
      "epoch": 1.7998142989786445,
      "grad_norm": 1.3989335725897727,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8736006021499634,
      "num_tokens": 337251871.0,
      "step": 9692
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.4213584676120103,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8801658749580383,
      "num_tokens": 337287544.0,
      "step": 9693
    },
    {
      "epoch": 1.8001857010213556,
      "grad_norm": 1.551955862036217,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8767290115356445,
      "num_tokens": 337319011.0,
      "step": 9694
    },
    {
      "epoch": 1.8003714020427113,
      "grad_norm": 1.624156196530315,
      "learning_rate": 1e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8537789583206177,
      "num_tokens": 337353046.0,
      "step": 9695
    },
    {
      "epoch": 1.8005571030640668,
      "grad_norm": 1.4286617577058816,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.887417197227478,
      "num_tokens": 337385452.0,
      "step": 9696
    },
    {
      "epoch": 1.8007428040854223,
      "grad_norm": 1.380761755076861,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.888818621635437,
      "num_tokens": 337425665.0,
      "step": 9697
    },
    {
      "epoch": 1.800928505106778,
      "grad_norm": 1.5024204889248793,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8762259483337402,
      "num_tokens": 337461520.0,
      "step": 9698
    },
    {
      "epoch": 1.8011142061281338,
      "grad_norm": 1.6396863223409266,
      "learning_rate": 1e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8679322004318237,
      "num_tokens": 337492877.0,
      "step": 9699
    },
    {
      "epoch": 1.8012999071494893,
      "grad_norm": 1.537134805290461,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.880012571811676,
      "num_tokens": 337526093.0,
      "step": 9700
    },
    {
      "epoch": 1.8014856081708448,
      "grad_norm": 1.5112598196153861,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.872529149055481,
      "num_tokens": 337559867.0,
      "step": 9701
    },
    {
      "epoch": 1.8016713091922005,
      "grad_norm": 1.5244617470661317,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.886309027671814,
      "num_tokens": 337591285.0,
      "step": 9702
    },
    {
      "epoch": 1.8018570102135563,
      "grad_norm": 1.527620143720097,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8802996873855591,
      "num_tokens": 337627120.0,
      "step": 9703
    },
    {
      "epoch": 1.8020427112349118,
      "grad_norm": 1.5757317279354566,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8824113607406616,
      "num_tokens": 337664465.0,
      "step": 9704
    },
    {
      "epoch": 1.8022284122562673,
      "grad_norm": 1.6196221653552223,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8860749006271362,
      "num_tokens": 337694414.0,
      "step": 9705
    },
    {
      "epoch": 1.802414113277623,
      "grad_norm": 1.4492242861846238,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8705329895019531,
      "num_tokens": 337732401.0,
      "step": 9706
    },
    {
      "epoch": 1.8025998142989788,
      "grad_norm": 1.7331904758834409,
      "learning_rate": 1e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8697913885116577,
      "num_tokens": 337758369.0,
      "step": 9707
    },
    {
      "epoch": 1.8027855153203343,
      "grad_norm": 1.686486188259655,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8728407025337219,
      "num_tokens": 337791402.0,
      "step": 9708
    },
    {
      "epoch": 1.8029712163416898,
      "grad_norm": 1.6137986991601,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8936094045639038,
      "num_tokens": 337820819.0,
      "step": 9709
    },
    {
      "epoch": 1.8031569173630455,
      "grad_norm": 1.8795617751019849,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8696162104606628,
      "num_tokens": 337847304.0,
      "step": 9710
    },
    {
      "epoch": 1.8033426183844012,
      "grad_norm": 1.5082749490478318,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8817423582077026,
      "num_tokens": 337883016.0,
      "step": 9711
    },
    {
      "epoch": 1.8035283194057568,
      "grad_norm": 1.5171166093153283,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.884233832359314,
      "num_tokens": 337918263.0,
      "step": 9712
    },
    {
      "epoch": 1.8037140204271123,
      "grad_norm": 1.478784682426473,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8839277029037476,
      "num_tokens": 337950110.0,
      "step": 9713
    },
    {
      "epoch": 1.803899721448468,
      "grad_norm": 1.4293756460284242,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8749565482139587,
      "num_tokens": 337987818.0,
      "step": 9714
    },
    {
      "epoch": 1.8040854224698237,
      "grad_norm": 1.6798067097373075,
      "learning_rate": 1e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8605151176452637,
      "num_tokens": 338023422.0,
      "step": 9715
    },
    {
      "epoch": 1.8042711234911792,
      "grad_norm": 1.8034829324419794,
      "learning_rate": 1e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.869024395942688,
      "num_tokens": 338052189.0,
      "step": 9716
    },
    {
      "epoch": 1.8044568245125348,
      "grad_norm": 1.5594059066424037,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8792766332626343,
      "num_tokens": 338088370.0,
      "step": 9717
    },
    {
      "epoch": 1.8046425255338905,
      "grad_norm": 1.581098906762258,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8691563606262207,
      "num_tokens": 338121703.0,
      "step": 9718
    },
    {
      "epoch": 1.8048282265552462,
      "grad_norm": 1.534242343505563,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8794918060302734,
      "num_tokens": 338158225.0,
      "step": 9719
    },
    {
      "epoch": 1.8050139275766015,
      "grad_norm": 1.5400075174881067,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8695259094238281,
      "num_tokens": 338194919.0,
      "step": 9720
    },
    {
      "epoch": 1.8051996285979572,
      "grad_norm": 1.6005472937621865,
      "learning_rate": 1e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8677855134010315,
      "num_tokens": 338228113.0,
      "step": 9721
    },
    {
      "epoch": 1.805385329619313,
      "grad_norm": 1.4897863580297221,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8791019916534424,
      "num_tokens": 338260431.0,
      "step": 9722
    },
    {
      "epoch": 1.8055710306406685,
      "grad_norm": 1.6617644707665817,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8763666152954102,
      "num_tokens": 338288890.0,
      "step": 9723
    },
    {
      "epoch": 1.805756731662024,
      "grad_norm": 1.4460760733982687,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8828213810920715,
      "num_tokens": 338327302.0,
      "step": 9724
    },
    {
      "epoch": 1.8059424326833797,
      "grad_norm": 1.4913593698768042,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8763257265090942,
      "num_tokens": 338363659.0,
      "step": 9725
    },
    {
      "epoch": 1.8061281337047355,
      "grad_norm": 1.5323297548768748,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8834462761878967,
      "num_tokens": 338397346.0,
      "step": 9726
    },
    {
      "epoch": 1.806313834726091,
      "grad_norm": 1.4374789171204085,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8858515620231628,
      "num_tokens": 338432214.0,
      "step": 9727
    },
    {
      "epoch": 1.8064995357474465,
      "grad_norm": 1.7922986412669748,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8809090852737427,
      "num_tokens": 338462895.0,
      "step": 9728
    },
    {
      "epoch": 1.8066852367688022,
      "grad_norm": 1.6239891488226217,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8697006702423096,
      "num_tokens": 338496597.0,
      "step": 9729
    },
    {
      "epoch": 1.806870937790158,
      "grad_norm": 1.468177114970843,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8778129816055298,
      "num_tokens": 338532950.0,
      "step": 9730
    },
    {
      "epoch": 1.8070566388115135,
      "grad_norm": 1.6260835016143846,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8757274150848389,
      "num_tokens": 338569406.0,
      "step": 9731
    },
    {
      "epoch": 1.807242339832869,
      "grad_norm": 1.4899670179330413,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8829317688941956,
      "num_tokens": 338602002.0,
      "step": 9732
    },
    {
      "epoch": 1.8074280408542247,
      "grad_norm": 1.4891126055647421,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8737099170684814,
      "num_tokens": 338635503.0,
      "step": 9733
    },
    {
      "epoch": 1.8076137418755804,
      "grad_norm": 1.450180433927759,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8753584027290344,
      "num_tokens": 338670353.0,
      "step": 9734
    },
    {
      "epoch": 1.807799442896936,
      "grad_norm": 1.4392036921493163,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8732428550720215,
      "num_tokens": 338711005.0,
      "step": 9735
    },
    {
      "epoch": 1.8079851439182915,
      "grad_norm": 1.4602387776462087,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8749580383300781,
      "num_tokens": 338747189.0,
      "step": 9736
    },
    {
      "epoch": 1.8081708449396472,
      "grad_norm": 1.6099694324674247,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8821903467178345,
      "num_tokens": 338773792.0,
      "step": 9737
    },
    {
      "epoch": 1.808356545961003,
      "grad_norm": 1.6345286184152434,
      "learning_rate": 1e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8635896444320679,
      "num_tokens": 338804524.0,
      "step": 9738
    },
    {
      "epoch": 1.8085422469823584,
      "grad_norm": 1.390257129005434,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8920274376869202,
      "num_tokens": 338840552.0,
      "step": 9739
    },
    {
      "epoch": 1.808727948003714,
      "grad_norm": 1.545559911238161,
      "learning_rate": 1e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8629055023193359,
      "num_tokens": 338881056.0,
      "step": 9740
    },
    {
      "epoch": 1.8089136490250697,
      "grad_norm": 1.4277388092744074,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8789992332458496,
      "num_tokens": 338919777.0,
      "step": 9741
    },
    {
      "epoch": 1.8090993500464254,
      "grad_norm": 1.4032218466841035,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.8810268044471741,
      "num_tokens": 338956843.0,
      "step": 9742
    },
    {
      "epoch": 1.809285051067781,
      "grad_norm": 1.4610768320081726,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8743476867675781,
      "num_tokens": 338996738.0,
      "step": 9743
    },
    {
      "epoch": 1.8094707520891364,
      "grad_norm": 1.5215140143478498,
      "learning_rate": 1e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8558846116065979,
      "num_tokens": 339034171.0,
      "step": 9744
    },
    {
      "epoch": 1.8096564531104922,
      "grad_norm": 1.585635142280607,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8819909691810608,
      "num_tokens": 339063781.0,
      "step": 9745
    },
    {
      "epoch": 1.8098421541318477,
      "grad_norm": 1.5334532140386579,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8826096057891846,
      "num_tokens": 339094394.0,
      "step": 9746
    },
    {
      "epoch": 1.8100278551532032,
      "grad_norm": 1.5551490738585472,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.87939453125,
      "num_tokens": 339124406.0,
      "step": 9747
    },
    {
      "epoch": 1.810213556174559,
      "grad_norm": 1.4680346561688946,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8870775699615479,
      "num_tokens": 339162458.0,
      "step": 9748
    },
    {
      "epoch": 1.8103992571959147,
      "grad_norm": 1.3959832827765997,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8741730451583862,
      "num_tokens": 339200083.0,
      "step": 9749
    },
    {
      "epoch": 1.8105849582172702,
      "grad_norm": 1.633059356434082,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8793777227401733,
      "num_tokens": 339229149.0,
      "step": 9750
    },
    {
      "epoch": 1.8107706592386257,
      "grad_norm": 1.411639370772201,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.883434534072876,
      "num_tokens": 339265978.0,
      "step": 9751
    },
    {
      "epoch": 1.8109563602599814,
      "grad_norm": 1.454508898977928,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8764380216598511,
      "num_tokens": 339300787.0,
      "step": 9752
    },
    {
      "epoch": 1.8111420612813371,
      "grad_norm": 1.398344018163686,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8768118023872375,
      "num_tokens": 339336507.0,
      "step": 9753
    },
    {
      "epoch": 1.8113277623026927,
      "grad_norm": 1.6838897126478998,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.88690185546875,
      "num_tokens": 339364232.0,
      "step": 9754
    },
    {
      "epoch": 1.8115134633240482,
      "grad_norm": 1.4305714942874714,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8896256685256958,
      "num_tokens": 339397278.0,
      "step": 9755
    },
    {
      "epoch": 1.811699164345404,
      "grad_norm": 1.3866865337157763,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8749527335166931,
      "num_tokens": 339432781.0,
      "step": 9756
    },
    {
      "epoch": 1.8118848653667596,
      "grad_norm": 1.4515748952685974,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8736224174499512,
      "num_tokens": 339468271.0,
      "step": 9757
    },
    {
      "epoch": 1.8120705663881151,
      "grad_norm": 1.5377451136103837,
      "learning_rate": 1e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8577291369438171,
      "num_tokens": 339506853.0,
      "step": 9758
    },
    {
      "epoch": 1.8122562674094707,
      "grad_norm": 1.534394215271549,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8667163252830505,
      "num_tokens": 339544198.0,
      "step": 9759
    },
    {
      "epoch": 1.8124419684308264,
      "grad_norm": 1.4785077541040224,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8684550523757935,
      "num_tokens": 339581310.0,
      "step": 9760
    },
    {
      "epoch": 1.8126276694521821,
      "grad_norm": 1.4908015390077964,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8733887076377869,
      "num_tokens": 339617363.0,
      "step": 9761
    },
    {
      "epoch": 1.8128133704735376,
      "grad_norm": 1.4871407901198337,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8650095462799072,
      "num_tokens": 339654633.0,
      "step": 9762
    },
    {
      "epoch": 1.8129990714948931,
      "grad_norm": 1.4062337901067488,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8766413331031799,
      "num_tokens": 339689462.0,
      "step": 9763
    },
    {
      "epoch": 1.8131847725162489,
      "grad_norm": 1.5395134140221596,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8760252594947815,
      "num_tokens": 339722319.0,
      "step": 9764
    },
    {
      "epoch": 1.8133704735376046,
      "grad_norm": 1.3792211908620833,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8815582990646362,
      "num_tokens": 339758925.0,
      "step": 9765
    },
    {
      "epoch": 1.8135561745589601,
      "grad_norm": 1.7072335580120048,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8834024667739868,
      "num_tokens": 339789348.0,
      "step": 9766
    },
    {
      "epoch": 1.8137418755803156,
      "grad_norm": 1.4901819885586751,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8714212775230408,
      "num_tokens": 339825091.0,
      "step": 9767
    },
    {
      "epoch": 1.8139275766016714,
      "grad_norm": 1.4856071617875235,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8702090382575989,
      "num_tokens": 339866803.0,
      "step": 9768
    },
    {
      "epoch": 1.8141132776230269,
      "grad_norm": 1.5115030423105749,
      "learning_rate": 1e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.8713653087615967,
      "num_tokens": 339900960.0,
      "step": 9769
    },
    {
      "epoch": 1.8142989786443824,
      "grad_norm": 1.538456465727653,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8741164207458496,
      "num_tokens": 339931818.0,
      "step": 9770
    },
    {
      "epoch": 1.8144846796657381,
      "grad_norm": 1.5400555470073685,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8892079591751099,
      "num_tokens": 339962041.0,
      "step": 9771
    },
    {
      "epoch": 1.8146703806870939,
      "grad_norm": 1.5154646058595307,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8769546747207642,
      "num_tokens": 339996786.0,
      "step": 9772
    },
    {
      "epoch": 1.8148560817084494,
      "grad_norm": 1.4725546446855635,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8881134986877441,
      "num_tokens": 340031367.0,
      "step": 9773
    },
    {
      "epoch": 1.8150417827298049,
      "grad_norm": 1.4606321556090374,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8628289103507996,
      "num_tokens": 340069732.0,
      "step": 9774
    },
    {
      "epoch": 1.8152274837511606,
      "grad_norm": 1.5304582119766068,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8929527401924133,
      "num_tokens": 340099586.0,
      "step": 9775
    },
    {
      "epoch": 1.8154131847725163,
      "grad_norm": 1.5220527295234898,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8815089464187622,
      "num_tokens": 340134606.0,
      "step": 9776
    },
    {
      "epoch": 1.8155988857938719,
      "grad_norm": 1.567055700275439,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8808919191360474,
      "num_tokens": 340166709.0,
      "step": 9777
    },
    {
      "epoch": 1.8157845868152274,
      "grad_norm": 1.521198137196826,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8792524337768555,
      "num_tokens": 340199721.0,
      "step": 9778
    },
    {
      "epoch": 1.815970287836583,
      "grad_norm": 1.5538297807504098,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.869391679763794,
      "num_tokens": 340231880.0,
      "step": 9779
    },
    {
      "epoch": 1.8161559888579388,
      "grad_norm": 1.4484291596861072,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.8824281692504883,
      "num_tokens": 340268537.0,
      "step": 9780
    },
    {
      "epoch": 1.8163416898792943,
      "grad_norm": 1.3939212837426613,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8870782852172852,
      "num_tokens": 340305192.0,
      "step": 9781
    },
    {
      "epoch": 1.8165273909006499,
      "grad_norm": 1.3434927607409946,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8887931704521179,
      "num_tokens": 340348357.0,
      "step": 9782
    },
    {
      "epoch": 1.8167130919220056,
      "grad_norm": 1.494794988304355,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8889071941375732,
      "num_tokens": 340380105.0,
      "step": 9783
    },
    {
      "epoch": 1.8168987929433613,
      "grad_norm": 1.4615345903257866,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8726882934570312,
      "num_tokens": 340416067.0,
      "step": 9784
    },
    {
      "epoch": 1.8170844939647168,
      "grad_norm": 1.503644951893644,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8904303312301636,
      "num_tokens": 340448836.0,
      "step": 9785
    },
    {
      "epoch": 1.8172701949860723,
      "grad_norm": 1.5818391071620523,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8682610988616943,
      "num_tokens": 340481654.0,
      "step": 9786
    },
    {
      "epoch": 1.817455896007428,
      "grad_norm": 1.4836892262128434,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8738011717796326,
      "num_tokens": 340514922.0,
      "step": 9787
    },
    {
      "epoch": 1.8176415970287838,
      "grad_norm": 1.523753276569345,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8787322044372559,
      "num_tokens": 340549866.0,
      "step": 9788
    },
    {
      "epoch": 1.8178272980501393,
      "grad_norm": 1.6865526216605657,
      "learning_rate": 1e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8689481019973755,
      "num_tokens": 340582013.0,
      "step": 9789
    },
    {
      "epoch": 1.8180129990714948,
      "grad_norm": 1.5244452838907456,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8719348907470703,
      "num_tokens": 340617781.0,
      "step": 9790
    },
    {
      "epoch": 1.8181987000928506,
      "grad_norm": 1.4417453567266962,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8777912855148315,
      "num_tokens": 340653921.0,
      "step": 9791
    },
    {
      "epoch": 1.8183844011142063,
      "grad_norm": 1.466958452366472,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8851044178009033,
      "num_tokens": 340684884.0,
      "step": 9792
    },
    {
      "epoch": 1.8185701021355616,
      "grad_norm": 1.458741111629971,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8777291774749756,
      "num_tokens": 340720793.0,
      "step": 9793
    },
    {
      "epoch": 1.8187558031569173,
      "grad_norm": 1.5501261520655416,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8907351493835449,
      "num_tokens": 340748215.0,
      "step": 9794
    },
    {
      "epoch": 1.818941504178273,
      "grad_norm": 1.5367371768892004,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8893399238586426,
      "num_tokens": 340776151.0,
      "step": 9795
    },
    {
      "epoch": 1.8191272051996286,
      "grad_norm": 1.3911176499850657,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8748345375061035,
      "num_tokens": 340817041.0,
      "step": 9796
    },
    {
      "epoch": 1.819312906220984,
      "grad_norm": 1.4679615626762947,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8811743259429932,
      "num_tokens": 340850216.0,
      "step": 9797
    },
    {
      "epoch": 1.8194986072423398,
      "grad_norm": 1.537399816856427,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8761404752731323,
      "num_tokens": 340880610.0,
      "step": 9798
    },
    {
      "epoch": 1.8196843082636955,
      "grad_norm": 1.4028501899113006,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8850717544555664,
      "num_tokens": 340916513.0,
      "step": 9799
    },
    {
      "epoch": 1.819870009285051,
      "grad_norm": 1.4681675833700496,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8823403716087341,
      "num_tokens": 340954869.0,
      "step": 9800
    },
    {
      "epoch": 1.8200557103064066,
      "grad_norm": 1.3618200197471833,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.875584602355957,
      "num_tokens": 340992073.0,
      "step": 9801
    },
    {
      "epoch": 1.8202414113277623,
      "grad_norm": 1.500241568567171,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8691809773445129,
      "num_tokens": 341028547.0,
      "step": 9802
    },
    {
      "epoch": 1.820427112349118,
      "grad_norm": 1.3663514954810487,
      "learning_rate": 1e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8713735938072205,
      "num_tokens": 341068947.0,
      "step": 9803
    },
    {
      "epoch": 1.8206128133704735,
      "grad_norm": 1.3310199025285652,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8771656155586243,
      "num_tokens": 341112346.0,
      "step": 9804
    },
    {
      "epoch": 1.820798514391829,
      "grad_norm": 1.4105969008162298,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8786356449127197,
      "num_tokens": 341151635.0,
      "step": 9805
    },
    {
      "epoch": 1.8209842154131848,
      "grad_norm": 1.4760781178813414,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8825557231903076,
      "num_tokens": 341188248.0,
      "step": 9806
    },
    {
      "epoch": 1.8211699164345405,
      "grad_norm": 1.3993681468883845,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8780139088630676,
      "num_tokens": 341224841.0,
      "step": 9807
    },
    {
      "epoch": 1.821355617455896,
      "grad_norm": 1.4511862704981084,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8717916011810303,
      "num_tokens": 341262429.0,
      "step": 9808
    },
    {
      "epoch": 1.8215413184772515,
      "grad_norm": 1.3971963622805368,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8867009878158569,
      "num_tokens": 341298041.0,
      "step": 9809
    },
    {
      "epoch": 1.8217270194986073,
      "grad_norm": 1.5222200309005463,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8904015421867371,
      "num_tokens": 341327866.0,
      "step": 9810
    },
    {
      "epoch": 1.821912720519963,
      "grad_norm": 1.405620216699826,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8847286105155945,
      "num_tokens": 341362571.0,
      "step": 9811
    },
    {
      "epoch": 1.8220984215413185,
      "grad_norm": 1.4754082973537592,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8766101598739624,
      "num_tokens": 341401632.0,
      "step": 9812
    },
    {
      "epoch": 1.822284122562674,
      "grad_norm": 1.4444294583344852,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8790851831436157,
      "num_tokens": 341438509.0,
      "step": 9813
    },
    {
      "epoch": 1.8224698235840298,
      "grad_norm": 1.4908539628683808,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8793549537658691,
      "num_tokens": 341470473.0,
      "step": 9814
    },
    {
      "epoch": 1.8226555246053855,
      "grad_norm": 1.4350735874899152,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.897030234336853,
      "num_tokens": 341501189.0,
      "step": 9815
    },
    {
      "epoch": 1.822841225626741,
      "grad_norm": 1.5468949183559828,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8861551284790039,
      "num_tokens": 341533795.0,
      "step": 9816
    },
    {
      "epoch": 1.8230269266480965,
      "grad_norm": 1.4635484661508584,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8784959316253662,
      "num_tokens": 341569538.0,
      "step": 9817
    },
    {
      "epoch": 1.8232126276694522,
      "grad_norm": 1.5352177193996852,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8717011213302612,
      "num_tokens": 341602470.0,
      "step": 9818
    },
    {
      "epoch": 1.8233983286908078,
      "grad_norm": 1.506176564182583,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8727588057518005,
      "num_tokens": 341639902.0,
      "step": 9819
    },
    {
      "epoch": 1.8235840297121633,
      "grad_norm": 1.5950226151700089,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8718653917312622,
      "num_tokens": 341672225.0,
      "step": 9820
    },
    {
      "epoch": 1.823769730733519,
      "grad_norm": 1.4070106887229474,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8760881423950195,
      "num_tokens": 341712185.0,
      "step": 9821
    },
    {
      "epoch": 1.8239554317548747,
      "grad_norm": 1.3671388954703443,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8903078436851501,
      "num_tokens": 341749794.0,
      "step": 9822
    },
    {
      "epoch": 1.8241411327762302,
      "grad_norm": 1.4904513060952072,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8858740329742432,
      "num_tokens": 341787076.0,
      "step": 9823
    },
    {
      "epoch": 1.8243268337975858,
      "grad_norm": 1.4314178426621473,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8866567611694336,
      "num_tokens": 341821750.0,
      "step": 9824
    },
    {
      "epoch": 1.8245125348189415,
      "grad_norm": 1.5586129087322642,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8779691457748413,
      "num_tokens": 341851401.0,
      "step": 9825
    },
    {
      "epoch": 1.8246982358402972,
      "grad_norm": 1.4169122882563352,
      "learning_rate": 1e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8684812784194946,
      "num_tokens": 341891686.0,
      "step": 9826
    },
    {
      "epoch": 1.8248839368616527,
      "grad_norm": 1.6500181600683852,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8821561932563782,
      "num_tokens": 341921022.0,
      "step": 9827
    },
    {
      "epoch": 1.8250696378830082,
      "grad_norm": 1.6510559432095873,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8884994983673096,
      "num_tokens": 341948041.0,
      "step": 9828
    },
    {
      "epoch": 1.825255338904364,
      "grad_norm": 1.4660982426930578,
      "learning_rate": 1e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8665850162506104,
      "num_tokens": 341984177.0,
      "step": 9829
    },
    {
      "epoch": 1.8254410399257197,
      "grad_norm": 1.4673029539978055,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.879644513130188,
      "num_tokens": 342017266.0,
      "step": 9830
    },
    {
      "epoch": 1.8256267409470752,
      "grad_norm": 1.4086672078453049,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8807190656661987,
      "num_tokens": 342054145.0,
      "step": 9831
    },
    {
      "epoch": 1.8258124419684307,
      "grad_norm": 1.4930856830162014,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.8662229776382446,
      "num_tokens": 342091626.0,
      "step": 9832
    },
    {
      "epoch": 1.8259981429897865,
      "grad_norm": 1.3086583857003058,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8756393790245056,
      "num_tokens": 342136464.0,
      "step": 9833
    },
    {
      "epoch": 1.8261838440111422,
      "grad_norm": 1.6314469469709743,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8673840761184692,
      "num_tokens": 342168238.0,
      "step": 9834
    },
    {
      "epoch": 1.8263695450324977,
      "grad_norm": 1.4934537231057334,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8719750046730042,
      "num_tokens": 342209120.0,
      "step": 9835
    },
    {
      "epoch": 1.8265552460538532,
      "grad_norm": 1.7470052089635708,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8606750965118408,
      "num_tokens": 342240641.0,
      "step": 9836
    },
    {
      "epoch": 1.826740947075209,
      "grad_norm": 1.595236709286758,
      "learning_rate": 1e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8611050844192505,
      "num_tokens": 342273239.0,
      "step": 9837
    },
    {
      "epoch": 1.8269266480965647,
      "grad_norm": 1.507797985358021,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8833268880844116,
      "num_tokens": 342305672.0,
      "step": 9838
    },
    {
      "epoch": 1.8271123491179202,
      "grad_norm": 1.49201238353403,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8887506723403931,
      "num_tokens": 342342834.0,
      "step": 9839
    },
    {
      "epoch": 1.8272980501392757,
      "grad_norm": 1.4450625412057851,
      "learning_rate": 1e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8665285110473633,
      "num_tokens": 342380963.0,
      "step": 9840
    },
    {
      "epoch": 1.8274837511606314,
      "grad_norm": 1.418640414344653,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8730350136756897,
      "num_tokens": 342419153.0,
      "step": 9841
    },
    {
      "epoch": 1.827669452181987,
      "grad_norm": 1.570287581397965,
      "learning_rate": 1e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8617627620697021,
      "num_tokens": 342455830.0,
      "step": 9842
    },
    {
      "epoch": 1.8278551532033425,
      "grad_norm": 1.5086074891579242,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8775278329849243,
      "num_tokens": 342488568.0,
      "step": 9843
    },
    {
      "epoch": 1.8280408542246982,
      "grad_norm": 1.4336424702442252,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8761017918586731,
      "num_tokens": 342524390.0,
      "step": 9844
    },
    {
      "epoch": 1.828226555246054,
      "grad_norm": 1.446770274823116,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8723865747451782,
      "num_tokens": 342562773.0,
      "step": 9845
    },
    {
      "epoch": 1.8284122562674094,
      "grad_norm": 1.4918020345992644,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8700347542762756,
      "num_tokens": 342597611.0,
      "step": 9846
    },
    {
      "epoch": 1.828597957288765,
      "grad_norm": 1.4084523642062676,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8644986152648926,
      "num_tokens": 342637014.0,
      "step": 9847
    },
    {
      "epoch": 1.8287836583101207,
      "grad_norm": 1.464917219226691,
      "learning_rate": 1e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8649858832359314,
      "num_tokens": 342673189.0,
      "step": 9848
    },
    {
      "epoch": 1.8289693593314764,
      "grad_norm": 1.498691882703381,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8840965032577515,
      "num_tokens": 342709782.0,
      "step": 9849
    },
    {
      "epoch": 1.829155060352832,
      "grad_norm": 1.497186420774427,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8783475160598755,
      "num_tokens": 342745618.0,
      "step": 9850
    },
    {
      "epoch": 1.8293407613741874,
      "grad_norm": 1.7399200810907922,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.866921603679657,
      "num_tokens": 342773848.0,
      "step": 9851
    },
    {
      "epoch": 1.8295264623955432,
      "grad_norm": 1.495904013337282,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8802247643470764,
      "num_tokens": 342806794.0,
      "step": 9852
    },
    {
      "epoch": 1.829712163416899,
      "grad_norm": 1.5408740269100591,
      "learning_rate": 1e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8662071228027344,
      "num_tokens": 342839295.0,
      "step": 9853
    },
    {
      "epoch": 1.8298978644382544,
      "grad_norm": 1.570028725016637,
      "learning_rate": 1e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8660897016525269,
      "num_tokens": 342874680.0,
      "step": 9854
    },
    {
      "epoch": 1.83008356545961,
      "grad_norm": 1.5471226007485965,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8757901787757874,
      "num_tokens": 342905271.0,
      "step": 9855
    },
    {
      "epoch": 1.8302692664809657,
      "grad_norm": 1.4572100191861457,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8642367720603943,
      "num_tokens": 342941974.0,
      "step": 9856
    },
    {
      "epoch": 1.8304549675023214,
      "grad_norm": 1.5527719139747207,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8938188552856445,
      "num_tokens": 342972351.0,
      "step": 9857
    },
    {
      "epoch": 1.830640668523677,
      "grad_norm": 1.372590177409544,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8842911124229431,
      "num_tokens": 343007088.0,
      "step": 9858
    },
    {
      "epoch": 1.8308263695450324,
      "grad_norm": 1.373725456373844,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8913459777832031,
      "num_tokens": 343040722.0,
      "step": 9859
    },
    {
      "epoch": 1.8310120705663882,
      "grad_norm": 1.6596912641121373,
      "learning_rate": 1e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8591588139533997,
      "num_tokens": 343069977.0,
      "step": 9860
    },
    {
      "epoch": 1.8311977715877439,
      "grad_norm": 1.5036442832253853,
      "learning_rate": 1e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8649054765701294,
      "num_tokens": 343103294.0,
      "step": 9861
    },
    {
      "epoch": 1.8313834726090994,
      "grad_norm": 1.658678878553166,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8672118186950684,
      "num_tokens": 343130395.0,
      "step": 9862
    },
    {
      "epoch": 1.831569173630455,
      "grad_norm": 1.4621168329145608,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8825581669807434,
      "num_tokens": 343161578.0,
      "step": 9863
    },
    {
      "epoch": 1.8317548746518106,
      "grad_norm": 1.4220531281709792,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8842402696609497,
      "num_tokens": 343198521.0,
      "step": 9864
    },
    {
      "epoch": 1.8319405756731661,
      "grad_norm": 1.700789907160462,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8762014508247375,
      "num_tokens": 343226503.0,
      "step": 9865
    },
    {
      "epoch": 1.8321262766945217,
      "grad_norm": 1.4039352530923541,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8854354619979858,
      "num_tokens": 343263988.0,
      "step": 9866
    },
    {
      "epoch": 1.8323119777158774,
      "grad_norm": 1.4805278292997741,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8769485950469971,
      "num_tokens": 343298420.0,
      "step": 9867
    },
    {
      "epoch": 1.8324976787372331,
      "grad_norm": 1.472706676747141,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8736521005630493,
      "num_tokens": 343332873.0,
      "step": 9868
    },
    {
      "epoch": 1.8326833797585886,
      "grad_norm": 1.3546253986464878,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8854584693908691,
      "num_tokens": 343369179.0,
      "step": 9869
    },
    {
      "epoch": 1.8328690807799441,
      "grad_norm": 1.4845372568785489,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8735747933387756,
      "num_tokens": 343405734.0,
      "step": 9870
    },
    {
      "epoch": 1.8330547818012999,
      "grad_norm": 1.6251887421003448,
      "learning_rate": 1e-06,
      "loss": 0.3871,
      "mean_token_accuracy": 0.869629979133606,
      "num_tokens": 343435958.0,
      "step": 9871
    },
    {
      "epoch": 1.8332404828226556,
      "grad_norm": 1.6968852569284805,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.866479218006134,
      "num_tokens": 343468823.0,
      "step": 9872
    },
    {
      "epoch": 1.8334261838440111,
      "grad_norm": 1.4840013078920722,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8886381387710571,
      "num_tokens": 343500276.0,
      "step": 9873
    },
    {
      "epoch": 1.8336118848653666,
      "grad_norm": 1.5809513989626598,
      "learning_rate": 1e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8585227727890015,
      "num_tokens": 343537111.0,
      "step": 9874
    },
    {
      "epoch": 1.8337975858867224,
      "grad_norm": 1.5478862231883448,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8780766725540161,
      "num_tokens": 343571492.0,
      "step": 9875
    },
    {
      "epoch": 1.833983286908078,
      "grad_norm": 1.4684878362287022,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8849714994430542,
      "num_tokens": 343602640.0,
      "step": 9876
    },
    {
      "epoch": 1.8341689879294336,
      "grad_norm": 1.7282273955749325,
      "learning_rate": 1e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8540427684783936,
      "num_tokens": 343631816.0,
      "step": 9877
    },
    {
      "epoch": 1.8343546889507891,
      "grad_norm": 1.3709709705710562,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8795082569122314,
      "num_tokens": 343671466.0,
      "step": 9878
    },
    {
      "epoch": 1.8345403899721449,
      "grad_norm": 1.5234188494340044,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8830385804176331,
      "num_tokens": 343708208.0,
      "step": 9879
    },
    {
      "epoch": 1.8347260909935006,
      "grad_norm": 1.4281541397269701,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8796721696853638,
      "num_tokens": 343745495.0,
      "step": 9880
    },
    {
      "epoch": 1.834911792014856,
      "grad_norm": 1.5728480792966257,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8724308013916016,
      "num_tokens": 343776752.0,
      "step": 9881
    },
    {
      "epoch": 1.8350974930362116,
      "grad_norm": 1.5549307375454613,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8818867206573486,
      "num_tokens": 343806921.0,
      "step": 9882
    },
    {
      "epoch": 1.8352831940575673,
      "grad_norm": 1.4770890690764191,
      "learning_rate": 1e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8538000583648682,
      "num_tokens": 343848716.0,
      "step": 9883
    },
    {
      "epoch": 1.835468895078923,
      "grad_norm": 1.5109529005296412,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8861623406410217,
      "num_tokens": 343880301.0,
      "step": 9884
    },
    {
      "epoch": 1.8356545961002786,
      "grad_norm": 1.6376061403871986,
      "learning_rate": 1e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8665147423744202,
      "num_tokens": 343911403.0,
      "step": 9885
    },
    {
      "epoch": 1.835840297121634,
      "grad_norm": 1.4899869625149587,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8756645917892456,
      "num_tokens": 343947965.0,
      "step": 9886
    },
    {
      "epoch": 1.8360259981429898,
      "grad_norm": 1.4394169679532605,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.878322958946228,
      "num_tokens": 343986215.0,
      "step": 9887
    },
    {
      "epoch": 1.8362116991643456,
      "grad_norm": 1.4795772142503953,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8801419138908386,
      "num_tokens": 344023100.0,
      "step": 9888
    },
    {
      "epoch": 1.8363974001857009,
      "grad_norm": 1.5934337313775195,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8819254636764526,
      "num_tokens": 344054853.0,
      "step": 9889
    },
    {
      "epoch": 1.8365831012070566,
      "grad_norm": 1.5026232765941665,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8819378018379211,
      "num_tokens": 344089171.0,
      "step": 9890
    },
    {
      "epoch": 1.8367688022284123,
      "grad_norm": 1.5744221611225973,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8738924860954285,
      "num_tokens": 344126090.0,
      "step": 9891
    },
    {
      "epoch": 1.8369545032497678,
      "grad_norm": 1.4781316331027632,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8937118053436279,
      "num_tokens": 344159507.0,
      "step": 9892
    },
    {
      "epoch": 1.8371402042711233,
      "grad_norm": 1.56759820647463,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8894912004470825,
      "num_tokens": 344188410.0,
      "step": 9893
    },
    {
      "epoch": 1.837325905292479,
      "grad_norm": 1.7211202462621984,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8841651678085327,
      "num_tokens": 344213779.0,
      "step": 9894
    },
    {
      "epoch": 1.8375116063138348,
      "grad_norm": 1.582413504691449,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8803718090057373,
      "num_tokens": 344246128.0,
      "step": 9895
    },
    {
      "epoch": 1.8376973073351903,
      "grad_norm": 1.4588353451979519,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8884970545768738,
      "num_tokens": 344280809.0,
      "step": 9896
    },
    {
      "epoch": 1.8378830083565458,
      "grad_norm": 1.7538886261222482,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8872369527816772,
      "num_tokens": 344313508.0,
      "step": 9897
    },
    {
      "epoch": 1.8380687093779016,
      "grad_norm": 1.5551764987942873,
      "learning_rate": 1e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8771709203720093,
      "num_tokens": 344346938.0,
      "step": 9898
    },
    {
      "epoch": 1.8382544103992573,
      "grad_norm": 1.4598702125236829,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8785926103591919,
      "num_tokens": 344386934.0,
      "step": 9899
    },
    {
      "epoch": 1.8384401114206128,
      "grad_norm": 1.4998902159960303,
      "learning_rate": 1e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8675446510314941,
      "num_tokens": 344426061.0,
      "step": 9900
    },
    {
      "epoch": 1.8386258124419683,
      "grad_norm": 1.404290311815885,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8951758146286011,
      "num_tokens": 344459563.0,
      "step": 9901
    },
    {
      "epoch": 1.838811513463324,
      "grad_norm": 1.4484190924339424,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8814459443092346,
      "num_tokens": 344494318.0,
      "step": 9902
    },
    {
      "epoch": 1.8389972144846798,
      "grad_norm": 1.5428076700057387,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8722222447395325,
      "num_tokens": 344524471.0,
      "step": 9903
    },
    {
      "epoch": 1.8391829155060353,
      "grad_norm": 1.4899114983327684,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8777493238449097,
      "num_tokens": 344557688.0,
      "step": 9904
    },
    {
      "epoch": 1.8393686165273908,
      "grad_norm": 1.580045767753519,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8719829320907593,
      "num_tokens": 344590794.0,
      "step": 9905
    },
    {
      "epoch": 1.8395543175487465,
      "grad_norm": 1.4321070290719857,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8770790100097656,
      "num_tokens": 344631234.0,
      "step": 9906
    },
    {
      "epoch": 1.8397400185701023,
      "grad_norm": 1.5511974403122417,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8712599873542786,
      "num_tokens": 344666983.0,
      "step": 9907
    },
    {
      "epoch": 1.8399257195914578,
      "grad_norm": 1.549379171811646,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8871344327926636,
      "num_tokens": 344696764.0,
      "step": 9908
    },
    {
      "epoch": 1.8401114206128133,
      "grad_norm": 1.409646355386981,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.876822829246521,
      "num_tokens": 344730543.0,
      "step": 9909
    },
    {
      "epoch": 1.840297121634169,
      "grad_norm": 1.5221965083074829,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.882381796836853,
      "num_tokens": 344762827.0,
      "step": 9910
    },
    {
      "epoch": 1.8404828226555248,
      "grad_norm": 1.5326660061607449,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8850537538528442,
      "num_tokens": 344793676.0,
      "step": 9911
    },
    {
      "epoch": 1.8406685236768803,
      "grad_norm": 1.579492146059876,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8731774091720581,
      "num_tokens": 344825673.0,
      "step": 9912
    },
    {
      "epoch": 1.8408542246982358,
      "grad_norm": 1.4530953438701597,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8796823024749756,
      "num_tokens": 344861495.0,
      "step": 9913
    },
    {
      "epoch": 1.8410399257195915,
      "grad_norm": 1.401077585290941,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8906851410865784,
      "num_tokens": 344894439.0,
      "step": 9914
    },
    {
      "epoch": 1.841225626740947,
      "grad_norm": 1.4563855324180828,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.886342465877533,
      "num_tokens": 344931911.0,
      "step": 9915
    },
    {
      "epoch": 1.8414113277623025,
      "grad_norm": 1.480457761254662,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8932114839553833,
      "num_tokens": 344965109.0,
      "step": 9916
    },
    {
      "epoch": 1.8415970287836583,
      "grad_norm": 1.5129651520516054,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8674227595329285,
      "num_tokens": 344996756.0,
      "step": 9917
    },
    {
      "epoch": 1.841782729805014,
      "grad_norm": 1.4649649101086941,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8906453847885132,
      "num_tokens": 345031207.0,
      "step": 9918
    },
    {
      "epoch": 1.8419684308263695,
      "grad_norm": 1.6191217674280676,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8644594550132751,
      "num_tokens": 345062923.0,
      "step": 9919
    },
    {
      "epoch": 1.842154131847725,
      "grad_norm": 1.4035868817102206,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8768593072891235,
      "num_tokens": 345102644.0,
      "step": 9920
    },
    {
      "epoch": 1.8423398328690808,
      "grad_norm": 1.5257323649226806,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8770661354064941,
      "num_tokens": 345134923.0,
      "step": 9921
    },
    {
      "epoch": 1.8425255338904365,
      "grad_norm": 1.370018534929136,
      "learning_rate": 1e-06,
      "loss": 0.3699,
      "mean_token_accuracy": 0.8689702153205872,
      "num_tokens": 345181576.0,
      "step": 9922
    },
    {
      "epoch": 1.842711234911792,
      "grad_norm": 1.5100204541003768,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8783078193664551,
      "num_tokens": 345223693.0,
      "step": 9923
    },
    {
      "epoch": 1.8428969359331475,
      "grad_norm": 1.4502524015673937,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8803010582923889,
      "num_tokens": 345257495.0,
      "step": 9924
    },
    {
      "epoch": 1.8430826369545033,
      "grad_norm": 1.4020174915833987,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8747206926345825,
      "num_tokens": 345296277.0,
      "step": 9925
    },
    {
      "epoch": 1.843268337975859,
      "grad_norm": 1.352410938837917,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8922521471977234,
      "num_tokens": 345334447.0,
      "step": 9926
    },
    {
      "epoch": 1.8434540389972145,
      "grad_norm": 1.3855623983116883,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8749415278434753,
      "num_tokens": 345374733.0,
      "step": 9927
    },
    {
      "epoch": 1.84363974001857,
      "grad_norm": 1.4461050687867147,
      "learning_rate": 1e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8694481253623962,
      "num_tokens": 345409975.0,
      "step": 9928
    },
    {
      "epoch": 1.8438254410399257,
      "grad_norm": 1.2675349324737393,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8891702890396118,
      "num_tokens": 345451273.0,
      "step": 9929
    },
    {
      "epoch": 1.8440111420612815,
      "grad_norm": 1.4192480739478943,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8870469927787781,
      "num_tokens": 345489517.0,
      "step": 9930
    },
    {
      "epoch": 1.844196843082637,
      "grad_norm": 1.5325406125103385,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8738435506820679,
      "num_tokens": 345521157.0,
      "step": 9931
    },
    {
      "epoch": 1.8443825441039925,
      "grad_norm": 1.4809069239897465,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8808298707008362,
      "num_tokens": 345556969.0,
      "step": 9932
    },
    {
      "epoch": 1.8445682451253482,
      "grad_norm": 1.3382130733706352,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.885040819644928,
      "num_tokens": 345598585.0,
      "step": 9933
    },
    {
      "epoch": 1.844753946146704,
      "grad_norm": 1.5320263317706335,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8783818483352661,
      "num_tokens": 345631069.0,
      "step": 9934
    },
    {
      "epoch": 1.8449396471680595,
      "grad_norm": 1.489452744396945,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8863364458084106,
      "num_tokens": 345667369.0,
      "step": 9935
    },
    {
      "epoch": 1.845125348189415,
      "grad_norm": 1.4624064520410036,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8645745515823364,
      "num_tokens": 345710157.0,
      "step": 9936
    },
    {
      "epoch": 1.8453110492107707,
      "grad_norm": 1.3121442099682346,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8834037184715271,
      "num_tokens": 345754393.0,
      "step": 9937
    },
    {
      "epoch": 1.8454967502321262,
      "grad_norm": 1.5355374091889742,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8961199522018433,
      "num_tokens": 345782278.0,
      "step": 9938
    },
    {
      "epoch": 1.8456824512534817,
      "grad_norm": 1.3897056758090076,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.890519380569458,
      "num_tokens": 345817511.0,
      "step": 9939
    },
    {
      "epoch": 1.8458681522748375,
      "grad_norm": 1.5822726381178327,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.887785792350769,
      "num_tokens": 345845544.0,
      "step": 9940
    },
    {
      "epoch": 1.8460538532961932,
      "grad_norm": 1.6018577130417784,
      "learning_rate": 1e-06,
      "loss": 0.3957,
      "mean_token_accuracy": 0.8618998527526855,
      "num_tokens": 345874363.0,
      "step": 9941
    },
    {
      "epoch": 1.8462395543175487,
      "grad_norm": 1.493562483486375,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8753475546836853,
      "num_tokens": 345913688.0,
      "step": 9942
    },
    {
      "epoch": 1.8464252553389042,
      "grad_norm": 1.512797149827285,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8761378526687622,
      "num_tokens": 345954021.0,
      "step": 9943
    },
    {
      "epoch": 1.84661095636026,
      "grad_norm": 1.4851886542793606,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8849252462387085,
      "num_tokens": 345989365.0,
      "step": 9944
    },
    {
      "epoch": 1.8467966573816157,
      "grad_norm": 1.421303587408595,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.874160885810852,
      "num_tokens": 346030648.0,
      "step": 9945
    },
    {
      "epoch": 1.8469823584029712,
      "grad_norm": 1.542804811051684,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8879523277282715,
      "num_tokens": 346060618.0,
      "step": 9946
    },
    {
      "epoch": 1.8471680594243267,
      "grad_norm": 1.5341475949009604,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8819664716720581,
      "num_tokens": 346100750.0,
      "step": 9947
    },
    {
      "epoch": 1.8473537604456824,
      "grad_norm": 1.443403088329909,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8806654214859009,
      "num_tokens": 346136523.0,
      "step": 9948
    },
    {
      "epoch": 1.8475394614670382,
      "grad_norm": 1.4912175246341752,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8695809841156006,
      "num_tokens": 346174561.0,
      "step": 9949
    },
    {
      "epoch": 1.8477251624883937,
      "grad_norm": 1.506252369864624,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8778543472290039,
      "num_tokens": 346211883.0,
      "step": 9950
    },
    {
      "epoch": 1.8479108635097492,
      "grad_norm": 1.4670685913810382,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8815381526947021,
      "num_tokens": 346247422.0,
      "step": 9951
    },
    {
      "epoch": 1.848096564531105,
      "grad_norm": 1.4629299325148593,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8891035318374634,
      "num_tokens": 346281579.0,
      "step": 9952
    },
    {
      "epoch": 1.8482822655524607,
      "grad_norm": 1.4806842794704973,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8733745813369751,
      "num_tokens": 346317167.0,
      "step": 9953
    },
    {
      "epoch": 1.8484679665738162,
      "grad_norm": 1.6674098549701857,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8805932998657227,
      "num_tokens": 346342963.0,
      "step": 9954
    },
    {
      "epoch": 1.8486536675951717,
      "grad_norm": 1.3727322923856966,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8793036937713623,
      "num_tokens": 346383918.0,
      "step": 9955
    },
    {
      "epoch": 1.8488393686165274,
      "grad_norm": 1.5461633551555691,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8854759931564331,
      "num_tokens": 346414711.0,
      "step": 9956
    },
    {
      "epoch": 1.8490250696378832,
      "grad_norm": 1.4240737725420884,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8846207857131958,
      "num_tokens": 346453668.0,
      "step": 9957
    },
    {
      "epoch": 1.8492107706592387,
      "grad_norm": 1.4262537409689644,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8737631440162659,
      "num_tokens": 346492402.0,
      "step": 9958
    },
    {
      "epoch": 1.8493964716805942,
      "grad_norm": 1.4799151659696328,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.885267436504364,
      "num_tokens": 346524376.0,
      "step": 9959
    },
    {
      "epoch": 1.84958217270195,
      "grad_norm": 1.6122743759666067,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8827862739562988,
      "num_tokens": 346555765.0,
      "step": 9960
    },
    {
      "epoch": 1.8497678737233056,
      "grad_norm": 1.5020536434196572,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8788696527481079,
      "num_tokens": 346591059.0,
      "step": 9961
    },
    {
      "epoch": 1.849953574744661,
      "grad_norm": 1.4871773420282557,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8810727596282959,
      "num_tokens": 346627815.0,
      "step": 9962
    },
    {
      "epoch": 1.8501392757660167,
      "grad_norm": 1.5647403655102867,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8705702424049377,
      "num_tokens": 346663522.0,
      "step": 9963
    },
    {
      "epoch": 1.8503249767873724,
      "grad_norm": 1.5459006405864193,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8813517689704895,
      "num_tokens": 346699734.0,
      "step": 9964
    },
    {
      "epoch": 1.850510677808728,
      "grad_norm": 1.4589583846293555,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8797298669815063,
      "num_tokens": 346739639.0,
      "step": 9965
    },
    {
      "epoch": 1.8506963788300834,
      "grad_norm": 1.3939699985108485,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8821520805358887,
      "num_tokens": 346780079.0,
      "step": 9966
    },
    {
      "epoch": 1.8508820798514392,
      "grad_norm": 1.5839229017215815,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8789958357810974,
      "num_tokens": 346811419.0,
      "step": 9967
    },
    {
      "epoch": 1.8510677808727949,
      "grad_norm": 1.4490110091826947,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8709156513214111,
      "num_tokens": 346848417.0,
      "step": 9968
    },
    {
      "epoch": 1.8512534818941504,
      "grad_norm": 1.3820515793520884,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8912628293037415,
      "num_tokens": 346884291.0,
      "step": 9969
    },
    {
      "epoch": 1.851439182915506,
      "grad_norm": 1.4588112982892523,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8869691491127014,
      "num_tokens": 346918076.0,
      "step": 9970
    },
    {
      "epoch": 1.8516248839368616,
      "grad_norm": 1.4562921221564025,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8949825167655945,
      "num_tokens": 346951806.0,
      "step": 9971
    },
    {
      "epoch": 1.8518105849582174,
      "grad_norm": 1.4668445016849792,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8685296773910522,
      "num_tokens": 346990162.0,
      "step": 9972
    },
    {
      "epoch": 1.8519962859795729,
      "grad_norm": 1.4844230882659428,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8796507716178894,
      "num_tokens": 347025807.0,
      "step": 9973
    },
    {
      "epoch": 1.8521819870009284,
      "grad_norm": 1.537241847337382,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8826765418052673,
      "num_tokens": 347056608.0,
      "step": 9974
    },
    {
      "epoch": 1.8523676880222841,
      "grad_norm": 1.5224128225505338,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8759983777999878,
      "num_tokens": 347090334.0,
      "step": 9975
    },
    {
      "epoch": 1.8525533890436399,
      "grad_norm": 1.4053777130422425,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8894279599189758,
      "num_tokens": 347127947.0,
      "step": 9976
    },
    {
      "epoch": 1.8527390900649954,
      "grad_norm": 1.4903541776331108,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8914958238601685,
      "num_tokens": 347157106.0,
      "step": 9977
    },
    {
      "epoch": 1.8529247910863509,
      "grad_norm": 1.3919738436581397,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8726416230201721,
      "num_tokens": 347193662.0,
      "step": 9978
    },
    {
      "epoch": 1.8531104921077066,
      "grad_norm": 1.6584886490819397,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8817965984344482,
      "num_tokens": 347232112.0,
      "step": 9979
    },
    {
      "epoch": 1.8532961931290624,
      "grad_norm": 1.458526646403853,
      "learning_rate": 1e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8702741861343384,
      "num_tokens": 347268757.0,
      "step": 9980
    },
    {
      "epoch": 1.8534818941504179,
      "grad_norm": 1.492457128232949,
      "learning_rate": 1e-06,
      "loss": 0.3952,
      "mean_token_accuracy": 0.8619204759597778,
      "num_tokens": 347306772.0,
      "step": 9981
    },
    {
      "epoch": 1.8536675951717734,
      "grad_norm": 1.665870507060786,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.866400957107544,
      "num_tokens": 347338168.0,
      "step": 9982
    },
    {
      "epoch": 1.853853296193129,
      "grad_norm": 1.6139327325450314,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8734285235404968,
      "num_tokens": 347368229.0,
      "step": 9983
    },
    {
      "epoch": 1.8540389972144848,
      "grad_norm": 1.5468673086064495,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8753494024276733,
      "num_tokens": 347401643.0,
      "step": 9984
    },
    {
      "epoch": 1.8542246982358404,
      "grad_norm": 1.2890813866608506,
      "learning_rate": 1e-06,
      "loss": 0.3098,
      "mean_token_accuracy": 0.890112042427063,
      "num_tokens": 347440438.0,
      "step": 9985
    },
    {
      "epoch": 1.8544103992571959,
      "grad_norm": 1.429975913775781,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.884385347366333,
      "num_tokens": 347478300.0,
      "step": 9986
    },
    {
      "epoch": 1.8545961002785516,
      "grad_norm": 1.4879270129212077,
      "learning_rate": 1e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8679085373878479,
      "num_tokens": 347515928.0,
      "step": 9987
    },
    {
      "epoch": 1.854781801299907,
      "grad_norm": 1.4727250209463556,
      "learning_rate": 1e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.870282769203186,
      "num_tokens": 347551220.0,
      "step": 9988
    },
    {
      "epoch": 1.8549675023212626,
      "grad_norm": 1.6079326220569217,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8762604594230652,
      "num_tokens": 347580657.0,
      "step": 9989
    },
    {
      "epoch": 1.8551532033426184,
      "grad_norm": 1.4198016113133096,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8794816732406616,
      "num_tokens": 347616975.0,
      "step": 9990
    },
    {
      "epoch": 1.855338904363974,
      "grad_norm": 1.4062378763533083,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.874221920967102,
      "num_tokens": 347654957.0,
      "step": 9991
    },
    {
      "epoch": 1.8555246053853296,
      "grad_norm": 1.5256272587191928,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8810751438140869,
      "num_tokens": 347689249.0,
      "step": 9992
    },
    {
      "epoch": 1.855710306406685,
      "grad_norm": 1.6077986154679063,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.871423065662384,
      "num_tokens": 347718569.0,
      "step": 9993
    },
    {
      "epoch": 1.8558960074280408,
      "grad_norm": 1.590515449144576,
      "learning_rate": 1e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8760942816734314,
      "num_tokens": 347751559.0,
      "step": 9994
    },
    {
      "epoch": 1.8560817084493966,
      "grad_norm": 1.4729932999544066,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8692413568496704,
      "num_tokens": 347789222.0,
      "step": 9995
    },
    {
      "epoch": 1.856267409470752,
      "grad_norm": 1.4119821567188136,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8898345232009888,
      "num_tokens": 347830056.0,
      "step": 9996
    },
    {
      "epoch": 1.8564531104921076,
      "grad_norm": 1.7314101642719844,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8874458074569702,
      "num_tokens": 347852052.0,
      "step": 9997
    },
    {
      "epoch": 1.8566388115134633,
      "grad_norm": 1.3686470500215122,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8893836736679077,
      "num_tokens": 347887466.0,
      "step": 9998
    },
    {
      "epoch": 1.856824512534819,
      "grad_norm": 1.4758756129582802,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8821016550064087,
      "num_tokens": 347921641.0,
      "step": 9999
    },
    {
      "epoch": 1.8570102135561746,
      "grad_norm": 1.6798711209136803,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8807435035705566,
      "num_tokens": 347949075.0,
      "step": 10000
    },
    {
      "epoch": 1.85719591457753,
      "grad_norm": 1.9150001327928365,
      "learning_rate": 1e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8548192977905273,
      "num_tokens": 347977684.0,
      "step": 10001
    },
    {
      "epoch": 1.8573816155988858,
      "grad_norm": 1.3987827873621874,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.884065568447113,
      "num_tokens": 348013655.0,
      "step": 10002
    },
    {
      "epoch": 1.8575673166202415,
      "grad_norm": 1.6269523671855821,
      "learning_rate": 1e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8653280735015869,
      "num_tokens": 348046927.0,
      "step": 10003
    },
    {
      "epoch": 1.857753017641597,
      "grad_norm": 1.5190058750546316,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8752484321594238,
      "num_tokens": 348083502.0,
      "step": 10004
    },
    {
      "epoch": 1.8579387186629526,
      "grad_norm": 1.4506832513531969,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8865788578987122,
      "num_tokens": 348119813.0,
      "step": 10005
    },
    {
      "epoch": 1.8581244196843083,
      "grad_norm": 1.5895862658786715,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8719692230224609,
      "num_tokens": 348153278.0,
      "step": 10006
    },
    {
      "epoch": 1.858310120705664,
      "grad_norm": 1.520502589286321,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8841280341148376,
      "num_tokens": 348185751.0,
      "step": 10007
    },
    {
      "epoch": 1.8584958217270195,
      "grad_norm": 1.3537291627560328,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.889995813369751,
      "num_tokens": 348224353.0,
      "step": 10008
    },
    {
      "epoch": 1.858681522748375,
      "grad_norm": 1.6264262321695708,
      "learning_rate": 1e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8620222806930542,
      "num_tokens": 348255444.0,
      "step": 10009
    },
    {
      "epoch": 1.8588672237697308,
      "grad_norm": 1.4744702740981666,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8756457567214966,
      "num_tokens": 348292322.0,
      "step": 10010
    },
    {
      "epoch": 1.8590529247910863,
      "grad_norm": 1.4473636408546253,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8854587078094482,
      "num_tokens": 348331908.0,
      "step": 10011
    },
    {
      "epoch": 1.8592386258124418,
      "grad_norm": 1.5927007757912788,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8768819570541382,
      "num_tokens": 348363519.0,
      "step": 10012
    },
    {
      "epoch": 1.8594243268337975,
      "grad_norm": 1.480422987772228,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8839254379272461,
      "num_tokens": 348398650.0,
      "step": 10013
    },
    {
      "epoch": 1.8596100278551533,
      "grad_norm": 1.4477267282240802,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8824588060379028,
      "num_tokens": 348435022.0,
      "step": 10014
    },
    {
      "epoch": 1.8597957288765088,
      "grad_norm": 1.5532338086306876,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8741394281387329,
      "num_tokens": 348467213.0,
      "step": 10015
    },
    {
      "epoch": 1.8599814298978643,
      "grad_norm": 1.5482627852393929,
      "learning_rate": 1e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8589102625846863,
      "num_tokens": 348502852.0,
      "step": 10016
    },
    {
      "epoch": 1.86016713091922,
      "grad_norm": 1.5375017704205096,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8824146389961243,
      "num_tokens": 348533729.0,
      "step": 10017
    },
    {
      "epoch": 1.8603528319405758,
      "grad_norm": 1.4154462012029507,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8683586120605469,
      "num_tokens": 348573690.0,
      "step": 10018
    },
    {
      "epoch": 1.8605385329619313,
      "grad_norm": 1.667950027099322,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8728735446929932,
      "num_tokens": 348600671.0,
      "step": 10019
    },
    {
      "epoch": 1.8607242339832868,
      "grad_norm": 1.4235855822147157,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8809151649475098,
      "num_tokens": 348636643.0,
      "step": 10020
    },
    {
      "epoch": 1.8609099350046425,
      "grad_norm": 1.5543096703321815,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8884264826774597,
      "num_tokens": 348667955.0,
      "step": 10021
    },
    {
      "epoch": 1.8610956360259983,
      "grad_norm": 1.4632195747215382,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8910953998565674,
      "num_tokens": 348701569.0,
      "step": 10022
    },
    {
      "epoch": 1.8612813370473538,
      "grad_norm": 1.4778138598183774,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8719680309295654,
      "num_tokens": 348736399.0,
      "step": 10023
    },
    {
      "epoch": 1.8614670380687093,
      "grad_norm": 1.4549102818688198,
      "learning_rate": 1e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8682219982147217,
      "num_tokens": 348779105.0,
      "step": 10024
    },
    {
      "epoch": 1.861652739090065,
      "grad_norm": 1.6960934195691786,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8811377286911011,
      "num_tokens": 348804692.0,
      "step": 10025
    },
    {
      "epoch": 1.8618384401114207,
      "grad_norm": 1.6073141007492393,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8755900859832764,
      "num_tokens": 348837872.0,
      "step": 10026
    },
    {
      "epoch": 1.8620241411327763,
      "grad_norm": 1.5921887462249562,
      "learning_rate": 1e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.857846736907959,
      "num_tokens": 348872114.0,
      "step": 10027
    },
    {
      "epoch": 1.8622098421541318,
      "grad_norm": 1.427588158043939,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8752395510673523,
      "num_tokens": 348909910.0,
      "step": 10028
    },
    {
      "epoch": 1.8623955431754875,
      "grad_norm": 1.6529250312595392,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8821579217910767,
      "num_tokens": 348938103.0,
      "step": 10029
    },
    {
      "epoch": 1.8625812441968432,
      "grad_norm": 1.625243805601499,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8771291375160217,
      "num_tokens": 348969471.0,
      "step": 10030
    },
    {
      "epoch": 1.8627669452181987,
      "grad_norm": 1.4457880097557996,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8718293905258179,
      "num_tokens": 349005185.0,
      "step": 10031
    },
    {
      "epoch": 1.8629526462395543,
      "grad_norm": 1.5813908394745424,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8774369955062866,
      "num_tokens": 349036210.0,
      "step": 10032
    },
    {
      "epoch": 1.86313834726091,
      "grad_norm": 1.4658339395621252,
      "learning_rate": 1e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8650385141372681,
      "num_tokens": 349075109.0,
      "step": 10033
    },
    {
      "epoch": 1.8633240482822655,
      "grad_norm": 1.3743053534311125,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8779772520065308,
      "num_tokens": 349116281.0,
      "step": 10034
    },
    {
      "epoch": 1.863509749303621,
      "grad_norm": 1.5244417469797347,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8737898468971252,
      "num_tokens": 349151583.0,
      "step": 10035
    },
    {
      "epoch": 1.8636954503249767,
      "grad_norm": 1.5251078964750042,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8745485544204712,
      "num_tokens": 349187154.0,
      "step": 10036
    },
    {
      "epoch": 1.8638811513463325,
      "grad_norm": 1.5378481719392125,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8693156242370605,
      "num_tokens": 349222191.0,
      "step": 10037
    },
    {
      "epoch": 1.864066852367688,
      "grad_norm": 1.3318517253538107,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8901612758636475,
      "num_tokens": 349261668.0,
      "step": 10038
    },
    {
      "epoch": 1.8642525533890435,
      "grad_norm": 1.3500055943871003,
      "learning_rate": 1e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8750268220901489,
      "num_tokens": 349302559.0,
      "step": 10039
    },
    {
      "epoch": 1.8644382544103992,
      "grad_norm": 1.5871904475110108,
      "learning_rate": 1e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8764438629150391,
      "num_tokens": 349333232.0,
      "step": 10040
    },
    {
      "epoch": 1.864623955431755,
      "grad_norm": 1.566036780748036,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8791146278381348,
      "num_tokens": 349362928.0,
      "step": 10041
    },
    {
      "epoch": 1.8648096564531105,
      "grad_norm": 1.4651604310764683,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8772996664047241,
      "num_tokens": 349400141.0,
      "step": 10042
    },
    {
      "epoch": 1.864995357474466,
      "grad_norm": 1.548300475661046,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8829443454742432,
      "num_tokens": 349434238.0,
      "step": 10043
    },
    {
      "epoch": 1.8651810584958217,
      "grad_norm": 1.4899929705857147,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8792010545730591,
      "num_tokens": 349468561.0,
      "step": 10044
    },
    {
      "epoch": 1.8653667595171775,
      "grad_norm": 1.4911435395450487,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8779085874557495,
      "num_tokens": 349503053.0,
      "step": 10045
    },
    {
      "epoch": 1.865552460538533,
      "grad_norm": 1.5220698487526303,
      "learning_rate": 1e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8693650960922241,
      "num_tokens": 349539230.0,
      "step": 10046
    },
    {
      "epoch": 1.8657381615598885,
      "grad_norm": 1.5383948756904169,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8794554471969604,
      "num_tokens": 349573779.0,
      "step": 10047
    },
    {
      "epoch": 1.8659238625812442,
      "grad_norm": 1.7146790921035604,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.873270571231842,
      "num_tokens": 349602932.0,
      "step": 10048
    },
    {
      "epoch": 1.8661095636026,
      "grad_norm": 1.4975397705949756,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8815230131149292,
      "num_tokens": 349636039.0,
      "step": 10049
    },
    {
      "epoch": 1.8662952646239555,
      "grad_norm": 1.471189647490486,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8820721507072449,
      "num_tokens": 349671200.0,
      "step": 10050
    },
    {
      "epoch": 1.866480965645311,
      "grad_norm": 1.4236167476682473,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8784093856811523,
      "num_tokens": 349709427.0,
      "step": 10051
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 1.5118800433144424,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.883000373840332,
      "num_tokens": 349742003.0,
      "step": 10052
    },
    {
      "epoch": 1.8668523676880224,
      "grad_norm": 1.483474765584275,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8763617277145386,
      "num_tokens": 349778709.0,
      "step": 10053
    },
    {
      "epoch": 1.867038068709378,
      "grad_norm": 1.6146450309587927,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8916664719581604,
      "num_tokens": 349814143.0,
      "step": 10054
    },
    {
      "epoch": 1.8672237697307335,
      "grad_norm": 1.567576701870214,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8820334076881409,
      "num_tokens": 349846081.0,
      "step": 10055
    },
    {
      "epoch": 1.8674094707520892,
      "grad_norm": 1.6360347181792572,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8707320690155029,
      "num_tokens": 349873479.0,
      "step": 10056
    },
    {
      "epoch": 1.867595171773445,
      "grad_norm": 1.5295116924895031,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8886934518814087,
      "num_tokens": 349902539.0,
      "step": 10057
    },
    {
      "epoch": 1.8677808727948002,
      "grad_norm": 1.4258576519014965,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.873531699180603,
      "num_tokens": 349940830.0,
      "step": 10058
    },
    {
      "epoch": 1.867966573816156,
      "grad_norm": 1.4534092789016917,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8813663125038147,
      "num_tokens": 349980404.0,
      "step": 10059
    },
    {
      "epoch": 1.8681522748375117,
      "grad_norm": 1.7045786845841333,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8730430603027344,
      "num_tokens": 350009544.0,
      "step": 10060
    },
    {
      "epoch": 1.8683379758588672,
      "grad_norm": 1.4127797810988711,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8918840885162354,
      "num_tokens": 350042752.0,
      "step": 10061
    },
    {
      "epoch": 1.8685236768802227,
      "grad_norm": 1.3833733340774428,
      "learning_rate": 1e-06,
      "loss": 0.2875,
      "mean_token_accuracy": 0.8982768058776855,
      "num_tokens": 350077423.0,
      "step": 10062
    },
    {
      "epoch": 1.8687093779015784,
      "grad_norm": 1.5972032792927742,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8738415241241455,
      "num_tokens": 350109370.0,
      "step": 10063
    },
    {
      "epoch": 1.8688950789229342,
      "grad_norm": 1.5043267935985773,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8816967606544495,
      "num_tokens": 350139046.0,
      "step": 10064
    },
    {
      "epoch": 1.8690807799442897,
      "grad_norm": 1.5288998879293567,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.869321882724762,
      "num_tokens": 350173553.0,
      "step": 10065
    },
    {
      "epoch": 1.8692664809656452,
      "grad_norm": 1.5160203268301273,
      "learning_rate": 1e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8703904151916504,
      "num_tokens": 350211522.0,
      "step": 10066
    },
    {
      "epoch": 1.869452181987001,
      "grad_norm": 1.5371688910829335,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8829745650291443,
      "num_tokens": 350247505.0,
      "step": 10067
    },
    {
      "epoch": 1.8696378830083566,
      "grad_norm": 1.4018679103958633,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8862369060516357,
      "num_tokens": 350285950.0,
      "step": 10068
    },
    {
      "epoch": 1.8698235840297122,
      "grad_norm": 1.3603789800923989,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8809176683425903,
      "num_tokens": 350324711.0,
      "step": 10069
    },
    {
      "epoch": 1.8700092850510677,
      "grad_norm": 1.3723071693981244,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8706784844398499,
      "num_tokens": 350364301.0,
      "step": 10070
    },
    {
      "epoch": 1.8701949860724234,
      "grad_norm": 1.600340388710274,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8774489760398865,
      "num_tokens": 350394056.0,
      "step": 10071
    },
    {
      "epoch": 1.8703806870937791,
      "grad_norm": 1.57387404021225,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8935911059379578,
      "num_tokens": 350431576.0,
      "step": 10072
    },
    {
      "epoch": 1.8705663881151346,
      "grad_norm": 1.5519891418459522,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8764818906784058,
      "num_tokens": 350465693.0,
      "step": 10073
    },
    {
      "epoch": 1.8707520891364902,
      "grad_norm": 1.509669734038062,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8841537237167358,
      "num_tokens": 350499876.0,
      "step": 10074
    },
    {
      "epoch": 1.870937790157846,
      "grad_norm": 1.3708682358413176,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8859557509422302,
      "num_tokens": 350538596.0,
      "step": 10075
    },
    {
      "epoch": 1.8711234911792016,
      "grad_norm": 1.5546447971176616,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8813172578811646,
      "num_tokens": 350569043.0,
      "step": 10076
    },
    {
      "epoch": 1.8713091922005571,
      "grad_norm": 1.7318027353423961,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8764956593513489,
      "num_tokens": 350596165.0,
      "step": 10077
    },
    {
      "epoch": 1.8714948932219126,
      "grad_norm": 1.7184674429718936,
      "learning_rate": 1e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8530478477478027,
      "num_tokens": 350627671.0,
      "step": 10078
    },
    {
      "epoch": 1.8716805942432684,
      "grad_norm": 1.461823707404492,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8908296823501587,
      "num_tokens": 350661632.0,
      "step": 10079
    },
    {
      "epoch": 1.8718662952646241,
      "grad_norm": 1.5595462353784446,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8892945051193237,
      "num_tokens": 350693086.0,
      "step": 10080
    },
    {
      "epoch": 1.8720519962859796,
      "grad_norm": 1.5912740472466538,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8785549998283386,
      "num_tokens": 350729135.0,
      "step": 10081
    },
    {
      "epoch": 1.8722376973073351,
      "grad_norm": 1.5923116589483854,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.885850191116333,
      "num_tokens": 350755344.0,
      "step": 10082
    },
    {
      "epoch": 1.8724233983286909,
      "grad_norm": 1.5298272471061123,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8910534381866455,
      "num_tokens": 350786159.0,
      "step": 10083
    },
    {
      "epoch": 1.8726090993500464,
      "grad_norm": 1.4811568428176187,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8955003023147583,
      "num_tokens": 350819570.0,
      "step": 10084
    },
    {
      "epoch": 1.872794800371402,
      "grad_norm": 1.4615819379856572,
      "learning_rate": 1e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8699046969413757,
      "num_tokens": 350855951.0,
      "step": 10085
    },
    {
      "epoch": 1.8729805013927576,
      "grad_norm": 1.351433451555748,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8748803734779358,
      "num_tokens": 350898135.0,
      "step": 10086
    },
    {
      "epoch": 1.8731662024141134,
      "grad_norm": 1.3885299125261112,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8783573508262634,
      "num_tokens": 350935216.0,
      "step": 10087
    },
    {
      "epoch": 1.8733519034354689,
      "grad_norm": 1.3650037197371696,
      "learning_rate": 1e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8558614253997803,
      "num_tokens": 350981349.0,
      "step": 10088
    },
    {
      "epoch": 1.8735376044568244,
      "grad_norm": 1.4016675549962223,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8755254149436951,
      "num_tokens": 351022672.0,
      "step": 10089
    },
    {
      "epoch": 1.8737233054781801,
      "grad_norm": 1.4992626931556081,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.878951907157898,
      "num_tokens": 351055214.0,
      "step": 10090
    },
    {
      "epoch": 1.8739090064995358,
      "grad_norm": 1.4260185221703345,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8807608485221863,
      "num_tokens": 351093214.0,
      "step": 10091
    },
    {
      "epoch": 1.8740947075208914,
      "grad_norm": 1.4499037002006527,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8864268064498901,
      "num_tokens": 351127661.0,
      "step": 10092
    },
    {
      "epoch": 1.8742804085422469,
      "grad_norm": 1.7549160089754905,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.864508330821991,
      "num_tokens": 351153036.0,
      "step": 10093
    },
    {
      "epoch": 1.8744661095636026,
      "grad_norm": 1.4852549263087884,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8792161345481873,
      "num_tokens": 351186833.0,
      "step": 10094
    },
    {
      "epoch": 1.8746518105849583,
      "grad_norm": 1.414048756880824,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.882831335067749,
      "num_tokens": 351225063.0,
      "step": 10095
    },
    {
      "epoch": 1.8748375116063138,
      "grad_norm": 1.4076831317688576,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8958649635314941,
      "num_tokens": 351258862.0,
      "step": 10096
    },
    {
      "epoch": 1.8750232126276694,
      "grad_norm": 1.511547557694282,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8923885226249695,
      "num_tokens": 351294752.0,
      "step": 10097
    },
    {
      "epoch": 1.875208913649025,
      "grad_norm": 1.5288484567217586,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8845717906951904,
      "num_tokens": 351327072.0,
      "step": 10098
    },
    {
      "epoch": 1.8753946146703808,
      "grad_norm": 1.4026718002016254,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8790199756622314,
      "num_tokens": 351365625.0,
      "step": 10099
    },
    {
      "epoch": 1.8755803156917363,
      "grad_norm": 1.4732818954173745,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8908572196960449,
      "num_tokens": 351399799.0,
      "step": 10100
    },
    {
      "epoch": 1.8757660167130918,
      "grad_norm": 1.5198533991829206,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8785570859909058,
      "num_tokens": 351432700.0,
      "step": 10101
    },
    {
      "epoch": 1.8759517177344476,
      "grad_norm": 1.3817807715666661,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8864424228668213,
      "num_tokens": 351466634.0,
      "step": 10102
    },
    {
      "epoch": 1.8761374187558033,
      "grad_norm": 1.3678803611206125,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8904130458831787,
      "num_tokens": 351508693.0,
      "step": 10103
    },
    {
      "epoch": 1.8763231197771588,
      "grad_norm": 1.7066217677645608,
      "learning_rate": 1e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8701215982437134,
      "num_tokens": 351537709.0,
      "step": 10104
    },
    {
      "epoch": 1.8765088207985143,
      "grad_norm": 1.5874293532911234,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8709437847137451,
      "num_tokens": 351570350.0,
      "step": 10105
    },
    {
      "epoch": 1.87669452181987,
      "grad_norm": 1.6538954751418005,
      "learning_rate": 1e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8672807216644287,
      "num_tokens": 351601092.0,
      "step": 10106
    },
    {
      "epoch": 1.8768802228412256,
      "grad_norm": 1.506753083340637,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8796308636665344,
      "num_tokens": 351636348.0,
      "step": 10107
    },
    {
      "epoch": 1.877065923862581,
      "grad_norm": 1.5665887554243143,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8815054893493652,
      "num_tokens": 351666078.0,
      "step": 10108
    },
    {
      "epoch": 1.8772516248839368,
      "grad_norm": 1.4401576360564106,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8911294341087341,
      "num_tokens": 351703542.0,
      "step": 10109
    },
    {
      "epoch": 1.8774373259052926,
      "grad_norm": 1.352129445642533,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8872456550598145,
      "num_tokens": 351741411.0,
      "step": 10110
    },
    {
      "epoch": 1.877623026926648,
      "grad_norm": 1.742698056063349,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8766095638275146,
      "num_tokens": 351767281.0,
      "step": 10111
    },
    {
      "epoch": 1.8778087279480036,
      "grad_norm": 1.419404688114952,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8824682235717773,
      "num_tokens": 351803896.0,
      "step": 10112
    },
    {
      "epoch": 1.8779944289693593,
      "grad_norm": 1.5082916307616685,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8810701966285706,
      "num_tokens": 351839800.0,
      "step": 10113
    },
    {
      "epoch": 1.878180129990715,
      "grad_norm": 1.498433333313314,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8772069215774536,
      "num_tokens": 351873316.0,
      "step": 10114
    },
    {
      "epoch": 1.8783658310120706,
      "grad_norm": 1.5923035552625302,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8790667057037354,
      "num_tokens": 351904309.0,
      "step": 10115
    },
    {
      "epoch": 1.878551532033426,
      "grad_norm": 1.3483693730803505,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.893761157989502,
      "num_tokens": 351940689.0,
      "step": 10116
    },
    {
      "epoch": 1.8787372330547818,
      "grad_norm": 1.497343112226875,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8796001672744751,
      "num_tokens": 351974695.0,
      "step": 10117
    },
    {
      "epoch": 1.8789229340761375,
      "grad_norm": 1.500217858974929,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8868869543075562,
      "num_tokens": 352007398.0,
      "step": 10118
    },
    {
      "epoch": 1.879108635097493,
      "grad_norm": 1.5446654715249,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8790233731269836,
      "num_tokens": 352045291.0,
      "step": 10119
    },
    {
      "epoch": 1.8792943361188486,
      "grad_norm": 1.477398279949944,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8900481462478638,
      "num_tokens": 352076995.0,
      "step": 10120
    },
    {
      "epoch": 1.8794800371402043,
      "grad_norm": 1.402883579460917,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.874955952167511,
      "num_tokens": 352117715.0,
      "step": 10121
    },
    {
      "epoch": 1.87966573816156,
      "grad_norm": 1.608018749639368,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8767899870872498,
      "num_tokens": 352149009.0,
      "step": 10122
    },
    {
      "epoch": 1.8798514391829155,
      "grad_norm": 1.4167890685510252,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8893730640411377,
      "num_tokens": 352186288.0,
      "step": 10123
    },
    {
      "epoch": 1.880037140204271,
      "grad_norm": 1.3665160756084316,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8852428793907166,
      "num_tokens": 352227074.0,
      "step": 10124
    },
    {
      "epoch": 1.8802228412256268,
      "grad_norm": 1.4258728661364743,
      "learning_rate": 1e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.8675242066383362,
      "num_tokens": 352266016.0,
      "step": 10125
    },
    {
      "epoch": 1.8804085422469825,
      "grad_norm": 1.4179346219209603,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8848023414611816,
      "num_tokens": 352300529.0,
      "step": 10126
    },
    {
      "epoch": 1.880594243268338,
      "grad_norm": 1.5638108870892522,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8729680776596069,
      "num_tokens": 352336697.0,
      "step": 10127
    },
    {
      "epoch": 1.8807799442896935,
      "grad_norm": 1.4314346738919645,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8899673223495483,
      "num_tokens": 352373616.0,
      "step": 10128
    },
    {
      "epoch": 1.8809656453110493,
      "grad_norm": 1.5671656516973387,
      "learning_rate": 1e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8606942892074585,
      "num_tokens": 352411983.0,
      "step": 10129
    },
    {
      "epoch": 1.881151346332405,
      "grad_norm": 1.4468866004223078,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8885512351989746,
      "num_tokens": 352449552.0,
      "step": 10130
    },
    {
      "epoch": 1.8813370473537603,
      "grad_norm": 1.3764513107641911,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8798027038574219,
      "num_tokens": 352491445.0,
      "step": 10131
    },
    {
      "epoch": 1.881522748375116,
      "grad_norm": 1.6180203913484539,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8668193817138672,
      "num_tokens": 352526715.0,
      "step": 10132
    },
    {
      "epoch": 1.8817084493964717,
      "grad_norm": 1.3192986232205004,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8773030638694763,
      "num_tokens": 352568994.0,
      "step": 10133
    },
    {
      "epoch": 1.8818941504178273,
      "grad_norm": 1.4119938335208997,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8757321834564209,
      "num_tokens": 352607023.0,
      "step": 10134
    },
    {
      "epoch": 1.8820798514391828,
      "grad_norm": 1.3585531044649304,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8835077285766602,
      "num_tokens": 352647819.0,
      "step": 10135
    },
    {
      "epoch": 1.8822655524605385,
      "grad_norm": 1.5523656401261559,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8806257843971252,
      "num_tokens": 352678862.0,
      "step": 10136
    },
    {
      "epoch": 1.8824512534818942,
      "grad_norm": 1.5623770462904647,
      "learning_rate": 1e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8764492869377136,
      "num_tokens": 352713001.0,
      "step": 10137
    },
    {
      "epoch": 1.8826369545032497,
      "grad_norm": 1.4217514002696645,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8855681419372559,
      "num_tokens": 352752683.0,
      "step": 10138
    },
    {
      "epoch": 1.8828226555246053,
      "grad_norm": 1.5607455086751298,
      "learning_rate": 1e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.8660954833030701,
      "num_tokens": 352787788.0,
      "step": 10139
    },
    {
      "epoch": 1.883008356545961,
      "grad_norm": 1.5261336673164967,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.8940251469612122,
      "num_tokens": 352817894.0,
      "step": 10140
    },
    {
      "epoch": 1.8831940575673167,
      "grad_norm": 1.4310367790731937,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8682190775871277,
      "num_tokens": 352857648.0,
      "step": 10141
    },
    {
      "epoch": 1.8833797585886722,
      "grad_norm": 1.5489024815391348,
      "learning_rate": 1e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8658356070518494,
      "num_tokens": 352892736.0,
      "step": 10142
    },
    {
      "epoch": 1.8835654596100277,
      "grad_norm": 1.4447466478237205,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8859879374504089,
      "num_tokens": 352927795.0,
      "step": 10143
    },
    {
      "epoch": 1.8837511606313835,
      "grad_norm": 1.4418973153211256,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8889763355255127,
      "num_tokens": 352959040.0,
      "step": 10144
    },
    {
      "epoch": 1.8839368616527392,
      "grad_norm": 1.420369459604838,
      "learning_rate": 1e-06,
      "loss": 0.3114,
      "mean_token_accuracy": 0.8904685974121094,
      "num_tokens": 352993748.0,
      "step": 10145
    },
    {
      "epoch": 1.8841225626740947,
      "grad_norm": 1.5027981196185705,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8807136416435242,
      "num_tokens": 353028253.0,
      "step": 10146
    },
    {
      "epoch": 1.8843082636954502,
      "grad_norm": 1.6058825764226066,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8699583411216736,
      "num_tokens": 353057310.0,
      "step": 10147
    },
    {
      "epoch": 1.884493964716806,
      "grad_norm": 1.5531308548311835,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8850682973861694,
      "num_tokens": 353085609.0,
      "step": 10148
    },
    {
      "epoch": 1.8846796657381617,
      "grad_norm": 1.3738954134869485,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8752874732017517,
      "num_tokens": 353128827.0,
      "step": 10149
    },
    {
      "epoch": 1.8848653667595172,
      "grad_norm": 1.5773223241317618,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8757742047309875,
      "num_tokens": 353165326.0,
      "step": 10150
    },
    {
      "epoch": 1.8850510677808727,
      "grad_norm": 1.5537147801530407,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.88588547706604,
      "num_tokens": 353196023.0,
      "step": 10151
    },
    {
      "epoch": 1.8852367688022285,
      "grad_norm": 1.5971392828635553,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8761394619941711,
      "num_tokens": 353229063.0,
      "step": 10152
    },
    {
      "epoch": 1.8854224698235842,
      "grad_norm": 1.6544490239786345,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.892060399055481,
      "num_tokens": 353255605.0,
      "step": 10153
    },
    {
      "epoch": 1.8856081708449397,
      "grad_norm": 1.5244078404424337,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8826122283935547,
      "num_tokens": 353292989.0,
      "step": 10154
    },
    {
      "epoch": 1.8857938718662952,
      "grad_norm": 1.442960678771425,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8904467225074768,
      "num_tokens": 353329982.0,
      "step": 10155
    },
    {
      "epoch": 1.885979572887651,
      "grad_norm": 1.385074264763532,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8860671520233154,
      "num_tokens": 353368579.0,
      "step": 10156
    },
    {
      "epoch": 1.8861652739090065,
      "grad_norm": 1.420842272213704,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8856217265129089,
      "num_tokens": 353406521.0,
      "step": 10157
    },
    {
      "epoch": 1.886350974930362,
      "grad_norm": 1.4699737704711737,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8779029250144958,
      "num_tokens": 353440336.0,
      "step": 10158
    },
    {
      "epoch": 1.8865366759517177,
      "grad_norm": 1.4717630929673462,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8771901726722717,
      "num_tokens": 353480812.0,
      "step": 10159
    },
    {
      "epoch": 1.8867223769730734,
      "grad_norm": 1.5175523280188825,
      "learning_rate": 1e-06,
      "loss": 0.3027,
      "mean_token_accuracy": 0.8936712741851807,
      "num_tokens": 353515587.0,
      "step": 10160
    },
    {
      "epoch": 1.886908077994429,
      "grad_norm": 1.6960821401682908,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8760436773300171,
      "num_tokens": 353546502.0,
      "step": 10161
    },
    {
      "epoch": 1.8870937790157845,
      "grad_norm": 1.3439895450608668,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.886844277381897,
      "num_tokens": 353584111.0,
      "step": 10162
    },
    {
      "epoch": 1.8872794800371402,
      "grad_norm": 1.4784390798655809,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8811147212982178,
      "num_tokens": 353620037.0,
      "step": 10163
    },
    {
      "epoch": 1.887465181058496,
      "grad_norm": 1.3967824108376867,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8791754245758057,
      "num_tokens": 353658161.0,
      "step": 10164
    },
    {
      "epoch": 1.8876508820798514,
      "grad_norm": 1.6737522888329943,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8771445751190186,
      "num_tokens": 353687281.0,
      "step": 10165
    },
    {
      "epoch": 1.887836583101207,
      "grad_norm": 1.3304943769064377,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8747531771659851,
      "num_tokens": 353734932.0,
      "step": 10166
    },
    {
      "epoch": 1.8880222841225627,
      "grad_norm": 1.6479139656078439,
      "learning_rate": 1e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8611243367195129,
      "num_tokens": 353768282.0,
      "step": 10167
    },
    {
      "epoch": 1.8882079851439184,
      "grad_norm": 1.3944855251592863,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8837107419967651,
      "num_tokens": 353804459.0,
      "step": 10168
    },
    {
      "epoch": 1.888393686165274,
      "grad_norm": 1.5613568714760384,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8757383823394775,
      "num_tokens": 353837138.0,
      "step": 10169
    },
    {
      "epoch": 1.8885793871866294,
      "grad_norm": 1.5387758310276036,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.87556391954422,
      "num_tokens": 353872903.0,
      "step": 10170
    },
    {
      "epoch": 1.8887650882079852,
      "grad_norm": 1.486500696656961,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8794279098510742,
      "num_tokens": 353907497.0,
      "step": 10171
    },
    {
      "epoch": 1.888950789229341,
      "grad_norm": 1.4702317311226485,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8760626316070557,
      "num_tokens": 353940756.0,
      "step": 10172
    },
    {
      "epoch": 1.8891364902506964,
      "grad_norm": 1.4193417515326179,
      "learning_rate": 1e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8661491870880127,
      "num_tokens": 353979794.0,
      "step": 10173
    },
    {
      "epoch": 1.889322191272052,
      "grad_norm": 1.561290422919943,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8743579387664795,
      "num_tokens": 354014505.0,
      "step": 10174
    },
    {
      "epoch": 1.8895078922934077,
      "grad_norm": 1.4572103064207667,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.871353030204773,
      "num_tokens": 354053258.0,
      "step": 10175
    },
    {
      "epoch": 1.8896935933147634,
      "grad_norm": 1.4761713830213041,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8873673677444458,
      "num_tokens": 354086249.0,
      "step": 10176
    },
    {
      "epoch": 1.889879294336119,
      "grad_norm": 1.625745582541674,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.875669538974762,
      "num_tokens": 354115916.0,
      "step": 10177
    },
    {
      "epoch": 1.8900649953574744,
      "grad_norm": 1.4584405183417926,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8709942102432251,
      "num_tokens": 354154983.0,
      "step": 10178
    },
    {
      "epoch": 1.8902506963788301,
      "grad_norm": 1.6043293439357613,
      "learning_rate": 1e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8513097763061523,
      "num_tokens": 354194757.0,
      "step": 10179
    },
    {
      "epoch": 1.8904363974001857,
      "grad_norm": 1.769435745665925,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8722657561302185,
      "num_tokens": 354224515.0,
      "step": 10180
    },
    {
      "epoch": 1.8906220984215412,
      "grad_norm": 1.452285444968318,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8769757747650146,
      "num_tokens": 354263260.0,
      "step": 10181
    },
    {
      "epoch": 1.890807799442897,
      "grad_norm": 1.546879504247275,
      "learning_rate": 1e-06,
      "loss": 0.3857,
      "mean_token_accuracy": 0.8646972179412842,
      "num_tokens": 354299756.0,
      "step": 10182
    },
    {
      "epoch": 1.8909935004642526,
      "grad_norm": 1.5390319648065975,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8703833818435669,
      "num_tokens": 354333672.0,
      "step": 10183
    },
    {
      "epoch": 1.8911792014856081,
      "grad_norm": 1.6401441940557138,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8832986354827881,
      "num_tokens": 354365360.0,
      "step": 10184
    },
    {
      "epoch": 1.8913649025069637,
      "grad_norm": 1.4742516160798402,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8751665353775024,
      "num_tokens": 354401153.0,
      "step": 10185
    },
    {
      "epoch": 1.8915506035283194,
      "grad_norm": 1.5730034274509013,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8853565454483032,
      "num_tokens": 354425867.0,
      "step": 10186
    },
    {
      "epoch": 1.8917363045496751,
      "grad_norm": 1.5192534938981272,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8782092332839966,
      "num_tokens": 354458041.0,
      "step": 10187
    },
    {
      "epoch": 1.8919220055710306,
      "grad_norm": 1.627125749227938,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8630639910697937,
      "num_tokens": 354489513.0,
      "step": 10188
    },
    {
      "epoch": 1.8921077065923861,
      "grad_norm": 1.6055949176326327,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8758809566497803,
      "num_tokens": 354517448.0,
      "step": 10189
    },
    {
      "epoch": 1.8922934076137419,
      "grad_norm": 1.5406572130147527,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8778735399246216,
      "num_tokens": 354547756.0,
      "step": 10190
    },
    {
      "epoch": 1.8924791086350976,
      "grad_norm": 1.442733130088194,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8835843801498413,
      "num_tokens": 354581766.0,
      "step": 10191
    },
    {
      "epoch": 1.8926648096564531,
      "grad_norm": 1.4913815040108747,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8721931576728821,
      "num_tokens": 354616702.0,
      "step": 10192
    },
    {
      "epoch": 1.8928505106778086,
      "grad_norm": 1.4919279074542855,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8753161430358887,
      "num_tokens": 354653140.0,
      "step": 10193
    },
    {
      "epoch": 1.8930362116991644,
      "grad_norm": 1.5794439935952176,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8710741996765137,
      "num_tokens": 354687153.0,
      "step": 10194
    },
    {
      "epoch": 1.89322191272052,
      "grad_norm": 1.514213705523017,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8832427263259888,
      "num_tokens": 354718176.0,
      "step": 10195
    },
    {
      "epoch": 1.8934076137418756,
      "grad_norm": 1.5335176849679217,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8809556365013123,
      "num_tokens": 354753072.0,
      "step": 10196
    },
    {
      "epoch": 1.8935933147632311,
      "grad_norm": 1.555260799495849,
      "learning_rate": 1e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8665554523468018,
      "num_tokens": 354788555.0,
      "step": 10197
    },
    {
      "epoch": 1.8937790157845868,
      "grad_norm": 1.5723740107771325,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8879247903823853,
      "num_tokens": 354817509.0,
      "step": 10198
    },
    {
      "epoch": 1.8939647168059426,
      "grad_norm": 1.5468523939644583,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8787637948989868,
      "num_tokens": 354850879.0,
      "step": 10199
    },
    {
      "epoch": 1.894150417827298,
      "grad_norm": 1.396633976381537,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8923565149307251,
      "num_tokens": 354886179.0,
      "step": 10200
    },
    {
      "epoch": 1.8943361188486536,
      "grad_norm": 1.5116755711564296,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8675083518028259,
      "num_tokens": 354925046.0,
      "step": 10201
    },
    {
      "epoch": 1.8945218198700093,
      "grad_norm": 1.658591753125543,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.876071572303772,
      "num_tokens": 354954307.0,
      "step": 10202
    },
    {
      "epoch": 1.8947075208913648,
      "grad_norm": 1.878506021756521,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8719667792320251,
      "num_tokens": 354980017.0,
      "step": 10203
    },
    {
      "epoch": 1.8948932219127204,
      "grad_norm": 1.4799441460280185,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8784583210945129,
      "num_tokens": 355018475.0,
      "step": 10204
    },
    {
      "epoch": 1.895078922934076,
      "grad_norm": 1.4651724891255702,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8776700496673584,
      "num_tokens": 355053248.0,
      "step": 10205
    },
    {
      "epoch": 1.8952646239554318,
      "grad_norm": 1.5015484344522367,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8906973600387573,
      "num_tokens": 355082956.0,
      "step": 10206
    },
    {
      "epoch": 1.8954503249767873,
      "grad_norm": 1.3794707358149698,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.88160240650177,
      "num_tokens": 355121658.0,
      "step": 10207
    },
    {
      "epoch": 1.8956360259981428,
      "grad_norm": 1.592544524484437,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8765822649002075,
      "num_tokens": 355153483.0,
      "step": 10208
    },
    {
      "epoch": 1.8958217270194986,
      "grad_norm": 1.3841138308197476,
      "learning_rate": 1e-06,
      "loss": 0.2928,
      "mean_token_accuracy": 0.897383451461792,
      "num_tokens": 355189794.0,
      "step": 10209
    },
    {
      "epoch": 1.8960074280408543,
      "grad_norm": 1.4876324175839428,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8784371614456177,
      "num_tokens": 355224876.0,
      "step": 10210
    },
    {
      "epoch": 1.8961931290622098,
      "grad_norm": 1.4481236441807859,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8806629776954651,
      "num_tokens": 355258781.0,
      "step": 10211
    },
    {
      "epoch": 1.8963788300835653,
      "grad_norm": 1.347911994485511,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8856780529022217,
      "num_tokens": 355297215.0,
      "step": 10212
    },
    {
      "epoch": 1.896564531104921,
      "grad_norm": 1.4979866865053058,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8769100904464722,
      "num_tokens": 355330907.0,
      "step": 10213
    },
    {
      "epoch": 1.8967502321262768,
      "grad_norm": 1.3769479878231585,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8899590969085693,
      "num_tokens": 355367185.0,
      "step": 10214
    },
    {
      "epoch": 1.8969359331476323,
      "grad_norm": 1.455834382836815,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8785935044288635,
      "num_tokens": 355401394.0,
      "step": 10215
    },
    {
      "epoch": 1.8971216341689878,
      "grad_norm": 1.447433427679154,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8827189803123474,
      "num_tokens": 355439937.0,
      "step": 10216
    },
    {
      "epoch": 1.8973073351903436,
      "grad_norm": 1.6589489248929068,
      "learning_rate": 1e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8657577633857727,
      "num_tokens": 355466775.0,
      "step": 10217
    },
    {
      "epoch": 1.8974930362116993,
      "grad_norm": 1.4649299465084313,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.886027991771698,
      "num_tokens": 355498695.0,
      "step": 10218
    },
    {
      "epoch": 1.8976787372330548,
      "grad_norm": 1.5283625349918086,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8890429735183716,
      "num_tokens": 355533169.0,
      "step": 10219
    },
    {
      "epoch": 1.8978644382544103,
      "grad_norm": 1.4080580547531985,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8858304619789124,
      "num_tokens": 355568358.0,
      "step": 10220
    },
    {
      "epoch": 1.898050139275766,
      "grad_norm": 1.4456985148590482,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8828876614570618,
      "num_tokens": 355606075.0,
      "step": 10221
    },
    {
      "epoch": 1.8982358402971218,
      "grad_norm": 1.587896235879802,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8826267719268799,
      "num_tokens": 355638013.0,
      "step": 10222
    },
    {
      "epoch": 1.8984215413184773,
      "grad_norm": 1.4697332811039767,
      "learning_rate": 1e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8679012060165405,
      "num_tokens": 355678246.0,
      "step": 10223
    },
    {
      "epoch": 1.8986072423398328,
      "grad_norm": 1.4572862136256541,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8779328465461731,
      "num_tokens": 355718603.0,
      "step": 10224
    },
    {
      "epoch": 1.8987929433611885,
      "grad_norm": 1.5257245011233136,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8718411922454834,
      "num_tokens": 355750291.0,
      "step": 10225
    },
    {
      "epoch": 1.8989786443825443,
      "grad_norm": 1.55293592761479,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8807774782180786,
      "num_tokens": 355782173.0,
      "step": 10226
    },
    {
      "epoch": 1.8991643454038996,
      "grad_norm": 1.501511127258292,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8875882029533386,
      "num_tokens": 355814434.0,
      "step": 10227
    },
    {
      "epoch": 1.8993500464252553,
      "grad_norm": 1.4825798092093245,
      "learning_rate": 1e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8642087578773499,
      "num_tokens": 355855677.0,
      "step": 10228
    },
    {
      "epoch": 1.899535747446611,
      "grad_norm": 1.3778326151567717,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8775913119316101,
      "num_tokens": 355897755.0,
      "step": 10229
    },
    {
      "epoch": 1.8997214484679665,
      "grad_norm": 1.4028848876306328,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8827482461929321,
      "num_tokens": 355938010.0,
      "step": 10230
    },
    {
      "epoch": 1.899907149489322,
      "grad_norm": 1.4568572798790467,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.892317533493042,
      "num_tokens": 355970095.0,
      "step": 10231
    },
    {
      "epoch": 1.9000928505106778,
      "grad_norm": 1.548367207909601,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8837054967880249,
      "num_tokens": 356001936.0,
      "step": 10232
    },
    {
      "epoch": 1.9002785515320335,
      "grad_norm": 1.6558990939101677,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8735709190368652,
      "num_tokens": 356030620.0,
      "step": 10233
    },
    {
      "epoch": 1.900464252553389,
      "grad_norm": 1.4798020448041642,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8788679838180542,
      "num_tokens": 356065492.0,
      "step": 10234
    },
    {
      "epoch": 1.9006499535747445,
      "grad_norm": 1.536740243906233,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8794188499450684,
      "num_tokens": 356099622.0,
      "step": 10235
    },
    {
      "epoch": 1.9008356545961003,
      "grad_norm": 1.4886552333580834,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8699094653129578,
      "num_tokens": 356135180.0,
      "step": 10236
    },
    {
      "epoch": 1.901021355617456,
      "grad_norm": 1.3651588139454287,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8925603032112122,
      "num_tokens": 356174687.0,
      "step": 10237
    },
    {
      "epoch": 1.9012070566388115,
      "grad_norm": 1.5995688309593337,
      "learning_rate": 1e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.8693383932113647,
      "num_tokens": 356205552.0,
      "step": 10238
    },
    {
      "epoch": 1.901392757660167,
      "grad_norm": 1.2993015661945733,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8796588182449341,
      "num_tokens": 356250845.0,
      "step": 10239
    },
    {
      "epoch": 1.9015784586815228,
      "grad_norm": 1.426932765868662,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8764651417732239,
      "num_tokens": 356289567.0,
      "step": 10240
    },
    {
      "epoch": 1.9017641597028785,
      "grad_norm": 1.5912423630878818,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8813439607620239,
      "num_tokens": 356322040.0,
      "step": 10241
    },
    {
      "epoch": 1.901949860724234,
      "grad_norm": 1.3637150651261336,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8846805095672607,
      "num_tokens": 356362854.0,
      "step": 10242
    },
    {
      "epoch": 1.9021355617455895,
      "grad_norm": 1.523614395126096,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8680608868598938,
      "num_tokens": 356398881.0,
      "step": 10243
    },
    {
      "epoch": 1.9023212627669452,
      "grad_norm": 1.3988726010212944,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8791841864585876,
      "num_tokens": 356436847.0,
      "step": 10244
    },
    {
      "epoch": 1.902506963788301,
      "grad_norm": 1.6061987345886415,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8725823163986206,
      "num_tokens": 356470389.0,
      "step": 10245
    },
    {
      "epoch": 1.9026926648096565,
      "grad_norm": 1.4770770415483903,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8876182436943054,
      "num_tokens": 356501761.0,
      "step": 10246
    },
    {
      "epoch": 1.902878365831012,
      "grad_norm": 1.519048969086648,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8860352635383606,
      "num_tokens": 356537291.0,
      "step": 10247
    },
    {
      "epoch": 1.9030640668523677,
      "grad_norm": 1.4045468474055816,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8897689580917358,
      "num_tokens": 356571417.0,
      "step": 10248
    },
    {
      "epoch": 1.9032497678737235,
      "grad_norm": 1.4888882579785168,
      "learning_rate": 1e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8584863543510437,
      "num_tokens": 356611825.0,
      "step": 10249
    },
    {
      "epoch": 1.903435468895079,
      "grad_norm": 1.4488446546343834,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.888351559638977,
      "num_tokens": 356647669.0,
      "step": 10250
    },
    {
      "epoch": 1.9036211699164345,
      "grad_norm": 1.4222459948197979,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8796564340591431,
      "num_tokens": 356680572.0,
      "step": 10251
    },
    {
      "epoch": 1.9038068709377902,
      "grad_norm": 1.4474776566887906,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8855741024017334,
      "num_tokens": 356714225.0,
      "step": 10252
    },
    {
      "epoch": 1.9039925719591457,
      "grad_norm": 1.5661811417383906,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8804299831390381,
      "num_tokens": 356746809.0,
      "step": 10253
    },
    {
      "epoch": 1.9041782729805012,
      "grad_norm": 1.473724340643766,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8882413506507874,
      "num_tokens": 356778705.0,
      "step": 10254
    },
    {
      "epoch": 1.904363974001857,
      "grad_norm": 1.4081469801747013,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8767679929733276,
      "num_tokens": 356817366.0,
      "step": 10255
    },
    {
      "epoch": 1.9045496750232127,
      "grad_norm": 1.3475954511438082,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.876097559928894,
      "num_tokens": 356858373.0,
      "step": 10256
    },
    {
      "epoch": 1.9047353760445682,
      "grad_norm": 1.416968379564681,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8798906803131104,
      "num_tokens": 356896629.0,
      "step": 10257
    },
    {
      "epoch": 1.9049210770659237,
      "grad_norm": 1.453627926097216,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8844373822212219,
      "num_tokens": 356929900.0,
      "step": 10258
    },
    {
      "epoch": 1.9051067780872795,
      "grad_norm": 1.643831721181482,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8855580687522888,
      "num_tokens": 356958434.0,
      "step": 10259
    },
    {
      "epoch": 1.9052924791086352,
      "grad_norm": 1.5460066813911721,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8672242164611816,
      "num_tokens": 356991704.0,
      "step": 10260
    },
    {
      "epoch": 1.9054781801299907,
      "grad_norm": 1.4437702121549107,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8873621225357056,
      "num_tokens": 357024756.0,
      "step": 10261
    },
    {
      "epoch": 1.9056638811513462,
      "grad_norm": 1.3856962383063418,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8790910243988037,
      "num_tokens": 357062622.0,
      "step": 10262
    },
    {
      "epoch": 1.905849582172702,
      "grad_norm": 1.3807533975067845,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8882982134819031,
      "num_tokens": 357099021.0,
      "step": 10263
    },
    {
      "epoch": 1.9060352831940577,
      "grad_norm": 1.4619767620901687,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8781968355178833,
      "num_tokens": 357136190.0,
      "step": 10264
    },
    {
      "epoch": 1.9062209842154132,
      "grad_norm": 1.688152143196208,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8728481531143188,
      "num_tokens": 357166090.0,
      "step": 10265
    },
    {
      "epoch": 1.9064066852367687,
      "grad_norm": 1.4218871143998375,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8731447458267212,
      "num_tokens": 357203364.0,
      "step": 10266
    },
    {
      "epoch": 1.9065923862581244,
      "grad_norm": 1.4259114714864567,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8787425756454468,
      "num_tokens": 357243951.0,
      "step": 10267
    },
    {
      "epoch": 1.9067780872794802,
      "grad_norm": 1.558112061889828,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8766750693321228,
      "num_tokens": 357275554.0,
      "step": 10268
    },
    {
      "epoch": 1.9069637883008357,
      "grad_norm": 1.5584281284086576,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8801752328872681,
      "num_tokens": 357301872.0,
      "step": 10269
    },
    {
      "epoch": 1.9071494893221912,
      "grad_norm": 1.4509487441427635,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8723266124725342,
      "num_tokens": 357340068.0,
      "step": 10270
    },
    {
      "epoch": 1.907335190343547,
      "grad_norm": 1.4399179844709855,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8760043978691101,
      "num_tokens": 357378874.0,
      "step": 10271
    },
    {
      "epoch": 1.9075208913649027,
      "grad_norm": 1.5049770656325652,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8626405000686646,
      "num_tokens": 357419152.0,
      "step": 10272
    },
    {
      "epoch": 1.9077065923862582,
      "grad_norm": 1.500855849045873,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.888392984867096,
      "num_tokens": 357450694.0,
      "step": 10273
    },
    {
      "epoch": 1.9078922934076137,
      "grad_norm": 1.5579629541402962,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8726105093955994,
      "num_tokens": 357482515.0,
      "step": 10274
    },
    {
      "epoch": 1.9080779944289694,
      "grad_norm": 1.3868766326958115,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.877070426940918,
      "num_tokens": 357523215.0,
      "step": 10275
    },
    {
      "epoch": 1.908263695450325,
      "grad_norm": 1.4681825181324217,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8754732608795166,
      "num_tokens": 357559992.0,
      "step": 10276
    },
    {
      "epoch": 1.9084493964716804,
      "grad_norm": 1.5157510198175692,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8787606358528137,
      "num_tokens": 357593484.0,
      "step": 10277
    },
    {
      "epoch": 1.9086350974930362,
      "grad_norm": 1.5698704443937204,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8710684776306152,
      "num_tokens": 357627256.0,
      "step": 10278
    },
    {
      "epoch": 1.908820798514392,
      "grad_norm": 1.728070643773267,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8786580562591553,
      "num_tokens": 357653595.0,
      "step": 10279
    },
    {
      "epoch": 1.9090064995357474,
      "grad_norm": 1.5325177449843272,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8778566122055054,
      "num_tokens": 357689262.0,
      "step": 10280
    },
    {
      "epoch": 1.909192200557103,
      "grad_norm": 1.4847210746877066,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8813266754150391,
      "num_tokens": 357725542.0,
      "step": 10281
    },
    {
      "epoch": 1.9093779015784587,
      "grad_norm": 1.5425103710376795,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8776899576187134,
      "num_tokens": 357760001.0,
      "step": 10282
    },
    {
      "epoch": 1.9095636025998144,
      "grad_norm": 1.4489061446598142,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8935049772262573,
      "num_tokens": 357791389.0,
      "step": 10283
    },
    {
      "epoch": 1.90974930362117,
      "grad_norm": 1.4014706635562055,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8884417414665222,
      "num_tokens": 357826882.0,
      "step": 10284
    },
    {
      "epoch": 1.9099350046425254,
      "grad_norm": 1.4607811798562143,
      "learning_rate": 1e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.871464729309082,
      "num_tokens": 357862319.0,
      "step": 10285
    },
    {
      "epoch": 1.9101207056638811,
      "grad_norm": 1.4093736578234923,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8865987658500671,
      "num_tokens": 357895072.0,
      "step": 10286
    },
    {
      "epoch": 1.9103064066852369,
      "grad_norm": 1.4646227930109839,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8821142911911011,
      "num_tokens": 357929531.0,
      "step": 10287
    },
    {
      "epoch": 1.9104921077065924,
      "grad_norm": 1.6722931501767386,
      "learning_rate": 1e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8531670570373535,
      "num_tokens": 357959527.0,
      "step": 10288
    },
    {
      "epoch": 1.910677808727948,
      "grad_norm": 1.5215895810219788,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.877212405204773,
      "num_tokens": 357993069.0,
      "step": 10289
    },
    {
      "epoch": 1.9108635097493036,
      "grad_norm": 1.4405786886744927,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8817578554153442,
      "num_tokens": 358028894.0,
      "step": 10290
    },
    {
      "epoch": 1.9110492107706594,
      "grad_norm": 1.467204410928112,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8803204894065857,
      "num_tokens": 358064016.0,
      "step": 10291
    },
    {
      "epoch": 1.9112349117920149,
      "grad_norm": 1.505551924890468,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8884460926055908,
      "num_tokens": 358097191.0,
      "step": 10292
    },
    {
      "epoch": 1.9114206128133704,
      "grad_norm": 1.4686800292331859,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8808392286300659,
      "num_tokens": 358130853.0,
      "step": 10293
    },
    {
      "epoch": 1.9116063138347261,
      "grad_norm": 1.4424772645580846,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8808186054229736,
      "num_tokens": 358166028.0,
      "step": 10294
    },
    {
      "epoch": 1.9117920148560819,
      "grad_norm": 1.387568794262973,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8843680620193481,
      "num_tokens": 358205138.0,
      "step": 10295
    },
    {
      "epoch": 1.9119777158774374,
      "grad_norm": 1.674249681205891,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8878672122955322,
      "num_tokens": 358237638.0,
      "step": 10296
    },
    {
      "epoch": 1.9121634168987929,
      "grad_norm": 1.4706014026430854,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.88255774974823,
      "num_tokens": 358272079.0,
      "step": 10297
    },
    {
      "epoch": 1.9123491179201486,
      "grad_norm": 1.4836833200328432,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8869112730026245,
      "num_tokens": 358306052.0,
      "step": 10298
    },
    {
      "epoch": 1.9125348189415043,
      "grad_norm": 1.4950700333312716,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.876915693283081,
      "num_tokens": 358342382.0,
      "step": 10299
    },
    {
      "epoch": 1.9127205199628596,
      "grad_norm": 1.5740140241352327,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8780736923217773,
      "num_tokens": 358373847.0,
      "step": 10300
    },
    {
      "epoch": 1.9129062209842154,
      "grad_norm": 1.7088297069969032,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8892970085144043,
      "num_tokens": 358400499.0,
      "step": 10301
    },
    {
      "epoch": 1.913091922005571,
      "grad_norm": 1.3789252547569568,
      "learning_rate": 1e-06,
      "loss": 0.3876,
      "mean_token_accuracy": 0.8670658469200134,
      "num_tokens": 358442363.0,
      "step": 10302
    },
    {
      "epoch": 1.9132776230269266,
      "grad_norm": 1.5892831376128984,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.864242672920227,
      "num_tokens": 358474191.0,
      "step": 10303
    },
    {
      "epoch": 1.9134633240482821,
      "grad_norm": 1.510615706641424,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8822357654571533,
      "num_tokens": 358508678.0,
      "step": 10304
    },
    {
      "epoch": 1.9136490250696379,
      "grad_norm": 1.4338826133764264,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8824105262756348,
      "num_tokens": 358544729.0,
      "step": 10305
    },
    {
      "epoch": 1.9138347260909936,
      "grad_norm": 1.4612457115075836,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.876078724861145,
      "num_tokens": 358579257.0,
      "step": 10306
    },
    {
      "epoch": 1.914020427112349,
      "grad_norm": 1.4136197878259287,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8822614550590515,
      "num_tokens": 358618995.0,
      "step": 10307
    },
    {
      "epoch": 1.9142061281337046,
      "grad_norm": 1.5427491708487397,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8805832862854004,
      "num_tokens": 358651739.0,
      "step": 10308
    },
    {
      "epoch": 1.9143918291550603,
      "grad_norm": 1.5652242315271963,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8704460263252258,
      "num_tokens": 358686890.0,
      "step": 10309
    },
    {
      "epoch": 1.914577530176416,
      "grad_norm": 1.6488097252781115,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8782424330711365,
      "num_tokens": 358717167.0,
      "step": 10310
    },
    {
      "epoch": 1.9147632311977716,
      "grad_norm": 1.5045496528823246,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8812255859375,
      "num_tokens": 358753729.0,
      "step": 10311
    },
    {
      "epoch": 1.914948932219127,
      "grad_norm": 1.4914821464704127,
      "learning_rate": 1e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8642085790634155,
      "num_tokens": 358791981.0,
      "step": 10312
    },
    {
      "epoch": 1.9151346332404828,
      "grad_norm": 1.4389329446379242,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8836584687232971,
      "num_tokens": 358829638.0,
      "step": 10313
    },
    {
      "epoch": 1.9153203342618386,
      "grad_norm": 1.4717680617334783,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8826998472213745,
      "num_tokens": 358863057.0,
      "step": 10314
    },
    {
      "epoch": 1.915506035283194,
      "grad_norm": 1.5351609731291285,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8887141942977905,
      "num_tokens": 358896849.0,
      "step": 10315
    },
    {
      "epoch": 1.9156917363045496,
      "grad_norm": 1.477894920384933,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8694335222244263,
      "num_tokens": 358935162.0,
      "step": 10316
    },
    {
      "epoch": 1.9158774373259053,
      "grad_norm": 1.4717892221314315,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8899048566818237,
      "num_tokens": 358968953.0,
      "step": 10317
    },
    {
      "epoch": 1.916063138347261,
      "grad_norm": 1.4760807767651707,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.875910222530365,
      "num_tokens": 359006091.0,
      "step": 10318
    },
    {
      "epoch": 1.9162488393686166,
      "grad_norm": 1.4405982483218565,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.878167986869812,
      "num_tokens": 359042894.0,
      "step": 10319
    },
    {
      "epoch": 1.916434540389972,
      "grad_norm": 1.3133603160257787,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8802851438522339,
      "num_tokens": 359086733.0,
      "step": 10320
    },
    {
      "epoch": 1.9166202414113278,
      "grad_norm": 1.4830600711151847,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.887953519821167,
      "num_tokens": 359117534.0,
      "step": 10321
    },
    {
      "epoch": 1.9168059424326835,
      "grad_norm": 1.5725543862579354,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.872383177280426,
      "num_tokens": 359149098.0,
      "step": 10322
    },
    {
      "epoch": 1.916991643454039,
      "grad_norm": 1.5031181399876699,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8820319771766663,
      "num_tokens": 359182329.0,
      "step": 10323
    },
    {
      "epoch": 1.9171773444753946,
      "grad_norm": 1.494650983611269,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.879677951335907,
      "num_tokens": 359218403.0,
      "step": 10324
    },
    {
      "epoch": 1.9173630454967503,
      "grad_norm": 1.4767463031293437,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8858959674835205,
      "num_tokens": 359252097.0,
      "step": 10325
    },
    {
      "epoch": 1.9175487465181058,
      "grad_norm": 1.3344748300093605,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8823651075363159,
      "num_tokens": 359292839.0,
      "step": 10326
    },
    {
      "epoch": 1.9177344475394613,
      "grad_norm": 1.4141232343903842,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8712693452835083,
      "num_tokens": 359329937.0,
      "step": 10327
    },
    {
      "epoch": 1.917920148560817,
      "grad_norm": 1.603908305004563,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.872800350189209,
      "num_tokens": 359358816.0,
      "step": 10328
    },
    {
      "epoch": 1.9181058495821728,
      "grad_norm": 1.4443786195067332,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8752540349960327,
      "num_tokens": 359398729.0,
      "step": 10329
    },
    {
      "epoch": 1.9182915506035283,
      "grad_norm": 1.3951612616568756,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8841450214385986,
      "num_tokens": 359437280.0,
      "step": 10330
    },
    {
      "epoch": 1.9184772516248838,
      "grad_norm": 1.4253053930207042,
      "learning_rate": 1e-06,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8724874258041382,
      "num_tokens": 359477416.0,
      "step": 10331
    },
    {
      "epoch": 1.9186629526462395,
      "grad_norm": 1.472254448498576,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8773660659790039,
      "num_tokens": 359511018.0,
      "step": 10332
    },
    {
      "epoch": 1.9188486536675953,
      "grad_norm": 1.4160672536408223,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8796850442886353,
      "num_tokens": 359547316.0,
      "step": 10333
    },
    {
      "epoch": 1.9190343546889508,
      "grad_norm": 1.5459850188597413,
      "learning_rate": 1e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8593895435333252,
      "num_tokens": 359583977.0,
      "step": 10334
    },
    {
      "epoch": 1.9192200557103063,
      "grad_norm": 1.6455724467546653,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8934564590454102,
      "num_tokens": 359607832.0,
      "step": 10335
    },
    {
      "epoch": 1.919405756731662,
      "grad_norm": 1.641974843972357,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8745037317276001,
      "num_tokens": 359636677.0,
      "step": 10336
    },
    {
      "epoch": 1.9195914577530178,
      "grad_norm": 1.5516244686950726,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8731346130371094,
      "num_tokens": 359679272.0,
      "step": 10337
    },
    {
      "epoch": 1.9197771587743733,
      "grad_norm": 1.3483628263805612,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8864275217056274,
      "num_tokens": 359718652.0,
      "step": 10338
    },
    {
      "epoch": 1.9199628597957288,
      "grad_norm": 1.2910436196361332,
      "learning_rate": 1e-06,
      "loss": 0.2998,
      "mean_token_accuracy": 0.8967956304550171,
      "num_tokens": 359759453.0,
      "step": 10339
    },
    {
      "epoch": 1.9201485608170845,
      "grad_norm": 1.546568762392476,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8696665167808533,
      "num_tokens": 359793119.0,
      "step": 10340
    },
    {
      "epoch": 1.9203342618384402,
      "grad_norm": 1.3740342297666814,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8855228424072266,
      "num_tokens": 359830348.0,
      "step": 10341
    },
    {
      "epoch": 1.9205199628597958,
      "grad_norm": 1.488682901068431,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8808717131614685,
      "num_tokens": 359865700.0,
      "step": 10342
    },
    {
      "epoch": 1.9207056638811513,
      "grad_norm": 1.375960098764619,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8867828249931335,
      "num_tokens": 359905069.0,
      "step": 10343
    },
    {
      "epoch": 1.920891364902507,
      "grad_norm": 1.4606758306116163,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8822302222251892,
      "num_tokens": 359941252.0,
      "step": 10344
    },
    {
      "epoch": 1.9210770659238627,
      "grad_norm": 1.5321566556011557,
      "learning_rate": 1e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8545436263084412,
      "num_tokens": 359977692.0,
      "step": 10345
    },
    {
      "epoch": 1.9212627669452182,
      "grad_norm": 1.560493291013245,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8813180923461914,
      "num_tokens": 360009984.0,
      "step": 10346
    },
    {
      "epoch": 1.9214484679665738,
      "grad_norm": 1.4906312467718406,
      "learning_rate": 1e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8635600805282593,
      "num_tokens": 360046287.0,
      "step": 10347
    },
    {
      "epoch": 1.9216341689879295,
      "grad_norm": 1.556636537601496,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.8943769931793213,
      "num_tokens": 360072993.0,
      "step": 10348
    },
    {
      "epoch": 1.921819870009285,
      "grad_norm": 1.3790858698643542,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8827042579650879,
      "num_tokens": 360109682.0,
      "step": 10349
    },
    {
      "epoch": 1.9220055710306405,
      "grad_norm": 1.5449477930425155,
      "learning_rate": 1e-06,
      "loss": 0.3584,
      "mean_token_accuracy": 0.8763250112533569,
      "num_tokens": 360143900.0,
      "step": 10350
    },
    {
      "epoch": 1.9221912720519962,
      "grad_norm": 1.4249211394063932,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8788597583770752,
      "num_tokens": 360183194.0,
      "step": 10351
    },
    {
      "epoch": 1.922376973073352,
      "grad_norm": 1.5464385737363429,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8810758590698242,
      "num_tokens": 360216820.0,
      "step": 10352
    },
    {
      "epoch": 1.9225626740947075,
      "grad_norm": 1.4343953473412152,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8908612728118896,
      "num_tokens": 360252986.0,
      "step": 10353
    },
    {
      "epoch": 1.922748375116063,
      "grad_norm": 1.774193086612566,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8772106170654297,
      "num_tokens": 360287752.0,
      "step": 10354
    },
    {
      "epoch": 1.9229340761374187,
      "grad_norm": 1.5992223844953968,
      "learning_rate": 1e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8602488040924072,
      "num_tokens": 360321381.0,
      "step": 10355
    },
    {
      "epoch": 1.9231197771587745,
      "grad_norm": 1.491816655728042,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8829012513160706,
      "num_tokens": 360355140.0,
      "step": 10356
    },
    {
      "epoch": 1.92330547818013,
      "grad_norm": 1.4321266250545301,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8772808909416199,
      "num_tokens": 360393405.0,
      "step": 10357
    },
    {
      "epoch": 1.9234911792014855,
      "grad_norm": 1.6112951839346914,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.8978879451751709,
      "num_tokens": 360419526.0,
      "step": 10358
    },
    {
      "epoch": 1.9236768802228412,
      "grad_norm": 1.5113620298003627,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8760255575180054,
      "num_tokens": 360453964.0,
      "step": 10359
    },
    {
      "epoch": 1.923862581244197,
      "grad_norm": 1.4382217180308006,
      "learning_rate": 1e-06,
      "loss": 0.3793,
      "mean_token_accuracy": 0.8663702011108398,
      "num_tokens": 360491009.0,
      "step": 10360
    },
    {
      "epoch": 1.9240482822655525,
      "grad_norm": 1.3784672727606169,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8741455674171448,
      "num_tokens": 360536162.0,
      "step": 10361
    },
    {
      "epoch": 1.924233983286908,
      "grad_norm": 1.6519731288013126,
      "learning_rate": 1e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8603041172027588,
      "num_tokens": 360569816.0,
      "step": 10362
    },
    {
      "epoch": 1.9244196843082637,
      "grad_norm": 1.3498783572520496,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8917096257209778,
      "num_tokens": 360607627.0,
      "step": 10363
    },
    {
      "epoch": 1.9246053853296194,
      "grad_norm": 1.3179856995448715,
      "learning_rate": 1e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8635693788528442,
      "num_tokens": 360653435.0,
      "step": 10364
    },
    {
      "epoch": 1.924791086350975,
      "grad_norm": 1.5037479330130974,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8679834604263306,
      "num_tokens": 360687346.0,
      "step": 10365
    },
    {
      "epoch": 1.9249767873723305,
      "grad_norm": 1.5571217410607037,
      "learning_rate": 1e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8707317113876343,
      "num_tokens": 360723115.0,
      "step": 10366
    },
    {
      "epoch": 1.9251624883936862,
      "grad_norm": 1.4408217860856922,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8834555149078369,
      "num_tokens": 360761305.0,
      "step": 10367
    },
    {
      "epoch": 1.925348189415042,
      "grad_norm": 1.5702570776053624,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8793011903762817,
      "num_tokens": 360791844.0,
      "step": 10368
    },
    {
      "epoch": 1.9255338904363974,
      "grad_norm": 1.4617550940168667,
      "learning_rate": 1e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8652788400650024,
      "num_tokens": 360829447.0,
      "step": 10369
    },
    {
      "epoch": 1.925719591457753,
      "grad_norm": 1.5008365782642414,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8765289187431335,
      "num_tokens": 360871840.0,
      "step": 10370
    },
    {
      "epoch": 1.9259052924791087,
      "grad_norm": 1.5876364468770356,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8807544708251953,
      "num_tokens": 360902477.0,
      "step": 10371
    },
    {
      "epoch": 1.9260909935004642,
      "grad_norm": 1.4775347486839105,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8702728748321533,
      "num_tokens": 360941525.0,
      "step": 10372
    },
    {
      "epoch": 1.9262766945218197,
      "grad_norm": 1.4730181440392718,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8821824789047241,
      "num_tokens": 360977409.0,
      "step": 10373
    },
    {
      "epoch": 1.9264623955431754,
      "grad_norm": 1.53399519017365,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8761600255966187,
      "num_tokens": 361010022.0,
      "step": 10374
    },
    {
      "epoch": 1.9266480965645312,
      "grad_norm": 1.4838023213130747,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8816916942596436,
      "num_tokens": 361044670.0,
      "step": 10375
    },
    {
      "epoch": 1.9268337975858867,
      "grad_norm": 1.4122724051881084,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8743029832839966,
      "num_tokens": 361081420.0,
      "step": 10376
    },
    {
      "epoch": 1.9270194986072422,
      "grad_norm": 1.5951538149521682,
      "learning_rate": 1e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.864538311958313,
      "num_tokens": 361114102.0,
      "step": 10377
    },
    {
      "epoch": 1.927205199628598,
      "grad_norm": 1.580627120689398,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8803658485412598,
      "num_tokens": 361141381.0,
      "step": 10378
    },
    {
      "epoch": 1.9273909006499537,
      "grad_norm": 1.4665665072020178,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8894458413124084,
      "num_tokens": 361175692.0,
      "step": 10379
    },
    {
      "epoch": 1.9275766016713092,
      "grad_norm": 1.4403518320714646,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8769917488098145,
      "num_tokens": 361214315.0,
      "step": 10380
    },
    {
      "epoch": 1.9277623026926647,
      "grad_norm": 1.4971511775838138,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8743243217468262,
      "num_tokens": 361248905.0,
      "step": 10381
    },
    {
      "epoch": 1.9279480037140204,
      "grad_norm": 1.714912593208052,
      "learning_rate": 1e-06,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8675902485847473,
      "num_tokens": 361278086.0,
      "step": 10382
    },
    {
      "epoch": 1.9281337047353762,
      "grad_norm": 1.4503816313180804,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8629903197288513,
      "num_tokens": 361319028.0,
      "step": 10383
    },
    {
      "epoch": 1.9283194057567317,
      "grad_norm": 1.4063144166770456,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8789342045783997,
      "num_tokens": 361355895.0,
      "step": 10384
    },
    {
      "epoch": 1.9285051067780872,
      "grad_norm": 1.488616649475024,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.875464677810669,
      "num_tokens": 361389373.0,
      "step": 10385
    },
    {
      "epoch": 1.928690807799443,
      "grad_norm": 1.4610666404577248,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8794941902160645,
      "num_tokens": 361424562.0,
      "step": 10386
    },
    {
      "epoch": 1.9288765088207986,
      "grad_norm": 1.5337780660741738,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8723121881484985,
      "num_tokens": 361456617.0,
      "step": 10387
    },
    {
      "epoch": 1.9290622098421542,
      "grad_norm": 1.4723813366961882,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8913036584854126,
      "num_tokens": 361487743.0,
      "step": 10388
    },
    {
      "epoch": 1.9292479108635097,
      "grad_norm": 1.553582921251157,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8788735866546631,
      "num_tokens": 361522171.0,
      "step": 10389
    },
    {
      "epoch": 1.9294336118848654,
      "grad_norm": 1.623736110058937,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8669097423553467,
      "num_tokens": 361554928.0,
      "step": 10390
    },
    {
      "epoch": 1.9296193129062211,
      "grad_norm": 1.5470317226041896,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8869365453720093,
      "num_tokens": 361584271.0,
      "step": 10391
    },
    {
      "epoch": 1.9298050139275766,
      "grad_norm": 1.337530976429004,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8780462741851807,
      "num_tokens": 361623305.0,
      "step": 10392
    },
    {
      "epoch": 1.9299907149489322,
      "grad_norm": 1.4179121474590781,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8720473051071167,
      "num_tokens": 361661061.0,
      "step": 10393
    },
    {
      "epoch": 1.9301764159702879,
      "grad_norm": 1.5287885539364543,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8679897785186768,
      "num_tokens": 361698456.0,
      "step": 10394
    },
    {
      "epoch": 1.9303621169916436,
      "grad_norm": 1.388281768816342,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8896034359931946,
      "num_tokens": 361738025.0,
      "step": 10395
    },
    {
      "epoch": 1.930547818012999,
      "grad_norm": 1.4390231336574508,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.887633740901947,
      "num_tokens": 361770979.0,
      "step": 10396
    },
    {
      "epoch": 1.9307335190343546,
      "grad_norm": 1.55037753142908,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8841865658760071,
      "num_tokens": 361802425.0,
      "step": 10397
    },
    {
      "epoch": 1.9309192200557104,
      "grad_norm": 1.4371810008692651,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8838813900947571,
      "num_tokens": 361839841.0,
      "step": 10398
    },
    {
      "epoch": 1.9311049210770659,
      "grad_norm": 1.3785991504626594,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8847248554229736,
      "num_tokens": 361877726.0,
      "step": 10399
    },
    {
      "epoch": 1.9312906220984214,
      "grad_norm": 1.3953869709269842,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8758656978607178,
      "num_tokens": 361918913.0,
      "step": 10400
    },
    {
      "epoch": 1.9314763231197771,
      "grad_norm": 1.6527493191160934,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8881822824478149,
      "num_tokens": 361945860.0,
      "step": 10401
    },
    {
      "epoch": 1.9316620241411329,
      "grad_norm": 1.4175737210963202,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8791921138763428,
      "num_tokens": 361982245.0,
      "step": 10402
    },
    {
      "epoch": 1.9318477251624884,
      "grad_norm": 1.353459659939687,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8887337446212769,
      "num_tokens": 362022218.0,
      "step": 10403
    },
    {
      "epoch": 1.9320334261838439,
      "grad_norm": 1.4857326869987675,
      "learning_rate": 1e-06,
      "loss": 0.3679,
      "mean_token_accuracy": 0.8742893934249878,
      "num_tokens": 362057298.0,
      "step": 10404
    },
    {
      "epoch": 1.9322191272051996,
      "grad_norm": 1.5529560745227697,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8817305564880371,
      "num_tokens": 362092602.0,
      "step": 10405
    },
    {
      "epoch": 1.9324048282265553,
      "grad_norm": 1.4174820484140314,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8768115639686584,
      "num_tokens": 362128617.0,
      "step": 10406
    },
    {
      "epoch": 1.9325905292479109,
      "grad_norm": 1.6790510714377325,
      "learning_rate": 1e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8643925189971924,
      "num_tokens": 362158638.0,
      "step": 10407
    },
    {
      "epoch": 1.9327762302692664,
      "grad_norm": 1.431761782505609,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.8765710592269897,
      "num_tokens": 362197375.0,
      "step": 10408
    },
    {
      "epoch": 1.932961931290622,
      "grad_norm": 1.5562359107868295,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8813722133636475,
      "num_tokens": 362227887.0,
      "step": 10409
    },
    {
      "epoch": 1.9331476323119778,
      "grad_norm": 1.4226004093513704,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8735738396644592,
      "num_tokens": 362265417.0,
      "step": 10410
    },
    {
      "epoch": 1.9333333333333333,
      "grad_norm": 1.4685702700185312,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8876559734344482,
      "num_tokens": 362303078.0,
      "step": 10411
    },
    {
      "epoch": 1.9335190343546889,
      "grad_norm": 1.45176954856456,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8870879411697388,
      "num_tokens": 362339640.0,
      "step": 10412
    },
    {
      "epoch": 1.9337047353760446,
      "grad_norm": 1.4224634189758947,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8799194693565369,
      "num_tokens": 362374067.0,
      "step": 10413
    },
    {
      "epoch": 1.9338904363974003,
      "grad_norm": 1.5838437964460845,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8864177465438843,
      "num_tokens": 362403052.0,
      "step": 10414
    },
    {
      "epoch": 1.9340761374187558,
      "grad_norm": 1.602106015763441,
      "learning_rate": 1e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8616936802864075,
      "num_tokens": 362440218.0,
      "step": 10415
    },
    {
      "epoch": 1.9342618384401113,
      "grad_norm": 1.3816677412410767,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8777779340744019,
      "num_tokens": 362481906.0,
      "step": 10416
    },
    {
      "epoch": 1.934447539461467,
      "grad_norm": 1.5987414745130484,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8809002637863159,
      "num_tokens": 362509755.0,
      "step": 10417
    },
    {
      "epoch": 1.9346332404828228,
      "grad_norm": 1.3915231531934746,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8846745491027832,
      "num_tokens": 362547509.0,
      "step": 10418
    },
    {
      "epoch": 1.9348189415041783,
      "grad_norm": 1.6096062469478998,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.872130274772644,
      "num_tokens": 362577916.0,
      "step": 10419
    },
    {
      "epoch": 1.9350046425255338,
      "grad_norm": 1.529475246302807,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.883724570274353,
      "num_tokens": 362609002.0,
      "step": 10420
    },
    {
      "epoch": 1.9351903435468896,
      "grad_norm": 1.4661189212632437,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8752280473709106,
      "num_tokens": 362647580.0,
      "step": 10421
    },
    {
      "epoch": 1.935376044568245,
      "grad_norm": 1.5835024032098697,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8821823596954346,
      "num_tokens": 362679537.0,
      "step": 10422
    },
    {
      "epoch": 1.9355617455896006,
      "grad_norm": 1.464997187600679,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8792985081672668,
      "num_tokens": 362716608.0,
      "step": 10423
    },
    {
      "epoch": 1.9357474466109563,
      "grad_norm": 1.538553870890454,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8898864388465881,
      "num_tokens": 362746066.0,
      "step": 10424
    },
    {
      "epoch": 1.935933147632312,
      "grad_norm": 1.454858660665108,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8782507181167603,
      "num_tokens": 362780624.0,
      "step": 10425
    },
    {
      "epoch": 1.9361188486536676,
      "grad_norm": 1.7839574277634225,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8708533644676208,
      "num_tokens": 362807042.0,
      "step": 10426
    },
    {
      "epoch": 1.936304549675023,
      "grad_norm": 1.5293706062021208,
      "learning_rate": 1e-06,
      "loss": 0.3728,
      "mean_token_accuracy": 0.8729177713394165,
      "num_tokens": 362841321.0,
      "step": 10427
    },
    {
      "epoch": 1.9364902506963788,
      "grad_norm": 1.668733440194377,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8741191029548645,
      "num_tokens": 362872315.0,
      "step": 10428
    },
    {
      "epoch": 1.9366759517177345,
      "grad_norm": 1.5441299543175728,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8903065323829651,
      "num_tokens": 362902780.0,
      "step": 10429
    },
    {
      "epoch": 1.93686165273909,
      "grad_norm": 1.6505715752522487,
      "learning_rate": 1e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.869409441947937,
      "num_tokens": 362933076.0,
      "step": 10430
    },
    {
      "epoch": 1.9370473537604456,
      "grad_norm": 1.6070421757508577,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8727986812591553,
      "num_tokens": 362968023.0,
      "step": 10431
    },
    {
      "epoch": 1.9372330547818013,
      "grad_norm": 1.5065228767999328,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8800600171089172,
      "num_tokens": 363002951.0,
      "step": 10432
    },
    {
      "epoch": 1.937418755803157,
      "grad_norm": 1.450290148645221,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8793729543685913,
      "num_tokens": 363036144.0,
      "step": 10433
    },
    {
      "epoch": 1.9376044568245125,
      "grad_norm": 1.527828180145762,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8755745887756348,
      "num_tokens": 363069229.0,
      "step": 10434
    },
    {
      "epoch": 1.937790157845868,
      "grad_norm": 1.4490830774307817,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8715571165084839,
      "num_tokens": 363106308.0,
      "step": 10435
    },
    {
      "epoch": 1.9379758588672238,
      "grad_norm": 1.642179891877865,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8872575759887695,
      "num_tokens": 363134828.0,
      "step": 10436
    },
    {
      "epoch": 1.9381615598885795,
      "grad_norm": 1.55562797771455,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8717089891433716,
      "num_tokens": 363165484.0,
      "step": 10437
    },
    {
      "epoch": 1.938347260909935,
      "grad_norm": 1.4615548789082955,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8869552612304688,
      "num_tokens": 363198468.0,
      "step": 10438
    },
    {
      "epoch": 1.9385329619312905,
      "grad_norm": 1.4347357271611376,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8802496790885925,
      "num_tokens": 363237587.0,
      "step": 10439
    },
    {
      "epoch": 1.9387186629526463,
      "grad_norm": 1.6092480341579947,
      "learning_rate": 1e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8728787302970886,
      "num_tokens": 363268683.0,
      "step": 10440
    },
    {
      "epoch": 1.938904363974002,
      "grad_norm": 1.4450905098488593,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8793635964393616,
      "num_tokens": 363305089.0,
      "step": 10441
    },
    {
      "epoch": 1.9390900649953575,
      "grad_norm": 1.5355021018949855,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8906575441360474,
      "num_tokens": 363336617.0,
      "step": 10442
    },
    {
      "epoch": 1.939275766016713,
      "grad_norm": 1.6517023168966838,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8790276646614075,
      "num_tokens": 363366094.0,
      "step": 10443
    },
    {
      "epoch": 1.9394614670380688,
      "grad_norm": 1.6818843192118862,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8760073184967041,
      "num_tokens": 363392222.0,
      "step": 10444
    },
    {
      "epoch": 1.9396471680594243,
      "grad_norm": 1.4107052648991398,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8788960576057434,
      "num_tokens": 363427574.0,
      "step": 10445
    },
    {
      "epoch": 1.9398328690807798,
      "grad_norm": 1.3620795066518385,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8902485966682434,
      "num_tokens": 363463843.0,
      "step": 10446
    },
    {
      "epoch": 1.9400185701021355,
      "grad_norm": 1.7496930448545227,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8911508917808533,
      "num_tokens": 363491604.0,
      "step": 10447
    },
    {
      "epoch": 1.9402042711234913,
      "grad_norm": 1.458560643420573,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8831313252449036,
      "num_tokens": 363525293.0,
      "step": 10448
    },
    {
      "epoch": 1.9403899721448468,
      "grad_norm": 1.4917914350029153,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.882490873336792,
      "num_tokens": 363559575.0,
      "step": 10449
    },
    {
      "epoch": 1.9405756731662023,
      "grad_norm": 1.4989799117246743,
      "learning_rate": 1e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8655261993408203,
      "num_tokens": 363600392.0,
      "step": 10450
    },
    {
      "epoch": 1.940761374187558,
      "grad_norm": 1.4254991004025797,
      "learning_rate": 1e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8625954985618591,
      "num_tokens": 363641121.0,
      "step": 10451
    },
    {
      "epoch": 1.9409470752089137,
      "grad_norm": 1.6155693556926423,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8756058812141418,
      "num_tokens": 363671591.0,
      "step": 10452
    },
    {
      "epoch": 1.9411327762302693,
      "grad_norm": 1.4009622062312699,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8806709051132202,
      "num_tokens": 363708654.0,
      "step": 10453
    },
    {
      "epoch": 1.9413184772516248,
      "grad_norm": 1.648488696261295,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8697695732116699,
      "num_tokens": 363740269.0,
      "step": 10454
    },
    {
      "epoch": 1.9415041782729805,
      "grad_norm": 1.4572403525040336,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8807213306427002,
      "num_tokens": 363775729.0,
      "step": 10455
    },
    {
      "epoch": 1.9416898792943362,
      "grad_norm": 1.3332742009351262,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8810909986495972,
      "num_tokens": 363819056.0,
      "step": 10456
    },
    {
      "epoch": 1.9418755803156917,
      "grad_norm": 1.388511831577297,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8768777847290039,
      "num_tokens": 363859853.0,
      "step": 10457
    },
    {
      "epoch": 1.9420612813370473,
      "grad_norm": 1.4873279868246063,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8813225626945496,
      "num_tokens": 363896030.0,
      "step": 10458
    },
    {
      "epoch": 1.942246982358403,
      "grad_norm": 1.6065580678880413,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8708311319351196,
      "num_tokens": 363927067.0,
      "step": 10459
    },
    {
      "epoch": 1.9424326833797587,
      "grad_norm": 1.410210681722715,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8856030106544495,
      "num_tokens": 363963478.0,
      "step": 10460
    },
    {
      "epoch": 1.9426183844011142,
      "grad_norm": 1.4802428203564064,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8832472562789917,
      "num_tokens": 363996936.0,
      "step": 10461
    },
    {
      "epoch": 1.9428040854224697,
      "grad_norm": 1.5061969052504924,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8663786053657532,
      "num_tokens": 364029582.0,
      "step": 10462
    },
    {
      "epoch": 1.9429897864438255,
      "grad_norm": 1.668876046171615,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8825373649597168,
      "num_tokens": 364056177.0,
      "step": 10463
    },
    {
      "epoch": 1.9431754874651812,
      "grad_norm": 1.4653657073514965,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8864229321479797,
      "num_tokens": 364087901.0,
      "step": 10464
    },
    {
      "epoch": 1.9433611884865367,
      "grad_norm": 1.4613230953200114,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8904246091842651,
      "num_tokens": 364122257.0,
      "step": 10465
    },
    {
      "epoch": 1.9435468895078922,
      "grad_norm": 1.7270915585741846,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8831901550292969,
      "num_tokens": 364148038.0,
      "step": 10466
    },
    {
      "epoch": 1.943732590529248,
      "grad_norm": 1.4508293720166943,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8881130218505859,
      "num_tokens": 364182563.0,
      "step": 10467
    },
    {
      "epoch": 1.9439182915506037,
      "grad_norm": 1.482134317307038,
      "learning_rate": 1e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8651913404464722,
      "num_tokens": 364225490.0,
      "step": 10468
    },
    {
      "epoch": 1.944103992571959,
      "grad_norm": 1.3937326482197983,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8938312530517578,
      "num_tokens": 364259327.0,
      "step": 10469
    },
    {
      "epoch": 1.9442896935933147,
      "grad_norm": 1.4487640053886084,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8872939944267273,
      "num_tokens": 364298186.0,
      "step": 10470
    },
    {
      "epoch": 1.9444753946146704,
      "grad_norm": 1.473708425667063,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8800202012062073,
      "num_tokens": 364331839.0,
      "step": 10471
    },
    {
      "epoch": 1.944661095636026,
      "grad_norm": 1.4118399846627991,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8839384317398071,
      "num_tokens": 364369508.0,
      "step": 10472
    },
    {
      "epoch": 1.9448467966573815,
      "grad_norm": 1.5299866172211454,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8783475756645203,
      "num_tokens": 364400571.0,
      "step": 10473
    },
    {
      "epoch": 1.9450324976787372,
      "grad_norm": 1.5799185266057871,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8776426315307617,
      "num_tokens": 364432029.0,
      "step": 10474
    },
    {
      "epoch": 1.945218198700093,
      "grad_norm": 1.451693507802656,
      "learning_rate": 1e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8645429015159607,
      "num_tokens": 364474986.0,
      "step": 10475
    },
    {
      "epoch": 1.9454038997214484,
      "grad_norm": 1.418244180850857,
      "learning_rate": 1e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8729530572891235,
      "num_tokens": 364515943.0,
      "step": 10476
    },
    {
      "epoch": 1.945589600742804,
      "grad_norm": 1.4154136691666068,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8832995891571045,
      "num_tokens": 364552788.0,
      "step": 10477
    },
    {
      "epoch": 1.9457753017641597,
      "grad_norm": 1.381245884645082,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8883363604545593,
      "num_tokens": 364590163.0,
      "step": 10478
    },
    {
      "epoch": 1.9459610027855154,
      "grad_norm": 1.6042171621645038,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8850352168083191,
      "num_tokens": 364621057.0,
      "step": 10479
    },
    {
      "epoch": 1.946146703806871,
      "grad_norm": 1.46439220228866,
      "learning_rate": 1e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8668962717056274,
      "num_tokens": 364662521.0,
      "step": 10480
    },
    {
      "epoch": 1.9463324048282264,
      "grad_norm": 1.6208415638232947,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8744767904281616,
      "num_tokens": 364694050.0,
      "step": 10481
    },
    {
      "epoch": 1.9465181058495822,
      "grad_norm": 1.652768991706644,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8743922710418701,
      "num_tokens": 364721309.0,
      "step": 10482
    },
    {
      "epoch": 1.946703806870938,
      "grad_norm": 1.4052486431445745,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8762827515602112,
      "num_tokens": 364758746.0,
      "step": 10483
    },
    {
      "epoch": 1.9468895078922934,
      "grad_norm": 1.529770870156684,
      "learning_rate": 1e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.870078444480896,
      "num_tokens": 364798744.0,
      "step": 10484
    },
    {
      "epoch": 1.947075208913649,
      "grad_norm": 1.4442755406027077,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8751646280288696,
      "num_tokens": 364836222.0,
      "step": 10485
    },
    {
      "epoch": 1.9472609099350047,
      "grad_norm": 1.6291369489410894,
      "learning_rate": 1e-06,
      "loss": 0.3606,
      "mean_token_accuracy": 0.874433696269989,
      "num_tokens": 364869500.0,
      "step": 10486
    },
    {
      "epoch": 1.9474466109563604,
      "grad_norm": 1.50516989488455,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8716636300086975,
      "num_tokens": 364904144.0,
      "step": 10487
    },
    {
      "epoch": 1.947632311977716,
      "grad_norm": 1.5554519961525022,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8764100074768066,
      "num_tokens": 364938217.0,
      "step": 10488
    },
    {
      "epoch": 1.9478180129990714,
      "grad_norm": 1.5743991282549967,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8821700811386108,
      "num_tokens": 364968425.0,
      "step": 10489
    },
    {
      "epoch": 1.9480037140204272,
      "grad_norm": 1.50060172767252,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8848389387130737,
      "num_tokens": 365003174.0,
      "step": 10490
    },
    {
      "epoch": 1.948189415041783,
      "grad_norm": 1.5475765339706704,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8845057487487793,
      "num_tokens": 365035556.0,
      "step": 10491
    },
    {
      "epoch": 1.9483751160631384,
      "grad_norm": 1.40372293933888,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8691838383674622,
      "num_tokens": 365077382.0,
      "step": 10492
    },
    {
      "epoch": 1.948560817084494,
      "grad_norm": 1.4922234685399607,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8685113191604614,
      "num_tokens": 365112671.0,
      "step": 10493
    },
    {
      "epoch": 1.9487465181058496,
      "grad_norm": 1.6658490962606904,
      "learning_rate": 1e-06,
      "loss": 0.3854,
      "mean_token_accuracy": 0.869533360004425,
      "num_tokens": 365140811.0,
      "step": 10494
    },
    {
      "epoch": 1.9489322191272052,
      "grad_norm": 1.6854136454056654,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8843801021575928,
      "num_tokens": 365168374.0,
      "step": 10495
    },
    {
      "epoch": 1.9491179201485607,
      "grad_norm": 1.4569456390109423,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8898433446884155,
      "num_tokens": 365202171.0,
      "step": 10496
    },
    {
      "epoch": 1.9493036211699164,
      "grad_norm": 1.4923280894316795,
      "learning_rate": 1e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8723979592323303,
      "num_tokens": 365237768.0,
      "step": 10497
    },
    {
      "epoch": 1.9494893221912721,
      "grad_norm": 1.5206612715114585,
      "learning_rate": 1e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8641844987869263,
      "num_tokens": 365272314.0,
      "step": 10498
    },
    {
      "epoch": 1.9496750232126276,
      "grad_norm": 1.4644869960544171,
      "learning_rate": 1e-06,
      "loss": 0.393,
      "mean_token_accuracy": 0.8656777739524841,
      "num_tokens": 365309449.0,
      "step": 10499
    },
    {
      "epoch": 1.9498607242339832,
      "grad_norm": 1.5555174973940695,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8713855147361755,
      "num_tokens": 365341035.0,
      "step": 10500
    },
    {
      "epoch": 1.9500464252553389,
      "grad_norm": 1.5700752699887566,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8850258588790894,
      "num_tokens": 365370426.0,
      "step": 10501
    },
    {
      "epoch": 1.9502321262766946,
      "grad_norm": 1.674467373130052,
      "learning_rate": 1e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8678009510040283,
      "num_tokens": 365400782.0,
      "step": 10502
    },
    {
      "epoch": 1.9504178272980501,
      "grad_norm": 1.4867645254027753,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8811346888542175,
      "num_tokens": 365435645.0,
      "step": 10503
    },
    {
      "epoch": 1.9506035283194056,
      "grad_norm": 1.6864347801346389,
      "learning_rate": 1e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8551328182220459,
      "num_tokens": 365467314.0,
      "step": 10504
    },
    {
      "epoch": 1.9507892293407614,
      "grad_norm": 1.536447614392064,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8721370697021484,
      "num_tokens": 365498899.0,
      "step": 10505
    },
    {
      "epoch": 1.950974930362117,
      "grad_norm": 1.5223119859104401,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8904008865356445,
      "num_tokens": 365531400.0,
      "step": 10506
    },
    {
      "epoch": 1.9511606313834726,
      "grad_norm": 1.329210177690776,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8958701491355896,
      "num_tokens": 365565186.0,
      "step": 10507
    },
    {
      "epoch": 1.9513463324048281,
      "grad_norm": 1.3653341506530687,
      "learning_rate": 1e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8677458167076111,
      "num_tokens": 365604847.0,
      "step": 10508
    },
    {
      "epoch": 1.9515320334261839,
      "grad_norm": 1.4292880766905274,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8780597448348999,
      "num_tokens": 365640672.0,
      "step": 10509
    },
    {
      "epoch": 1.9517177344475396,
      "grad_norm": 1.6099228773288048,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8886591196060181,
      "num_tokens": 365673945.0,
      "step": 10510
    },
    {
      "epoch": 1.951903435468895,
      "grad_norm": 1.43390897664641,
      "learning_rate": 1e-06,
      "loss": 0.3695,
      "mean_token_accuracy": 0.8735342025756836,
      "num_tokens": 365714369.0,
      "step": 10511
    },
    {
      "epoch": 1.9520891364902506,
      "grad_norm": 1.4760586216995186,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8780540823936462,
      "num_tokens": 365750367.0,
      "step": 10512
    },
    {
      "epoch": 1.9522748375116064,
      "grad_norm": 1.5882562124146806,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8651286363601685,
      "num_tokens": 365786064.0,
      "step": 10513
    },
    {
      "epoch": 1.952460538532962,
      "grad_norm": 1.5732996225401283,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8703610897064209,
      "num_tokens": 365819839.0,
      "step": 10514
    },
    {
      "epoch": 1.9526462395543176,
      "grad_norm": 1.3903655203850027,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8804123997688293,
      "num_tokens": 365857917.0,
      "step": 10515
    },
    {
      "epoch": 1.952831940575673,
      "grad_norm": 1.6536569567039654,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8791329860687256,
      "num_tokens": 365886015.0,
      "step": 10516
    },
    {
      "epoch": 1.9530176415970288,
      "grad_norm": 1.6253638901300154,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.888757050037384,
      "num_tokens": 365914751.0,
      "step": 10517
    },
    {
      "epoch": 1.9532033426183844,
      "grad_norm": 1.458504152964962,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8688992261886597,
      "num_tokens": 365956154.0,
      "step": 10518
    },
    {
      "epoch": 1.9533890436397399,
      "grad_norm": 1.434549454864699,
      "learning_rate": 1e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8635528087615967,
      "num_tokens": 365996330.0,
      "step": 10519
    },
    {
      "epoch": 1.9535747446610956,
      "grad_norm": 1.3286640501356104,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8800680637359619,
      "num_tokens": 366036634.0,
      "step": 10520
    },
    {
      "epoch": 1.9537604456824513,
      "grad_norm": 1.5745995607703982,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8647904992103577,
      "num_tokens": 366069291.0,
      "step": 10521
    },
    {
      "epoch": 1.9539461467038068,
      "grad_norm": 1.5943204381621483,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8782473802566528,
      "num_tokens": 366100060.0,
      "step": 10522
    },
    {
      "epoch": 1.9541318477251624,
      "grad_norm": 1.586536565555229,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8851536512374878,
      "num_tokens": 366125551.0,
      "step": 10523
    },
    {
      "epoch": 1.954317548746518,
      "grad_norm": 1.3982187894212468,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8803902268409729,
      "num_tokens": 366164821.0,
      "step": 10524
    },
    {
      "epoch": 1.9545032497678738,
      "grad_norm": 1.4796012927116315,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8754512071609497,
      "num_tokens": 366201529.0,
      "step": 10525
    },
    {
      "epoch": 1.9546889507892293,
      "grad_norm": 1.3959827823796809,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8753543496131897,
      "num_tokens": 366236877.0,
      "step": 10526
    },
    {
      "epoch": 1.9548746518105848,
      "grad_norm": 1.4459341650327833,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8745743036270142,
      "num_tokens": 366272351.0,
      "step": 10527
    },
    {
      "epoch": 1.9550603528319406,
      "grad_norm": 1.6303630576343406,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8767710328102112,
      "num_tokens": 366307751.0,
      "step": 10528
    },
    {
      "epoch": 1.9552460538532963,
      "grad_norm": 1.5570608910680583,
      "learning_rate": 1e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8708751797676086,
      "num_tokens": 366344882.0,
      "step": 10529
    },
    {
      "epoch": 1.9554317548746518,
      "grad_norm": 1.510267203859273,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8750172853469849,
      "num_tokens": 366382016.0,
      "step": 10530
    },
    {
      "epoch": 1.9556174558960073,
      "grad_norm": 1.4263942883620384,
      "learning_rate": 1e-06,
      "loss": 0.2802,
      "mean_token_accuracy": 0.8988995552062988,
      "num_tokens": 366414115.0,
      "step": 10531
    },
    {
      "epoch": 1.955803156917363,
      "grad_norm": 1.4421419043335133,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8726279735565186,
      "num_tokens": 366452372.0,
      "step": 10532
    },
    {
      "epoch": 1.9559888579387188,
      "grad_norm": 1.5715366225898706,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8760293126106262,
      "num_tokens": 366483305.0,
      "step": 10533
    },
    {
      "epoch": 1.9561745589600743,
      "grad_norm": 1.5520536118023336,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.892045259475708,
      "num_tokens": 366516493.0,
      "step": 10534
    },
    {
      "epoch": 1.9563602599814298,
      "grad_norm": 1.4036922879854556,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8861303329467773,
      "num_tokens": 366555718.0,
      "step": 10535
    },
    {
      "epoch": 1.9565459610027855,
      "grad_norm": 1.568766288073073,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8784796595573425,
      "num_tokens": 366590220.0,
      "step": 10536
    },
    {
      "epoch": 1.9567316620241413,
      "grad_norm": 1.4370375391037127,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.870283842086792,
      "num_tokens": 366629432.0,
      "step": 10537
    },
    {
      "epoch": 1.9569173630454968,
      "grad_norm": 1.5042851516137177,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8824799060821533,
      "num_tokens": 366664020.0,
      "step": 10538
    },
    {
      "epoch": 1.9571030640668523,
      "grad_norm": 1.4831065098305218,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8878905773162842,
      "num_tokens": 366698204.0,
      "step": 10539
    },
    {
      "epoch": 1.957288765088208,
      "grad_norm": 1.4578670141506846,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8812838196754456,
      "num_tokens": 366733344.0,
      "step": 10540
    },
    {
      "epoch": 1.9574744661095635,
      "grad_norm": 1.6222104369634633,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8685176372528076,
      "num_tokens": 366769992.0,
      "step": 10541
    },
    {
      "epoch": 1.957660167130919,
      "grad_norm": 1.5308018997393966,
      "learning_rate": 1e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8626888990402222,
      "num_tokens": 366810384.0,
      "step": 10542
    },
    {
      "epoch": 1.9578458681522748,
      "grad_norm": 1.4549173324044826,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8864498734474182,
      "num_tokens": 366845953.0,
      "step": 10543
    },
    {
      "epoch": 1.9580315691736305,
      "grad_norm": 1.53131544761099,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8688027858734131,
      "num_tokens": 366883274.0,
      "step": 10544
    },
    {
      "epoch": 1.958217270194986,
      "grad_norm": 1.5505580043011769,
      "learning_rate": 1e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8702293634414673,
      "num_tokens": 366917119.0,
      "step": 10545
    },
    {
      "epoch": 1.9584029712163415,
      "grad_norm": 1.6411767811881124,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8713550567626953,
      "num_tokens": 366947969.0,
      "step": 10546
    },
    {
      "epoch": 1.9585886722376973,
      "grad_norm": 1.377187561294068,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8798598051071167,
      "num_tokens": 366988092.0,
      "step": 10547
    },
    {
      "epoch": 1.958774373259053,
      "grad_norm": 1.5515740072706479,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8879383206367493,
      "num_tokens": 367016364.0,
      "step": 10548
    },
    {
      "epoch": 1.9589600742804085,
      "grad_norm": 1.4906497868688138,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8724489212036133,
      "num_tokens": 367051544.0,
      "step": 10549
    },
    {
      "epoch": 1.959145775301764,
      "grad_norm": 1.4784098798799112,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.869506299495697,
      "num_tokens": 367084059.0,
      "step": 10550
    },
    {
      "epoch": 1.9593314763231198,
      "grad_norm": 1.580713285914278,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8735393285751343,
      "num_tokens": 367116323.0,
      "step": 10551
    },
    {
      "epoch": 1.9595171773444755,
      "grad_norm": 1.5136855095071773,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8812893629074097,
      "num_tokens": 367154076.0,
      "step": 10552
    },
    {
      "epoch": 1.959702878365831,
      "grad_norm": 1.4763479082125996,
      "learning_rate": 1e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.8653560876846313,
      "num_tokens": 367192524.0,
      "step": 10553
    },
    {
      "epoch": 1.9598885793871865,
      "grad_norm": 1.4452949786798273,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8770866394042969,
      "num_tokens": 367229254.0,
      "step": 10554
    },
    {
      "epoch": 1.9600742804085423,
      "grad_norm": 1.5114201391501478,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.8760219216346741,
      "num_tokens": 367261530.0,
      "step": 10555
    },
    {
      "epoch": 1.960259981429898,
      "grad_norm": 1.473493688766215,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8761860132217407,
      "num_tokens": 367297430.0,
      "step": 10556
    },
    {
      "epoch": 1.9604456824512535,
      "grad_norm": 1.5432710807928134,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8754109740257263,
      "num_tokens": 367333521.0,
      "step": 10557
    },
    {
      "epoch": 1.960631383472609,
      "grad_norm": 1.4176678614496185,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.8896549940109253,
      "num_tokens": 367367137.0,
      "step": 10558
    },
    {
      "epoch": 1.9608170844939647,
      "grad_norm": 1.563320834877691,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.883310854434967,
      "num_tokens": 367396291.0,
      "step": 10559
    },
    {
      "epoch": 1.9610027855153205,
      "grad_norm": 1.5831209450658716,
      "learning_rate": 1e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8598123788833618,
      "num_tokens": 367431566.0,
      "step": 10560
    },
    {
      "epoch": 1.961188486536676,
      "grad_norm": 1.6135867643531605,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8761910200119019,
      "num_tokens": 367462025.0,
      "step": 10561
    },
    {
      "epoch": 1.9613741875580315,
      "grad_norm": 1.7332814836661758,
      "learning_rate": 1e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.8596417307853699,
      "num_tokens": 367497509.0,
      "step": 10562
    },
    {
      "epoch": 1.9615598885793872,
      "grad_norm": 1.473812982920192,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8861067295074463,
      "num_tokens": 367532815.0,
      "step": 10563
    },
    {
      "epoch": 1.961745589600743,
      "grad_norm": 1.491831184937671,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.8978245854377747,
      "num_tokens": 367566557.0,
      "step": 10564
    },
    {
      "epoch": 1.9619312906220983,
      "grad_norm": 1.4504621932440795,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8900640606880188,
      "num_tokens": 367601415.0,
      "step": 10565
    },
    {
      "epoch": 1.962116991643454,
      "grad_norm": 1.6548607651494471,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8935750722885132,
      "num_tokens": 367627518.0,
      "step": 10566
    },
    {
      "epoch": 1.9623026926648097,
      "grad_norm": 1.2962914635744143,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8839736580848694,
      "num_tokens": 367669151.0,
      "step": 10567
    },
    {
      "epoch": 1.9624883936861652,
      "grad_norm": 1.431902872681272,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8706256747245789,
      "num_tokens": 367707200.0,
      "step": 10568
    },
    {
      "epoch": 1.9626740947075207,
      "grad_norm": 1.5413898460876776,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.886042594909668,
      "num_tokens": 367736801.0,
      "step": 10569
    },
    {
      "epoch": 1.9628597957288765,
      "grad_norm": 1.4825702649084815,
      "learning_rate": 1e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8674460649490356,
      "num_tokens": 367771480.0,
      "step": 10570
    },
    {
      "epoch": 1.9630454967502322,
      "grad_norm": 1.4954660595298257,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8840030431747437,
      "num_tokens": 367802053.0,
      "step": 10571
    },
    {
      "epoch": 1.9632311977715877,
      "grad_norm": 1.5632824235497662,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8675352334976196,
      "num_tokens": 367835078.0,
      "step": 10572
    },
    {
      "epoch": 1.9634168987929432,
      "grad_norm": 1.6527760085840024,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8708847761154175,
      "num_tokens": 367865140.0,
      "step": 10573
    },
    {
      "epoch": 1.963602599814299,
      "grad_norm": 1.5445359048362275,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8802511692047119,
      "num_tokens": 367896564.0,
      "step": 10574
    },
    {
      "epoch": 1.9637883008356547,
      "grad_norm": 1.4565210129190689,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8745753765106201,
      "num_tokens": 367937839.0,
      "step": 10575
    },
    {
      "epoch": 1.9639740018570102,
      "grad_norm": 1.5056479145018726,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8719607591629028,
      "num_tokens": 367972677.0,
      "step": 10576
    },
    {
      "epoch": 1.9641597028783657,
      "grad_norm": 1.3863640614034887,
      "learning_rate": 1e-06,
      "loss": 0.3174,
      "mean_token_accuracy": 0.8901098966598511,
      "num_tokens": 368007829.0,
      "step": 10577
    },
    {
      "epoch": 1.9643454038997215,
      "grad_norm": 1.4499886232299708,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8843831419944763,
      "num_tokens": 368045752.0,
      "step": 10578
    },
    {
      "epoch": 1.9645311049210772,
      "grad_norm": 1.5908841331714898,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8823222517967224,
      "num_tokens": 368082219.0,
      "step": 10579
    },
    {
      "epoch": 1.9647168059424327,
      "grad_norm": 1.7061977904686427,
      "learning_rate": 1e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8660424947738647,
      "num_tokens": 368111254.0,
      "step": 10580
    },
    {
      "epoch": 1.9649025069637882,
      "grad_norm": 1.3210035691499185,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8869588375091553,
      "num_tokens": 368149555.0,
      "step": 10581
    },
    {
      "epoch": 1.965088207985144,
      "grad_norm": 1.8297341605417567,
      "learning_rate": 1e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8693716526031494,
      "num_tokens": 368174743.0,
      "step": 10582
    },
    {
      "epoch": 1.9652739090064997,
      "grad_norm": 1.6412521272273843,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8817116022109985,
      "num_tokens": 368204177.0,
      "step": 10583
    },
    {
      "epoch": 1.9654596100278552,
      "grad_norm": 1.4205273369985292,
      "learning_rate": 1e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8660314083099365,
      "num_tokens": 368246949.0,
      "step": 10584
    },
    {
      "epoch": 1.9656453110492107,
      "grad_norm": 1.5789814883588857,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8723887205123901,
      "num_tokens": 368279370.0,
      "step": 10585
    },
    {
      "epoch": 1.9658310120705664,
      "grad_norm": 1.528216606547901,
      "learning_rate": 1e-06,
      "loss": 0.3637,
      "mean_token_accuracy": 0.8700482249259949,
      "num_tokens": 368312238.0,
      "step": 10586
    },
    {
      "epoch": 1.9660167130919222,
      "grad_norm": 1.3702042804550176,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.8929412364959717,
      "num_tokens": 368347882.0,
      "step": 10587
    },
    {
      "epoch": 1.9662024141132777,
      "grad_norm": 1.4648335521668294,
      "learning_rate": 1e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8644365668296814,
      "num_tokens": 368386223.0,
      "step": 10588
    },
    {
      "epoch": 1.9663881151346332,
      "grad_norm": 1.5119859424675977,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8713108897209167,
      "num_tokens": 368419414.0,
      "step": 10589
    },
    {
      "epoch": 1.966573816155989,
      "grad_norm": 1.6278690908481712,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8761985301971436,
      "num_tokens": 368450208.0,
      "step": 10590
    },
    {
      "epoch": 1.9667595171773444,
      "grad_norm": 1.437822368930506,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8871413469314575,
      "num_tokens": 368485549.0,
      "step": 10591
    },
    {
      "epoch": 1.9669452181987,
      "grad_norm": 1.4383700679817826,
      "learning_rate": 1e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8568995594978333,
      "num_tokens": 368525389.0,
      "step": 10592
    },
    {
      "epoch": 1.9671309192200557,
      "grad_norm": 1.4209207121670868,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.8934808969497681,
      "num_tokens": 368560431.0,
      "step": 10593
    },
    {
      "epoch": 1.9673166202414114,
      "grad_norm": 1.5811969576375204,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.871345043182373,
      "num_tokens": 368593063.0,
      "step": 10594
    },
    {
      "epoch": 1.967502321262767,
      "grad_norm": 1.4956159629836836,
      "learning_rate": 1e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8651472330093384,
      "num_tokens": 368629654.0,
      "step": 10595
    },
    {
      "epoch": 1.9676880222841224,
      "grad_norm": 1.6604332629373932,
      "learning_rate": 1e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8776613473892212,
      "num_tokens": 368663603.0,
      "step": 10596
    },
    {
      "epoch": 1.9678737233054782,
      "grad_norm": 1.4666403223614237,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8826883435249329,
      "num_tokens": 368696863.0,
      "step": 10597
    },
    {
      "epoch": 1.968059424326834,
      "grad_norm": 1.2900313651277313,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.89236980676651,
      "num_tokens": 368737516.0,
      "step": 10598
    },
    {
      "epoch": 1.9682451253481894,
      "grad_norm": 1.5194737435473917,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8814082145690918,
      "num_tokens": 368767950.0,
      "step": 10599
    },
    {
      "epoch": 1.968430826369545,
      "grad_norm": 1.4850629852804083,
      "learning_rate": 1e-06,
      "loss": 0.3005,
      "mean_token_accuracy": 0.8916124105453491,
      "num_tokens": 368799429.0,
      "step": 10600
    },
    {
      "epoch": 1.9686165273909007,
      "grad_norm": 1.5441168944053625,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8790754079818726,
      "num_tokens": 368832986.0,
      "step": 10601
    },
    {
      "epoch": 1.9688022284122564,
      "grad_norm": 1.4339826628423278,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8751593828201294,
      "num_tokens": 368868417.0,
      "step": 10602
    },
    {
      "epoch": 1.968987929433612,
      "grad_norm": 1.5801550221321485,
      "learning_rate": 1e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8489822149276733,
      "num_tokens": 368902357.0,
      "step": 10603
    },
    {
      "epoch": 1.9691736304549674,
      "grad_norm": 1.6682034396876182,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8747996091842651,
      "num_tokens": 368928895.0,
      "step": 10604
    },
    {
      "epoch": 1.9693593314763231,
      "grad_norm": 1.4831374414495473,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8736832141876221,
      "num_tokens": 368962056.0,
      "step": 10605
    },
    {
      "epoch": 1.9695450324976789,
      "grad_norm": 1.448419331709317,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8893406391143799,
      "num_tokens": 368995120.0,
      "step": 10606
    },
    {
      "epoch": 1.9697307335190344,
      "grad_norm": 1.3163737560992637,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8867692947387695,
      "num_tokens": 369035605.0,
      "step": 10607
    },
    {
      "epoch": 1.96991643454039,
      "grad_norm": 1.7015661491122978,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8582078218460083,
      "num_tokens": 369068257.0,
      "step": 10608
    },
    {
      "epoch": 1.9701021355617456,
      "grad_norm": 1.3155656657783161,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8907575607299805,
      "num_tokens": 369108362.0,
      "step": 10609
    },
    {
      "epoch": 1.9702878365831014,
      "grad_norm": 1.5546502113750056,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8776801824569702,
      "num_tokens": 369143177.0,
      "step": 10610
    },
    {
      "epoch": 1.9704735376044569,
      "grad_norm": 1.618364319181353,
      "learning_rate": 1e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.862338662147522,
      "num_tokens": 369176475.0,
      "step": 10611
    },
    {
      "epoch": 1.9706592386258124,
      "grad_norm": 1.4933996744531701,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.888106107711792,
      "num_tokens": 369208109.0,
      "step": 10612
    },
    {
      "epoch": 1.9708449396471681,
      "grad_norm": 1.5428864883060698,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8710603713989258,
      "num_tokens": 369245672.0,
      "step": 10613
    },
    {
      "epoch": 1.9710306406685236,
      "grad_norm": 1.5446994329089865,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8881743550300598,
      "num_tokens": 369274448.0,
      "step": 10614
    },
    {
      "epoch": 1.9712163416898791,
      "grad_norm": 1.3306381197037278,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8829955458641052,
      "num_tokens": 369317036.0,
      "step": 10615
    },
    {
      "epoch": 1.9714020427112349,
      "grad_norm": 1.4845954289796879,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8731099963188171,
      "num_tokens": 369352647.0,
      "step": 10616
    },
    {
      "epoch": 1.9715877437325906,
      "grad_norm": 1.4962038413301681,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8732558488845825,
      "num_tokens": 369385599.0,
      "step": 10617
    },
    {
      "epoch": 1.9717734447539461,
      "grad_norm": 1.453189273990642,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8708423376083374,
      "num_tokens": 369424803.0,
      "step": 10618
    },
    {
      "epoch": 1.9719591457753016,
      "grad_norm": 1.5987781649910855,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8834262490272522,
      "num_tokens": 369452473.0,
      "step": 10619
    },
    {
      "epoch": 1.9721448467966574,
      "grad_norm": 1.615725856210639,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8867034316062927,
      "num_tokens": 369480395.0,
      "step": 10620
    },
    {
      "epoch": 1.972330547818013,
      "grad_norm": 1.3826275523223714,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8772032260894775,
      "num_tokens": 369519717.0,
      "step": 10621
    },
    {
      "epoch": 1.9725162488393686,
      "grad_norm": 1.5918449182275003,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8783277273178101,
      "num_tokens": 369548463.0,
      "step": 10622
    },
    {
      "epoch": 1.9727019498607241,
      "grad_norm": 1.3994337760491338,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8845059871673584,
      "num_tokens": 369586009.0,
      "step": 10623
    },
    {
      "epoch": 1.9728876508820798,
      "grad_norm": 1.3758914052814302,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8884419202804565,
      "num_tokens": 369623908.0,
      "step": 10624
    },
    {
      "epoch": 1.9730733519034356,
      "grad_norm": 1.4231585625914842,
      "learning_rate": 1e-06,
      "loss": 0.2839,
      "mean_token_accuracy": 0.8979185819625854,
      "num_tokens": 369655046.0,
      "step": 10625
    },
    {
      "epoch": 1.973259052924791,
      "grad_norm": 1.4213008287922826,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8906422853469849,
      "num_tokens": 369688659.0,
      "step": 10626
    },
    {
      "epoch": 1.9734447539461466,
      "grad_norm": 1.3359679642999438,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8648964762687683,
      "num_tokens": 369735089.0,
      "step": 10627
    },
    {
      "epoch": 1.9736304549675023,
      "grad_norm": 1.534607316313988,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.870917022228241,
      "num_tokens": 369772177.0,
      "step": 10628
    },
    {
      "epoch": 1.973816155988858,
      "grad_norm": 1.5172086777637792,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8755120635032654,
      "num_tokens": 369807882.0,
      "step": 10629
    },
    {
      "epoch": 1.9740018570102136,
      "grad_norm": 1.696474436999729,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8753132224082947,
      "num_tokens": 369834095.0,
      "step": 10630
    },
    {
      "epoch": 1.974187558031569,
      "grad_norm": 1.5560525242223444,
      "learning_rate": 1e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8649147748947144,
      "num_tokens": 369870483.0,
      "step": 10631
    },
    {
      "epoch": 1.9743732590529248,
      "grad_norm": 1.5810736220526513,
      "learning_rate": 1e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8642474412918091,
      "num_tokens": 369905365.0,
      "step": 10632
    },
    {
      "epoch": 1.9745589600742806,
      "grad_norm": 1.7362158481496137,
      "learning_rate": 1e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8644192814826965,
      "num_tokens": 369939694.0,
      "step": 10633
    },
    {
      "epoch": 1.974744661095636,
      "grad_norm": 1.4698387120276495,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8764699101448059,
      "num_tokens": 369976885.0,
      "step": 10634
    },
    {
      "epoch": 1.9749303621169916,
      "grad_norm": 1.515267672039782,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8812077045440674,
      "num_tokens": 370010333.0,
      "step": 10635
    },
    {
      "epoch": 1.9751160631383473,
      "grad_norm": 1.5769209439009606,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8704016804695129,
      "num_tokens": 370043207.0,
      "step": 10636
    },
    {
      "epoch": 1.975301764159703,
      "grad_norm": 1.4746015890708508,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8829897046089172,
      "num_tokens": 370077927.0,
      "step": 10637
    },
    {
      "epoch": 1.9754874651810583,
      "grad_norm": 1.432366479000388,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8893465399742126,
      "num_tokens": 370114619.0,
      "step": 10638
    },
    {
      "epoch": 1.975673166202414,
      "grad_norm": 1.4808873299396772,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8712368011474609,
      "num_tokens": 370149496.0,
      "step": 10639
    },
    {
      "epoch": 1.9758588672237698,
      "grad_norm": 1.4328947116028574,
      "learning_rate": 1e-06,
      "loss": 0.2969,
      "mean_token_accuracy": 0.8948130011558533,
      "num_tokens": 370181763.0,
      "step": 10640
    },
    {
      "epoch": 1.9760445682451253,
      "grad_norm": 1.565950359520713,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8765134811401367,
      "num_tokens": 370210783.0,
      "step": 10641
    },
    {
      "epoch": 1.9762302692664808,
      "grad_norm": 1.6120242545638017,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8813404440879822,
      "num_tokens": 370241106.0,
      "step": 10642
    },
    {
      "epoch": 1.9764159702878366,
      "grad_norm": 1.3664889181199418,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8801043033599854,
      "num_tokens": 370282114.0,
      "step": 10643
    },
    {
      "epoch": 1.9766016713091923,
      "grad_norm": 1.5813598063116405,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8820422887802124,
      "num_tokens": 370318196.0,
      "step": 10644
    },
    {
      "epoch": 1.9767873723305478,
      "grad_norm": 1.5370180269249967,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8936536312103271,
      "num_tokens": 370352870.0,
      "step": 10645
    },
    {
      "epoch": 1.9769730733519033,
      "grad_norm": 1.4431555074199867,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.874322235584259,
      "num_tokens": 370393590.0,
      "step": 10646
    },
    {
      "epoch": 1.977158774373259,
      "grad_norm": 1.477515996260581,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8861744403839111,
      "num_tokens": 370428120.0,
      "step": 10647
    },
    {
      "epoch": 1.9773444753946148,
      "grad_norm": 1.615695926460314,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8802734613418579,
      "num_tokens": 370459236.0,
      "step": 10648
    },
    {
      "epoch": 1.9775301764159703,
      "grad_norm": 1.3758137799526529,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8893316984176636,
      "num_tokens": 370496510.0,
      "step": 10649
    },
    {
      "epoch": 1.9777158774373258,
      "grad_norm": 1.5477936233376501,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8730823397636414,
      "num_tokens": 370530094.0,
      "step": 10650
    },
    {
      "epoch": 1.9779015784586815,
      "grad_norm": 1.464825409908375,
      "learning_rate": 1e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8707553148269653,
      "num_tokens": 370565080.0,
      "step": 10651
    },
    {
      "epoch": 1.9780872794800373,
      "grad_norm": 1.408425859897239,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.86676025390625,
      "num_tokens": 370608227.0,
      "step": 10652
    },
    {
      "epoch": 1.9782729805013928,
      "grad_norm": 1.6037668479855984,
      "learning_rate": 1e-06,
      "loss": 0.3928,
      "mean_token_accuracy": 0.868872880935669,
      "num_tokens": 370643527.0,
      "step": 10653
    },
    {
      "epoch": 1.9784586815227483,
      "grad_norm": 1.566282788913383,
      "learning_rate": 1e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8784679174423218,
      "num_tokens": 370676139.0,
      "step": 10654
    },
    {
      "epoch": 1.978644382544104,
      "grad_norm": 1.5383614059057378,
      "learning_rate": 1e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8638498783111572,
      "num_tokens": 370709847.0,
      "step": 10655
    },
    {
      "epoch": 1.9788300835654598,
      "grad_norm": 1.4308012539376131,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8769079446792603,
      "num_tokens": 370749275.0,
      "step": 10656
    },
    {
      "epoch": 1.9790157845868153,
      "grad_norm": 1.521784711758997,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8868223428726196,
      "num_tokens": 370782850.0,
      "step": 10657
    },
    {
      "epoch": 1.9792014856081708,
      "grad_norm": 1.4442375124451108,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.881898045539856,
      "num_tokens": 370821274.0,
      "step": 10658
    },
    {
      "epoch": 1.9793871866295265,
      "grad_norm": 1.4191782783918612,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8724638819694519,
      "num_tokens": 370858962.0,
      "step": 10659
    },
    {
      "epoch": 1.9795728876508822,
      "grad_norm": 1.5387874505556989,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8700117468833923,
      "num_tokens": 370894888.0,
      "step": 10660
    },
    {
      "epoch": 1.9797585886722378,
      "grad_norm": 1.3634279284157045,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8868406414985657,
      "num_tokens": 370932009.0,
      "step": 10661
    },
    {
      "epoch": 1.9799442896935933,
      "grad_norm": 1.5271050761918783,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.8932945728302002,
      "num_tokens": 370967851.0,
      "step": 10662
    },
    {
      "epoch": 1.980129990714949,
      "grad_norm": 1.4851142991648905,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8701257705688477,
      "num_tokens": 371003028.0,
      "step": 10663
    },
    {
      "epoch": 1.9803156917363045,
      "grad_norm": 1.5233663042095404,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8784309029579163,
      "num_tokens": 371034407.0,
      "step": 10664
    },
    {
      "epoch": 1.98050139275766,
      "grad_norm": 1.5811667180758515,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8864925503730774,
      "num_tokens": 371067117.0,
      "step": 10665
    },
    {
      "epoch": 1.9806870937790158,
      "grad_norm": 1.5308652935899556,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8824955224990845,
      "num_tokens": 371101624.0,
      "step": 10666
    },
    {
      "epoch": 1.9808727948003715,
      "grad_norm": 1.599016628396812,
      "learning_rate": 1e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8578386306762695,
      "num_tokens": 371138904.0,
      "step": 10667
    },
    {
      "epoch": 1.981058495821727,
      "grad_norm": 1.4383571684209848,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8815680742263794,
      "num_tokens": 371174532.0,
      "step": 10668
    },
    {
      "epoch": 1.9812441968430825,
      "grad_norm": 1.4900880853054008,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8781152367591858,
      "num_tokens": 371211093.0,
      "step": 10669
    },
    {
      "epoch": 1.9814298978644382,
      "grad_norm": 1.6247747718759125,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8723194599151611,
      "num_tokens": 371241457.0,
      "step": 10670
    },
    {
      "epoch": 1.981615598885794,
      "grad_norm": 1.4098386977095936,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8816858530044556,
      "num_tokens": 371279819.0,
      "step": 10671
    },
    {
      "epoch": 1.9818012999071495,
      "grad_norm": 1.5250116987453288,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8863956332206726,
      "num_tokens": 371314513.0,
      "step": 10672
    },
    {
      "epoch": 1.981987000928505,
      "grad_norm": 1.5564900487358908,
      "learning_rate": 1e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.8739218711853027,
      "num_tokens": 371348010.0,
      "step": 10673
    },
    {
      "epoch": 1.9821727019498607,
      "grad_norm": 1.4654876975711784,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8878961205482483,
      "num_tokens": 371381084.0,
      "step": 10674
    },
    {
      "epoch": 1.9823584029712165,
      "grad_norm": 1.460271390715603,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8701281547546387,
      "num_tokens": 371418715.0,
      "step": 10675
    },
    {
      "epoch": 1.982544103992572,
      "grad_norm": 1.3953347749880938,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8952528238296509,
      "num_tokens": 371454734.0,
      "step": 10676
    },
    {
      "epoch": 1.9827298050139275,
      "grad_norm": 1.6882466673144585,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8778847455978394,
      "num_tokens": 371481202.0,
      "step": 10677
    },
    {
      "epoch": 1.9829155060352832,
      "grad_norm": 1.3884268680497838,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8691501617431641,
      "num_tokens": 371522584.0,
      "step": 10678
    },
    {
      "epoch": 1.983101207056639,
      "grad_norm": 1.56771046345441,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.878433346748352,
      "num_tokens": 371555600.0,
      "step": 10679
    },
    {
      "epoch": 1.9832869080779945,
      "grad_norm": 1.4435300345866076,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8791893720626831,
      "num_tokens": 371589235.0,
      "step": 10680
    },
    {
      "epoch": 1.98347260909935,
      "grad_norm": 1.576006147043875,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8728183507919312,
      "num_tokens": 371620581.0,
      "step": 10681
    },
    {
      "epoch": 1.9836583101207057,
      "grad_norm": 1.4273497878955188,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8837945461273193,
      "num_tokens": 371655958.0,
      "step": 10682
    },
    {
      "epoch": 1.9838440111420614,
      "grad_norm": 1.6575702497381806,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8758325576782227,
      "num_tokens": 371684537.0,
      "step": 10683
    },
    {
      "epoch": 1.984029712163417,
      "grad_norm": 1.389442451584794,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8702877759933472,
      "num_tokens": 371725683.0,
      "step": 10684
    },
    {
      "epoch": 1.9842154131847725,
      "grad_norm": 1.444041290199748,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8809617757797241,
      "num_tokens": 371761768.0,
      "step": 10685
    },
    {
      "epoch": 1.9844011142061282,
      "grad_norm": 1.458758437602876,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8756154775619507,
      "num_tokens": 371796777.0,
      "step": 10686
    },
    {
      "epoch": 1.9845868152274837,
      "grad_norm": 1.5867002769953695,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8783435225486755,
      "num_tokens": 371827927.0,
      "step": 10687
    },
    {
      "epoch": 1.9847725162488392,
      "grad_norm": 1.4908432328507046,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8894906044006348,
      "num_tokens": 371864324.0,
      "step": 10688
    },
    {
      "epoch": 1.984958217270195,
      "grad_norm": 1.5706797149473792,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8778094053268433,
      "num_tokens": 371901436.0,
      "step": 10689
    },
    {
      "epoch": 1.9851439182915507,
      "grad_norm": 1.3330829779556705,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8771428465843201,
      "num_tokens": 371946745.0,
      "step": 10690
    },
    {
      "epoch": 1.9853296193129062,
      "grad_norm": 1.3934503040516917,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8841714859008789,
      "num_tokens": 371985582.0,
      "step": 10691
    },
    {
      "epoch": 1.9855153203342617,
      "grad_norm": 1.545994403161345,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.88136225938797,
      "num_tokens": 372019204.0,
      "step": 10692
    },
    {
      "epoch": 1.9857010213556174,
      "grad_norm": 1.659071602799536,
      "learning_rate": 1e-06,
      "loss": 0.3872,
      "mean_token_accuracy": 0.8724742531776428,
      "num_tokens": 372048976.0,
      "step": 10693
    },
    {
      "epoch": 1.9858867223769732,
      "grad_norm": 1.4115655103163787,
      "learning_rate": 1e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.8661971688270569,
      "num_tokens": 372090373.0,
      "step": 10694
    },
    {
      "epoch": 1.9860724233983287,
      "grad_norm": 1.35626264254889,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8925051093101501,
      "num_tokens": 372127264.0,
      "step": 10695
    },
    {
      "epoch": 1.9862581244196842,
      "grad_norm": 1.4882004266732387,
      "learning_rate": 1e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8693920373916626,
      "num_tokens": 372166997.0,
      "step": 10696
    },
    {
      "epoch": 1.98644382544104,
      "grad_norm": 1.4668472463077826,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8768446445465088,
      "num_tokens": 372201243.0,
      "step": 10697
    },
    {
      "epoch": 1.9866295264623957,
      "grad_norm": 1.322801269697461,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8954319953918457,
      "num_tokens": 372241495.0,
      "step": 10698
    },
    {
      "epoch": 1.9868152274837512,
      "grad_norm": 1.4100549534294649,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8729318976402283,
      "num_tokens": 372280497.0,
      "step": 10699
    },
    {
      "epoch": 1.9870009285051067,
      "grad_norm": 1.5544375514476862,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8898566365242004,
      "num_tokens": 372314114.0,
      "step": 10700
    },
    {
      "epoch": 1.9871866295264624,
      "grad_norm": 1.4700592193885675,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8840312957763672,
      "num_tokens": 372350423.0,
      "step": 10701
    },
    {
      "epoch": 1.9873723305478181,
      "grad_norm": 1.5644015941813765,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8803391456604004,
      "num_tokens": 372384423.0,
      "step": 10702
    },
    {
      "epoch": 1.9875580315691737,
      "grad_norm": 1.584570519564557,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8699957728385925,
      "num_tokens": 372418563.0,
      "step": 10703
    },
    {
      "epoch": 1.9877437325905292,
      "grad_norm": 1.4886216076566563,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8897584676742554,
      "num_tokens": 372453445.0,
      "step": 10704
    },
    {
      "epoch": 1.987929433611885,
      "grad_norm": 1.4753647715062075,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8775167465209961,
      "num_tokens": 372489001.0,
      "step": 10705
    },
    {
      "epoch": 1.9881151346332406,
      "grad_norm": 1.4858886714036648,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8858014345169067,
      "num_tokens": 372523622.0,
      "step": 10706
    },
    {
      "epoch": 1.9883008356545961,
      "grad_norm": 1.4379284278949775,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.889228105545044,
      "num_tokens": 372561657.0,
      "step": 10707
    },
    {
      "epoch": 1.9884865366759517,
      "grad_norm": 1.6866009928191754,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8835740089416504,
      "num_tokens": 372588545.0,
      "step": 10708
    },
    {
      "epoch": 1.9886722376973074,
      "grad_norm": 1.55580209042763,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8895862102508545,
      "num_tokens": 372625086.0,
      "step": 10709
    },
    {
      "epoch": 1.988857938718663,
      "grad_norm": 1.5449046617069906,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8776358366012573,
      "num_tokens": 372658085.0,
      "step": 10710
    },
    {
      "epoch": 1.9890436397400184,
      "grad_norm": 1.5527070133628653,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8772662281990051,
      "num_tokens": 372690053.0,
      "step": 10711
    },
    {
      "epoch": 1.9892293407613741,
      "grad_norm": 1.6272997409082612,
      "learning_rate": 1e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8624322414398193,
      "num_tokens": 372721077.0,
      "step": 10712
    },
    {
      "epoch": 1.9894150417827299,
      "grad_norm": 1.5917465455330453,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8745756149291992,
      "num_tokens": 372754107.0,
      "step": 10713
    },
    {
      "epoch": 1.9896007428040854,
      "grad_norm": 1.4228643559473506,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8880653381347656,
      "num_tokens": 372791445.0,
      "step": 10714
    },
    {
      "epoch": 1.989786443825441,
      "grad_norm": 1.3871545912588856,
      "learning_rate": 1e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.8658716678619385,
      "num_tokens": 372834567.0,
      "step": 10715
    },
    {
      "epoch": 1.9899721448467966,
      "grad_norm": 1.5636250998671442,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.891571044921875,
      "num_tokens": 372867426.0,
      "step": 10716
    },
    {
      "epoch": 1.9901578458681524,
      "grad_norm": 1.4769843903086453,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8772098422050476,
      "num_tokens": 372904108.0,
      "step": 10717
    },
    {
      "epoch": 1.9903435468895079,
      "grad_norm": 1.4247373745125287,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8837437629699707,
      "num_tokens": 372940430.0,
      "step": 10718
    },
    {
      "epoch": 1.9905292479108634,
      "grad_norm": 1.4335860637690618,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.882391631603241,
      "num_tokens": 372977120.0,
      "step": 10719
    },
    {
      "epoch": 1.9907149489322191,
      "grad_norm": 1.4729840016066578,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8805514574050903,
      "num_tokens": 373013353.0,
      "step": 10720
    },
    {
      "epoch": 1.9909006499535749,
      "grad_norm": 1.5942395901890765,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8734520673751831,
      "num_tokens": 373047574.0,
      "step": 10721
    },
    {
      "epoch": 1.9910863509749304,
      "grad_norm": 1.5360263467989892,
      "learning_rate": 1e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8642581701278687,
      "num_tokens": 373079582.0,
      "step": 10722
    },
    {
      "epoch": 1.9912720519962859,
      "grad_norm": 1.450560123084834,
      "learning_rate": 1e-06,
      "loss": 0.2859,
      "mean_token_accuracy": 0.8980594873428345,
      "num_tokens": 373110796.0,
      "step": 10723
    },
    {
      "epoch": 1.9914577530176416,
      "grad_norm": 1.4993750818715963,
      "learning_rate": 1e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8681946992874146,
      "num_tokens": 373148124.0,
      "step": 10724
    },
    {
      "epoch": 1.9916434540389973,
      "grad_norm": 1.5076630354975395,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8816580772399902,
      "num_tokens": 373183410.0,
      "step": 10725
    },
    {
      "epoch": 1.9918291550603529,
      "grad_norm": 1.4854635386657826,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8847107887268066,
      "num_tokens": 373215899.0,
      "step": 10726
    },
    {
      "epoch": 1.9920148560817084,
      "grad_norm": 1.5942348229415142,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8894832134246826,
      "num_tokens": 373242692.0,
      "step": 10727
    },
    {
      "epoch": 1.992200557103064,
      "grad_norm": 1.692694359808738,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8786718845367432,
      "num_tokens": 373273228.0,
      "step": 10728
    },
    {
      "epoch": 1.9923862581244198,
      "grad_norm": 1.3336083453687506,
      "learning_rate": 1e-06,
      "loss": 0.2845,
      "mean_token_accuracy": 0.8990179300308228,
      "num_tokens": 373311250.0,
      "step": 10729
    },
    {
      "epoch": 1.9925719591457753,
      "grad_norm": 1.3844488247795683,
      "learning_rate": 1e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8732333779335022,
      "num_tokens": 373354271.0,
      "step": 10730
    },
    {
      "epoch": 1.9927576601671309,
      "grad_norm": 1.559543210477297,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.8815058469772339,
      "num_tokens": 373384681.0,
      "step": 10731
    },
    {
      "epoch": 1.9929433611884866,
      "grad_norm": 1.5140597341137003,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8861403465270996,
      "num_tokens": 373415272.0,
      "step": 10732
    },
    {
      "epoch": 1.9931290622098423,
      "grad_norm": 1.34649821678119,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.876867413520813,
      "num_tokens": 373456542.0,
      "step": 10733
    },
    {
      "epoch": 1.9933147632311976,
      "grad_norm": 1.5207830057025506,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.861910343170166,
      "num_tokens": 373492996.0,
      "step": 10734
    },
    {
      "epoch": 1.9935004642525533,
      "grad_norm": 1.43624861928419,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.8812456130981445,
      "num_tokens": 373529154.0,
      "step": 10735
    },
    {
      "epoch": 1.993686165273909,
      "grad_norm": 1.5989389075226108,
      "learning_rate": 1e-06,
      "loss": 0.3883,
      "mean_token_accuracy": 0.8647154569625854,
      "num_tokens": 373562115.0,
      "step": 10736
    },
    {
      "epoch": 1.9938718662952646,
      "grad_norm": 1.6866594767200314,
      "learning_rate": 1e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8667343854904175,
      "num_tokens": 373593092.0,
      "step": 10737
    },
    {
      "epoch": 1.99405756731662,
      "grad_norm": 1.4945492144949333,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8785459995269775,
      "num_tokens": 373625357.0,
      "step": 10738
    },
    {
      "epoch": 1.9942432683379758,
      "grad_norm": 1.5238549053567785,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8721871376037598,
      "num_tokens": 373659809.0,
      "step": 10739
    },
    {
      "epoch": 1.9944289693593316,
      "grad_norm": 2.214133595010913,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8706782460212708,
      "num_tokens": 373697127.0,
      "step": 10740
    },
    {
      "epoch": 1.994614670380687,
      "grad_norm": 1.394865069545873,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8865814208984375,
      "num_tokens": 373729944.0,
      "step": 10741
    },
    {
      "epoch": 1.9948003714020426,
      "grad_norm": 1.4470068415853325,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8797589540481567,
      "num_tokens": 373768846.0,
      "step": 10742
    },
    {
      "epoch": 1.9949860724233983,
      "grad_norm": 1.3827251364655504,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8810155391693115,
      "num_tokens": 373805570.0,
      "step": 10743
    },
    {
      "epoch": 1.995171773444754,
      "grad_norm": 1.5047070973037104,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8688695430755615,
      "num_tokens": 373842065.0,
      "step": 10744
    },
    {
      "epoch": 1.9953574744661096,
      "grad_norm": 1.4177165145902513,
      "learning_rate": 1e-06,
      "loss": 0.3748,
      "mean_token_accuracy": 0.8764312267303467,
      "num_tokens": 373879577.0,
      "step": 10745
    },
    {
      "epoch": 1.995543175487465,
      "grad_norm": 1.4995381178750953,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8780222535133362,
      "num_tokens": 373918404.0,
      "step": 10746
    },
    {
      "epoch": 1.9957288765088208,
      "grad_norm": 1.46098473660683,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8775734901428223,
      "num_tokens": 373951400.0,
      "step": 10747
    },
    {
      "epoch": 1.9959145775301765,
      "grad_norm": 1.399029553250972,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.893821656703949,
      "num_tokens": 373984937.0,
      "step": 10748
    },
    {
      "epoch": 1.996100278551532,
      "grad_norm": 1.312876866779252,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8895415663719177,
      "num_tokens": 374024479.0,
      "step": 10749
    },
    {
      "epoch": 1.9962859795728876,
      "grad_norm": 1.3298056191877696,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.876915693283081,
      "num_tokens": 374067550.0,
      "step": 10750
    },
    {
      "epoch": 1.9964716805942433,
      "grad_norm": 1.5382922592755202,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8784908056259155,
      "num_tokens": 374101518.0,
      "step": 10751
    },
    {
      "epoch": 1.996657381615599,
      "grad_norm": 1.484176722993504,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8869295120239258,
      "num_tokens": 374136038.0,
      "step": 10752
    },
    {
      "epoch": 1.9968430826369545,
      "grad_norm": 1.52855785145475,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8777517080307007,
      "num_tokens": 374173834.0,
      "step": 10753
    },
    {
      "epoch": 1.99702878365831,
      "grad_norm": 1.8321453987730727,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8723989725112915,
      "num_tokens": 374197551.0,
      "step": 10754
    },
    {
      "epoch": 1.9972144846796658,
      "grad_norm": 1.4950153372267951,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8860377073287964,
      "num_tokens": 374227931.0,
      "step": 10755
    },
    {
      "epoch": 1.9974001857010215,
      "grad_norm": 1.5321224660023414,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8870094418525696,
      "num_tokens": 374259723.0,
      "step": 10756
    },
    {
      "epoch": 1.997585886722377,
      "grad_norm": 1.4845132473191165,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8900861740112305,
      "num_tokens": 374289790.0,
      "step": 10757
    },
    {
      "epoch": 1.9977715877437325,
      "grad_norm": 1.4088888261451207,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8838513493537903,
      "num_tokens": 374325750.0,
      "step": 10758
    },
    {
      "epoch": 1.9979572887650883,
      "grad_norm": 1.5956089398797122,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8778262138366699,
      "num_tokens": 374354749.0,
      "step": 10759
    },
    {
      "epoch": 1.9981429897864438,
      "grad_norm": 1.5262048765288998,
      "learning_rate": 1e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8616526126861572,
      "num_tokens": 374391379.0,
      "step": 10760
    },
    {
      "epoch": 1.9983286908077993,
      "grad_norm": 1.5889665786517286,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8698511719703674,
      "num_tokens": 374424459.0,
      "step": 10761
    },
    {
      "epoch": 1.998514391829155,
      "grad_norm": 1.4206388103907164,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8705437779426575,
      "num_tokens": 374462958.0,
      "step": 10762
    },
    {
      "epoch": 1.9987000928505108,
      "grad_norm": 1.519684696138038,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8781314492225647,
      "num_tokens": 374495409.0,
      "step": 10763
    },
    {
      "epoch": 1.9988857938718663,
      "grad_norm": 1.3716688805299844,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8799022436141968,
      "num_tokens": 374533628.0,
      "step": 10764
    },
    {
      "epoch": 1.9990714948932218,
      "grad_norm": 1.3826601323744552,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8778570890426636,
      "num_tokens": 374570878.0,
      "step": 10765
    },
    {
      "epoch": 1.9992571959145775,
      "grad_norm": 1.4129675174377778,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.879777193069458,
      "num_tokens": 374611088.0,
      "step": 10766
    },
    {
      "epoch": 1.9994428969359332,
      "grad_norm": 1.4194012311049073,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.8679375052452087,
      "num_tokens": 374653398.0,
      "step": 10767
    },
    {
      "epoch": 1.9996285979572888,
      "grad_norm": 1.45621301237432,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8700888156890869,
      "num_tokens": 374690451.0,
      "step": 10768
    },
    {
      "epoch": 1.9998142989786443,
      "grad_norm": 1.4780013422102942,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8861258029937744,
      "num_tokens": 374724245.0,
      "step": 10769
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4704585167465687,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8718933463096619,
      "num_tokens": 374760835.0,
      "step": 10770
    },
    {
      "epoch": 2.0001857010213557,
      "grad_norm": 1.5002881606145277,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.889019250869751,
      "num_tokens": 374798838.0,
      "step": 10771
    },
    {
      "epoch": 2.000371402042711,
      "grad_norm": 1.4183284421329476,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8927483558654785,
      "num_tokens": 374835138.0,
      "step": 10772
    },
    {
      "epoch": 2.0005571030640668,
      "grad_norm": 1.4015800294678256,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8913413286209106,
      "num_tokens": 374872702.0,
      "step": 10773
    },
    {
      "epoch": 2.0007428040854225,
      "grad_norm": 1.5910006331317907,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8794724941253662,
      "num_tokens": 374901963.0,
      "step": 10774
    },
    {
      "epoch": 2.000928505106778,
      "grad_norm": 1.4356316802861704,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8826208114624023,
      "num_tokens": 374939640.0,
      "step": 10775
    },
    {
      "epoch": 2.0011142061281335,
      "grad_norm": 1.3442023534204168,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8916493654251099,
      "num_tokens": 374978901.0,
      "step": 10776
    },
    {
      "epoch": 2.0012999071494892,
      "grad_norm": 1.4214112619644421,
      "learning_rate": 1e-06,
      "loss": 0.2795,
      "mean_token_accuracy": 0.899822473526001,
      "num_tokens": 375014498.0,
      "step": 10777
    },
    {
      "epoch": 2.001485608170845,
      "grad_norm": 1.450041013346496,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.893798828125,
      "num_tokens": 375048253.0,
      "step": 10778
    },
    {
      "epoch": 2.0016713091922007,
      "grad_norm": 1.627189879187797,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8749921321868896,
      "num_tokens": 375083780.0,
      "step": 10779
    },
    {
      "epoch": 2.001857010213556,
      "grad_norm": 1.5884218597808155,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8857653141021729,
      "num_tokens": 375117532.0,
      "step": 10780
    },
    {
      "epoch": 2.0020427112349117,
      "grad_norm": 1.5473841088798492,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8866158723831177,
      "num_tokens": 375151231.0,
      "step": 10781
    },
    {
      "epoch": 2.0022284122562675,
      "grad_norm": 1.5731439137923022,
      "learning_rate": 1e-06,
      "loss": 0.2731,
      "mean_token_accuracy": 0.9002057313919067,
      "num_tokens": 375181827.0,
      "step": 10782
    },
    {
      "epoch": 2.002414113277623,
      "grad_norm": 1.7141740061781399,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8896212577819824,
      "num_tokens": 375210107.0,
      "step": 10783
    },
    {
      "epoch": 2.0025998142989785,
      "grad_norm": 1.6182982928852798,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8914583325386047,
      "num_tokens": 375241712.0,
      "step": 10784
    },
    {
      "epoch": 2.002785515320334,
      "grad_norm": 1.5829096087469963,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.885259211063385,
      "num_tokens": 375279747.0,
      "step": 10785
    },
    {
      "epoch": 2.00297121634169,
      "grad_norm": 1.6053705100391122,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8829101920127869,
      "num_tokens": 375312775.0,
      "step": 10786
    },
    {
      "epoch": 2.0031569173630457,
      "grad_norm": 1.5036301712232554,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.8900225758552551,
      "num_tokens": 375347469.0,
      "step": 10787
    },
    {
      "epoch": 2.003342618384401,
      "grad_norm": 1.6247418523160633,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8842196464538574,
      "num_tokens": 375377937.0,
      "step": 10788
    },
    {
      "epoch": 2.0035283194057567,
      "grad_norm": 1.3110311928541807,
      "learning_rate": 1e-06,
      "loss": 0.3083,
      "mean_token_accuracy": 0.8914475440979004,
      "num_tokens": 375421470.0,
      "step": 10789
    },
    {
      "epoch": 2.0037140204271124,
      "grad_norm": 1.5435867437603985,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8818960189819336,
      "num_tokens": 375460621.0,
      "step": 10790
    },
    {
      "epoch": 2.003899721448468,
      "grad_norm": 1.5247912814439823,
      "learning_rate": 1e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8666360378265381,
      "num_tokens": 375499799.0,
      "step": 10791
    },
    {
      "epoch": 2.0040854224698235,
      "grad_norm": 1.334903103642262,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8969404101371765,
      "num_tokens": 375538923.0,
      "step": 10792
    },
    {
      "epoch": 2.004271123491179,
      "grad_norm": 1.5535212009568864,
      "learning_rate": 1e-06,
      "loss": 0.2775,
      "mean_token_accuracy": 0.9024122357368469,
      "num_tokens": 375567869.0,
      "step": 10793
    },
    {
      "epoch": 2.004456824512535,
      "grad_norm": 1.4483449376352828,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8843215703964233,
      "num_tokens": 375608277.0,
      "step": 10794
    },
    {
      "epoch": 2.00464252553389,
      "grad_norm": 1.5161565235708332,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8884828686714172,
      "num_tokens": 375640731.0,
      "step": 10795
    },
    {
      "epoch": 2.004828226555246,
      "grad_norm": 1.4877575761667934,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8839312195777893,
      "num_tokens": 375676379.0,
      "step": 10796
    },
    {
      "epoch": 2.0050139275766017,
      "grad_norm": 1.5842594242575883,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8872473835945129,
      "num_tokens": 375710645.0,
      "step": 10797
    },
    {
      "epoch": 2.0051996285979574,
      "grad_norm": 1.589269465616174,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8806769847869873,
      "num_tokens": 375744006.0,
      "step": 10798
    },
    {
      "epoch": 2.0053853296193127,
      "grad_norm": 1.4398792651935468,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.893104076385498,
      "num_tokens": 375776989.0,
      "step": 10799
    },
    {
      "epoch": 2.0055710306406684,
      "grad_norm": 1.4482445500143069,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8808850049972534,
      "num_tokens": 375816685.0,
      "step": 10800
    },
    {
      "epoch": 2.005756731662024,
      "grad_norm": 1.4600461344807414,
      "learning_rate": 1e-06,
      "loss": 0.2766,
      "mean_token_accuracy": 0.9021281003952026,
      "num_tokens": 375852322.0,
      "step": 10801
    },
    {
      "epoch": 2.00594243268338,
      "grad_norm": 1.649767791832117,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8766404390335083,
      "num_tokens": 375884237.0,
      "step": 10802
    },
    {
      "epoch": 2.006128133704735,
      "grad_norm": 1.5299328202270464,
      "learning_rate": 1e-06,
      "loss": 0.288,
      "mean_token_accuracy": 0.8972945213317871,
      "num_tokens": 375916920.0,
      "step": 10803
    },
    {
      "epoch": 2.006313834726091,
      "grad_norm": 1.4648046893912137,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8841018080711365,
      "num_tokens": 375956440.0,
      "step": 10804
    },
    {
      "epoch": 2.0064995357474467,
      "grad_norm": 1.526733560718419,
      "learning_rate": 1e-06,
      "loss": 0.3593,
      "mean_token_accuracy": 0.8768559098243713,
      "num_tokens": 375994565.0,
      "step": 10805
    },
    {
      "epoch": 2.0066852367688024,
      "grad_norm": 1.3621606756558498,
      "learning_rate": 1e-06,
      "loss": 0.2725,
      "mean_token_accuracy": 0.9023377895355225,
      "num_tokens": 376034281.0,
      "step": 10806
    },
    {
      "epoch": 2.0068709377901577,
      "grad_norm": 1.421719356874049,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8822875022888184,
      "num_tokens": 376074600.0,
      "step": 10807
    },
    {
      "epoch": 2.0070566388115134,
      "grad_norm": 1.5345952356372081,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8771042227745056,
      "num_tokens": 376106940.0,
      "step": 10808
    },
    {
      "epoch": 2.007242339832869,
      "grad_norm": 1.4849856268144557,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8878651857376099,
      "num_tokens": 376142749.0,
      "step": 10809
    },
    {
      "epoch": 2.007428040854225,
      "grad_norm": 1.4970284723279201,
      "learning_rate": 1e-06,
      "loss": 0.2818,
      "mean_token_accuracy": 0.9015620350837708,
      "num_tokens": 376180156.0,
      "step": 10810
    },
    {
      "epoch": 2.00761374187558,
      "grad_norm": 1.476470338426744,
      "learning_rate": 1e-06,
      "loss": 0.3005,
      "mean_token_accuracy": 0.8950074315071106,
      "num_tokens": 376217685.0,
      "step": 10811
    },
    {
      "epoch": 2.007799442896936,
      "grad_norm": 1.483959874744587,
      "learning_rate": 1e-06,
      "loss": 0.3335,
      "mean_token_accuracy": 0.8837865591049194,
      "num_tokens": 376258726.0,
      "step": 10812
    },
    {
      "epoch": 2.0079851439182916,
      "grad_norm": 1.642848634672837,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8898269534111023,
      "num_tokens": 376291548.0,
      "step": 10813
    },
    {
      "epoch": 2.0081708449396474,
      "grad_norm": 1.4745932089133484,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8908197283744812,
      "num_tokens": 376328037.0,
      "step": 10814
    },
    {
      "epoch": 2.0083565459610027,
      "grad_norm": 1.731554708633873,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8864076733589172,
      "num_tokens": 376356514.0,
      "step": 10815
    },
    {
      "epoch": 2.0085422469823584,
      "grad_norm": 1.567340268866878,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8907877802848816,
      "num_tokens": 376390044.0,
      "step": 10816
    },
    {
      "epoch": 2.008727948003714,
      "grad_norm": 1.535012938279411,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8876144289970398,
      "num_tokens": 376426868.0,
      "step": 10817
    },
    {
      "epoch": 2.00891364902507,
      "grad_norm": 1.5922508981914947,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8893662691116333,
      "num_tokens": 376463552.0,
      "step": 10818
    },
    {
      "epoch": 2.009099350046425,
      "grad_norm": 1.8788669141409904,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8779202103614807,
      "num_tokens": 376489498.0,
      "step": 10819
    },
    {
      "epoch": 2.009285051067781,
      "grad_norm": 1.7807357245498063,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8690678477287292,
      "num_tokens": 376517976.0,
      "step": 10820
    },
    {
      "epoch": 2.0094707520891366,
      "grad_norm": 1.5269001624379197,
      "learning_rate": 1e-06,
      "loss": 0.295,
      "mean_token_accuracy": 0.8948981165885925,
      "num_tokens": 376554267.0,
      "step": 10821
    },
    {
      "epoch": 2.009656453110492,
      "grad_norm": 1.8081215593318916,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8806848526000977,
      "num_tokens": 376579483.0,
      "step": 10822
    },
    {
      "epoch": 2.0098421541318476,
      "grad_norm": 1.5313175939079444,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8782044053077698,
      "num_tokens": 376616420.0,
      "step": 10823
    },
    {
      "epoch": 2.0100278551532034,
      "grad_norm": 1.4873489203010748,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.8963729739189148,
      "num_tokens": 376654342.0,
      "step": 10824
    },
    {
      "epoch": 2.010213556174559,
      "grad_norm": 1.6488990194385822,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8839328289031982,
      "num_tokens": 376685459.0,
      "step": 10825
    },
    {
      "epoch": 2.0103992571959144,
      "grad_norm": 1.5257731760055657,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.8987709283828735,
      "num_tokens": 376720078.0,
      "step": 10826
    },
    {
      "epoch": 2.01058495821727,
      "grad_norm": 1.5551557613649738,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8849027752876282,
      "num_tokens": 376754839.0,
      "step": 10827
    },
    {
      "epoch": 2.010770659238626,
      "grad_norm": 1.6583902122353347,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8705172538757324,
      "num_tokens": 376786231.0,
      "step": 10828
    },
    {
      "epoch": 2.0109563602599816,
      "grad_norm": 1.6678450052440865,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8885247707366943,
      "num_tokens": 376816659.0,
      "step": 10829
    },
    {
      "epoch": 2.011142061281337,
      "grad_norm": 1.4667599570381196,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8935015201568604,
      "num_tokens": 376854142.0,
      "step": 10830
    },
    {
      "epoch": 2.0113277623026926,
      "grad_norm": 1.6118739777465683,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8960772752761841,
      "num_tokens": 376891492.0,
      "step": 10831
    },
    {
      "epoch": 2.0115134633240483,
      "grad_norm": 1.4069558386354317,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8903579115867615,
      "num_tokens": 376931115.0,
      "step": 10832
    },
    {
      "epoch": 2.011699164345404,
      "grad_norm": 1.4731418014354045,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8833228349685669,
      "num_tokens": 376972906.0,
      "step": 10833
    },
    {
      "epoch": 2.0118848653667594,
      "grad_norm": 1.6159609063178988,
      "learning_rate": 1e-06,
      "loss": 0.3098,
      "mean_token_accuracy": 0.8924041986465454,
      "num_tokens": 377005627.0,
      "step": 10834
    },
    {
      "epoch": 2.012070566388115,
      "grad_norm": 1.5605748299130129,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8877521753311157,
      "num_tokens": 377042135.0,
      "step": 10835
    },
    {
      "epoch": 2.012256267409471,
      "grad_norm": 1.4378147447516842,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8921290636062622,
      "num_tokens": 377081294.0,
      "step": 10836
    },
    {
      "epoch": 2.0124419684308266,
      "grad_norm": 1.6263815924569343,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8874219655990601,
      "num_tokens": 377113985.0,
      "step": 10837
    },
    {
      "epoch": 2.012627669452182,
      "grad_norm": 1.5717167983161058,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8723910450935364,
      "num_tokens": 377152281.0,
      "step": 10838
    },
    {
      "epoch": 2.0128133704735376,
      "grad_norm": 1.5917153196946814,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8854900598526001,
      "num_tokens": 377182385.0,
      "step": 10839
    },
    {
      "epoch": 2.0129990714948933,
      "grad_norm": 1.4352005856434784,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.884022057056427,
      "num_tokens": 377224378.0,
      "step": 10840
    },
    {
      "epoch": 2.013184772516249,
      "grad_norm": 1.5384240407697711,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8799146413803101,
      "num_tokens": 377261593.0,
      "step": 10841
    },
    {
      "epoch": 2.0133704735376043,
      "grad_norm": 1.563535097514731,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8882608413696289,
      "num_tokens": 377294032.0,
      "step": 10842
    },
    {
      "epoch": 2.01355617455896,
      "grad_norm": 1.4824664912088268,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8879417181015015,
      "num_tokens": 377329800.0,
      "step": 10843
    },
    {
      "epoch": 2.013741875580316,
      "grad_norm": 1.5950109869413045,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8913845419883728,
      "num_tokens": 377360149.0,
      "step": 10844
    },
    {
      "epoch": 2.013927576601671,
      "grad_norm": 1.4955860370138607,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8853365182876587,
      "num_tokens": 377396526.0,
      "step": 10845
    },
    {
      "epoch": 2.014113277623027,
      "grad_norm": 1.5951748463682331,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8901129364967346,
      "num_tokens": 377429678.0,
      "step": 10846
    },
    {
      "epoch": 2.0142989786443826,
      "grad_norm": 1.5120844518742864,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8807404041290283,
      "num_tokens": 377469104.0,
      "step": 10847
    },
    {
      "epoch": 2.0144846796657383,
      "grad_norm": 1.5547798477789854,
      "learning_rate": 1e-06,
      "loss": 0.2856,
      "mean_token_accuracy": 0.899750292301178,
      "num_tokens": 377499595.0,
      "step": 10848
    },
    {
      "epoch": 2.0146703806870936,
      "grad_norm": 1.5412341238336658,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8842354416847229,
      "num_tokens": 377536892.0,
      "step": 10849
    },
    {
      "epoch": 2.0148560817084493,
      "grad_norm": 1.6312527737146214,
      "learning_rate": 1e-06,
      "loss": 0.2933,
      "mean_token_accuracy": 0.895675539970398,
      "num_tokens": 377574970.0,
      "step": 10850
    },
    {
      "epoch": 2.015041782729805,
      "grad_norm": 1.5264675006878032,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8861310482025146,
      "num_tokens": 377607872.0,
      "step": 10851
    },
    {
      "epoch": 2.015227483751161,
      "grad_norm": 1.7478245977121931,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8719281554222107,
      "num_tokens": 377637318.0,
      "step": 10852
    },
    {
      "epoch": 2.015413184772516,
      "grad_norm": 1.6948692561710206,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8791581392288208,
      "num_tokens": 377668640.0,
      "step": 10853
    },
    {
      "epoch": 2.015598885793872,
      "grad_norm": 1.5013636281988112,
      "learning_rate": 1e-06,
      "loss": 0.287,
      "mean_token_accuracy": 0.8971197605133057,
      "num_tokens": 377704731.0,
      "step": 10854
    },
    {
      "epoch": 2.0157845868152275,
      "grad_norm": 1.5805385239730814,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8948922157287598,
      "num_tokens": 377735910.0,
      "step": 10855
    },
    {
      "epoch": 2.0159702878365833,
      "grad_norm": 1.5789975889219443,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8836979269981384,
      "num_tokens": 377772926.0,
      "step": 10856
    },
    {
      "epoch": 2.0161559888579386,
      "grad_norm": 1.486753406388056,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8881658911705017,
      "num_tokens": 377812017.0,
      "step": 10857
    },
    {
      "epoch": 2.0163416898792943,
      "grad_norm": 1.638002165934511,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8769640922546387,
      "num_tokens": 377846076.0,
      "step": 10858
    },
    {
      "epoch": 2.01652739090065,
      "grad_norm": 1.3587972144448845,
      "learning_rate": 1e-06,
      "loss": 0.2794,
      "mean_token_accuracy": 0.903499960899353,
      "num_tokens": 377882718.0,
      "step": 10859
    },
    {
      "epoch": 2.0167130919220058,
      "grad_norm": 1.4956006255420118,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8876153826713562,
      "num_tokens": 377920113.0,
      "step": 10860
    },
    {
      "epoch": 2.016898792943361,
      "grad_norm": 1.6863029274539099,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8907970190048218,
      "num_tokens": 377950475.0,
      "step": 10861
    },
    {
      "epoch": 2.017084493964717,
      "grad_norm": 1.6224260098697185,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8780008554458618,
      "num_tokens": 377986436.0,
      "step": 10862
    },
    {
      "epoch": 2.0172701949860725,
      "grad_norm": 1.6174365657534113,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8806742429733276,
      "num_tokens": 378020255.0,
      "step": 10863
    },
    {
      "epoch": 2.0174558960074283,
      "grad_norm": 1.7860532522364216,
      "learning_rate": 1e-06,
      "loss": 0.3969,
      "mean_token_accuracy": 0.8662086725234985,
      "num_tokens": 378053585.0,
      "step": 10864
    },
    {
      "epoch": 2.0176415970287835,
      "grad_norm": 1.5481723142762016,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8815835118293762,
      "num_tokens": 378088958.0,
      "step": 10865
    },
    {
      "epoch": 2.0178272980501393,
      "grad_norm": 1.466784467078266,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8886759281158447,
      "num_tokens": 378128632.0,
      "step": 10866
    },
    {
      "epoch": 2.018012999071495,
      "grad_norm": 1.4947370396192174,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8903199434280396,
      "num_tokens": 378166153.0,
      "step": 10867
    },
    {
      "epoch": 2.0181987000928503,
      "grad_norm": 1.725418967503799,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8740948438644409,
      "num_tokens": 378196057.0,
      "step": 10868
    },
    {
      "epoch": 2.018384401114206,
      "grad_norm": 1.5869081825246187,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8800381422042847,
      "num_tokens": 378230950.0,
      "step": 10869
    },
    {
      "epoch": 2.0185701021355618,
      "grad_norm": 1.4868875483677213,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8845639228820801,
      "num_tokens": 378269426.0,
      "step": 10870
    },
    {
      "epoch": 2.0187558031569175,
      "grad_norm": 1.5327068692763155,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8809878826141357,
      "num_tokens": 378308121.0,
      "step": 10871
    },
    {
      "epoch": 2.018941504178273,
      "grad_norm": 1.528851051458869,
      "learning_rate": 1e-06,
      "loss": 0.2792,
      "mean_token_accuracy": 0.899274468421936,
      "num_tokens": 378339713.0,
      "step": 10872
    },
    {
      "epoch": 2.0191272051996285,
      "grad_norm": 1.5037541133055323,
      "learning_rate": 1e-06,
      "loss": 0.3776,
      "mean_token_accuracy": 0.8716254234313965,
      "num_tokens": 378378363.0,
      "step": 10873
    },
    {
      "epoch": 2.0193129062209842,
      "grad_norm": 1.6238598916703724,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.879149854183197,
      "num_tokens": 378413224.0,
      "step": 10874
    },
    {
      "epoch": 2.01949860724234,
      "grad_norm": 1.7027990327800355,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8910763263702393,
      "num_tokens": 378448554.0,
      "step": 10875
    },
    {
      "epoch": 2.0196843082636953,
      "grad_norm": 1.6077717576846606,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8809573650360107,
      "num_tokens": 378481699.0,
      "step": 10876
    },
    {
      "epoch": 2.019870009285051,
      "grad_norm": 1.5749032356818924,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.881382405757904,
      "num_tokens": 378516988.0,
      "step": 10877
    },
    {
      "epoch": 2.0200557103064067,
      "grad_norm": 1.554362083247208,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8859802484512329,
      "num_tokens": 378548842.0,
      "step": 10878
    },
    {
      "epoch": 2.0202414113277625,
      "grad_norm": 1.506950380898217,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.894689679145813,
      "num_tokens": 378582497.0,
      "step": 10879
    },
    {
      "epoch": 2.0204271123491178,
      "grad_norm": 1.7112735292350905,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8896859884262085,
      "num_tokens": 378613792.0,
      "step": 10880
    },
    {
      "epoch": 2.0206128133704735,
      "grad_norm": 1.5615521332158073,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8835767507553101,
      "num_tokens": 378647933.0,
      "step": 10881
    },
    {
      "epoch": 2.0207985143918292,
      "grad_norm": 1.5681464582993212,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8914421796798706,
      "num_tokens": 378680372.0,
      "step": 10882
    },
    {
      "epoch": 2.020984215413185,
      "grad_norm": 1.5397720878076073,
      "learning_rate": 1e-06,
      "loss": 0.2909,
      "mean_token_accuracy": 0.8971199989318848,
      "num_tokens": 378713084.0,
      "step": 10883
    },
    {
      "epoch": 2.0211699164345402,
      "grad_norm": 1.680123382354699,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8750479221343994,
      "num_tokens": 378742889.0,
      "step": 10884
    },
    {
      "epoch": 2.021355617455896,
      "grad_norm": 1.6916854598849265,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8881151676177979,
      "num_tokens": 378773126.0,
      "step": 10885
    },
    {
      "epoch": 2.0215413184772517,
      "grad_norm": 1.512007601984,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8813470602035522,
      "num_tokens": 378815409.0,
      "step": 10886
    },
    {
      "epoch": 2.0217270194986074,
      "grad_norm": 1.5935762209386684,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.891079306602478,
      "num_tokens": 378849716.0,
      "step": 10887
    },
    {
      "epoch": 2.0219127205199627,
      "grad_norm": 1.5712940229668357,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.873441219329834,
      "num_tokens": 378888773.0,
      "step": 10888
    },
    {
      "epoch": 2.0220984215413185,
      "grad_norm": 1.7531590525540168,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8923436403274536,
      "num_tokens": 378917635.0,
      "step": 10889
    },
    {
      "epoch": 2.022284122562674,
      "grad_norm": 1.401106043566016,
      "learning_rate": 1e-06,
      "loss": 0.2904,
      "mean_token_accuracy": 0.8975818157196045,
      "num_tokens": 378956598.0,
      "step": 10890
    },
    {
      "epoch": 2.0224698235840295,
      "grad_norm": 1.700571382912729,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8837288618087769,
      "num_tokens": 378986831.0,
      "step": 10891
    },
    {
      "epoch": 2.0226555246053852,
      "grad_norm": 1.546639950892486,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8994243144989014,
      "num_tokens": 379020304.0,
      "step": 10892
    },
    {
      "epoch": 2.022841225626741,
      "grad_norm": 1.4901655208509088,
      "learning_rate": 1e-06,
      "loss": 0.2983,
      "mean_token_accuracy": 0.8916677236557007,
      "num_tokens": 379055034.0,
      "step": 10893
    },
    {
      "epoch": 2.0230269266480967,
      "grad_norm": 1.5681904107628535,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8760523796081543,
      "num_tokens": 379091262.0,
      "step": 10894
    },
    {
      "epoch": 2.023212627669452,
      "grad_norm": 1.5317845671773118,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8910255432128906,
      "num_tokens": 379128632.0,
      "step": 10895
    },
    {
      "epoch": 2.0233983286908077,
      "grad_norm": 1.6252972853499168,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8836488723754883,
      "num_tokens": 379162860.0,
      "step": 10896
    },
    {
      "epoch": 2.0235840297121634,
      "grad_norm": 1.6992548780332308,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8930096626281738,
      "num_tokens": 379194706.0,
      "step": 10897
    },
    {
      "epoch": 2.023769730733519,
      "grad_norm": 1.789920650512693,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8731511235237122,
      "num_tokens": 379224858.0,
      "step": 10898
    },
    {
      "epoch": 2.0239554317548745,
      "grad_norm": 1.632744886614017,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8863461017608643,
      "num_tokens": 379259464.0,
      "step": 10899
    },
    {
      "epoch": 2.02414113277623,
      "grad_norm": 1.8075626173068713,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8914783000946045,
      "num_tokens": 379290086.0,
      "step": 10900
    },
    {
      "epoch": 2.024326833797586,
      "grad_norm": 1.523036426999459,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8889258503913879,
      "num_tokens": 379326053.0,
      "step": 10901
    },
    {
      "epoch": 2.0245125348189417,
      "grad_norm": 1.6755683295189256,
      "learning_rate": 1e-06,
      "loss": 0.2869,
      "mean_token_accuracy": 0.8983795046806335,
      "num_tokens": 379352615.0,
      "step": 10902
    },
    {
      "epoch": 2.024698235840297,
      "grad_norm": 1.670400111168298,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8848322629928589,
      "num_tokens": 379385463.0,
      "step": 10903
    },
    {
      "epoch": 2.0248839368616527,
      "grad_norm": 1.5277659259031258,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8862450122833252,
      "num_tokens": 379420271.0,
      "step": 10904
    },
    {
      "epoch": 2.0250696378830084,
      "grad_norm": 1.6150010521775824,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8926947116851807,
      "num_tokens": 379451517.0,
      "step": 10905
    },
    {
      "epoch": 2.025255338904364,
      "grad_norm": 2.252449209324364,
      "learning_rate": 1e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8635460138320923,
      "num_tokens": 379486392.0,
      "step": 10906
    },
    {
      "epoch": 2.0254410399257194,
      "grad_norm": 1.512334460028367,
      "learning_rate": 1e-06,
      "loss": 0.3798,
      "mean_token_accuracy": 0.8713946342468262,
      "num_tokens": 379527833.0,
      "step": 10907
    },
    {
      "epoch": 2.025626740947075,
      "grad_norm": 1.4111864740523556,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8854244947433472,
      "num_tokens": 379569666.0,
      "step": 10908
    },
    {
      "epoch": 2.025812441968431,
      "grad_norm": 1.5966929174594822,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8895581364631653,
      "num_tokens": 379599849.0,
      "step": 10909
    },
    {
      "epoch": 2.0259981429897866,
      "grad_norm": 1.773704519744588,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8872520923614502,
      "num_tokens": 379626022.0,
      "step": 10910
    },
    {
      "epoch": 2.026183844011142,
      "grad_norm": 1.472246142241892,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8847256302833557,
      "num_tokens": 379665548.0,
      "step": 10911
    },
    {
      "epoch": 2.0263695450324977,
      "grad_norm": 1.5130139813672325,
      "learning_rate": 1e-06,
      "loss": 0.2757,
      "mean_token_accuracy": 0.901486873626709,
      "num_tokens": 379700434.0,
      "step": 10912
    },
    {
      "epoch": 2.0265552460538534,
      "grad_norm": 1.4538270151785955,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8923064470291138,
      "num_tokens": 379738568.0,
      "step": 10913
    },
    {
      "epoch": 2.026740947075209,
      "grad_norm": 1.6386696031698504,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8811304569244385,
      "num_tokens": 379770881.0,
      "step": 10914
    },
    {
      "epoch": 2.0269266480965644,
      "grad_norm": 1.5017629251169702,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8784709572792053,
      "num_tokens": 379807983.0,
      "step": 10915
    },
    {
      "epoch": 2.02711234911792,
      "grad_norm": 1.607202087968128,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8894404768943787,
      "num_tokens": 379836993.0,
      "step": 10916
    },
    {
      "epoch": 2.027298050139276,
      "grad_norm": 1.5221395708551155,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.884876012802124,
      "num_tokens": 379871642.0,
      "step": 10917
    },
    {
      "epoch": 2.027483751160631,
      "grad_norm": 1.6671248983205946,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8806787133216858,
      "num_tokens": 379904461.0,
      "step": 10918
    },
    {
      "epoch": 2.027669452181987,
      "grad_norm": 1.5475296008678279,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.886934220790863,
      "num_tokens": 379942013.0,
      "step": 10919
    },
    {
      "epoch": 2.0278551532033426,
      "grad_norm": 1.4956170427925504,
      "learning_rate": 1e-06,
      "loss": 0.281,
      "mean_token_accuracy": 0.8983436822891235,
      "num_tokens": 379975522.0,
      "step": 10920
    },
    {
      "epoch": 2.0280408542246984,
      "grad_norm": 1.42471431899278,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.892714262008667,
      "num_tokens": 380016021.0,
      "step": 10921
    },
    {
      "epoch": 2.0282265552460537,
      "grad_norm": 1.5973140423024181,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8808472156524658,
      "num_tokens": 380050164.0,
      "step": 10922
    },
    {
      "epoch": 2.0284122562674094,
      "grad_norm": 1.529424632194954,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8933244347572327,
      "num_tokens": 380084454.0,
      "step": 10923
    },
    {
      "epoch": 2.028597957288765,
      "grad_norm": 1.6537342525271552,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8781353235244751,
      "num_tokens": 380115558.0,
      "step": 10924
    },
    {
      "epoch": 2.028783658310121,
      "grad_norm": 1.5815768516185413,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8852846622467041,
      "num_tokens": 380152517.0,
      "step": 10925
    },
    {
      "epoch": 2.028969359331476,
      "grad_norm": 1.446742529810866,
      "learning_rate": 1e-06,
      "loss": 0.3002,
      "mean_token_accuracy": 0.8948000073432922,
      "num_tokens": 380188839.0,
      "step": 10926
    },
    {
      "epoch": 2.029155060352832,
      "grad_norm": 1.6424000481863466,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8694789409637451,
      "num_tokens": 380221349.0,
      "step": 10927
    },
    {
      "epoch": 2.0293407613741876,
      "grad_norm": 1.7751842898841137,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8838355541229248,
      "num_tokens": 380247654.0,
      "step": 10928
    },
    {
      "epoch": 2.0295264623955434,
      "grad_norm": 1.699486549729312,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8877414464950562,
      "num_tokens": 380281238.0,
      "step": 10929
    },
    {
      "epoch": 2.0297121634168986,
      "grad_norm": 1.4447805474990032,
      "learning_rate": 1e-06,
      "loss": 0.2693,
      "mean_token_accuracy": 0.9044034481048584,
      "num_tokens": 380318836.0,
      "step": 10930
    },
    {
      "epoch": 2.0298978644382544,
      "grad_norm": 1.5483380555911186,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8842551112174988,
      "num_tokens": 380356940.0,
      "step": 10931
    },
    {
      "epoch": 2.03008356545961,
      "grad_norm": 1.6988105426545446,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8723134398460388,
      "num_tokens": 380391293.0,
      "step": 10932
    },
    {
      "epoch": 2.030269266480966,
      "grad_norm": 1.600468090198205,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8836665153503418,
      "num_tokens": 380426463.0,
      "step": 10933
    },
    {
      "epoch": 2.030454967502321,
      "grad_norm": 1.588505446135696,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8777521848678589,
      "num_tokens": 380464112.0,
      "step": 10934
    },
    {
      "epoch": 2.030640668523677,
      "grad_norm": 1.4729363025798103,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.894801914691925,
      "num_tokens": 380503046.0,
      "step": 10935
    },
    {
      "epoch": 2.0308263695450326,
      "grad_norm": 1.5828331342440258,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.87349534034729,
      "num_tokens": 380540096.0,
      "step": 10936
    },
    {
      "epoch": 2.0310120705663883,
      "grad_norm": 1.5774891216128797,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8886359333992004,
      "num_tokens": 380575631.0,
      "step": 10937
    },
    {
      "epoch": 2.0311977715877436,
      "grad_norm": 1.559701225971073,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8895577192306519,
      "num_tokens": 380609501.0,
      "step": 10938
    },
    {
      "epoch": 2.0313834726090993,
      "grad_norm": 1.5719668226427264,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8840860724449158,
      "num_tokens": 380643622.0,
      "step": 10939
    },
    {
      "epoch": 2.031569173630455,
      "grad_norm": 1.5672703353478723,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8879756927490234,
      "num_tokens": 380677844.0,
      "step": 10940
    },
    {
      "epoch": 2.0317548746518104,
      "grad_norm": 1.390583482044406,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8964202404022217,
      "num_tokens": 380716413.0,
      "step": 10941
    },
    {
      "epoch": 2.031940575673166,
      "grad_norm": 1.5581716330776272,
      "learning_rate": 1e-06,
      "loss": 0.2876,
      "mean_token_accuracy": 0.8985871076583862,
      "num_tokens": 380754067.0,
      "step": 10942
    },
    {
      "epoch": 2.032126276694522,
      "grad_norm": 1.64262188685583,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8743650913238525,
      "num_tokens": 380785239.0,
      "step": 10943
    },
    {
      "epoch": 2.0323119777158776,
      "grad_norm": 1.6115546338586715,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8848904967308044,
      "num_tokens": 380817303.0,
      "step": 10944
    },
    {
      "epoch": 2.032497678737233,
      "grad_norm": 1.5842224738321997,
      "learning_rate": 1e-06,
      "loss": 0.2536,
      "mean_token_accuracy": 0.9092784523963928,
      "num_tokens": 380846166.0,
      "step": 10945
    },
    {
      "epoch": 2.0326833797585886,
      "grad_norm": 1.4800266293009128,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8840320706367493,
      "num_tokens": 380884908.0,
      "step": 10946
    },
    {
      "epoch": 2.0328690807799443,
      "grad_norm": 1.5658369240191137,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8851568698883057,
      "num_tokens": 380923721.0,
      "step": 10947
    },
    {
      "epoch": 2.0330547818013,
      "grad_norm": 1.562376440584136,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8946096301078796,
      "num_tokens": 380957450.0,
      "step": 10948
    },
    {
      "epoch": 2.0332404828226553,
      "grad_norm": 1.4354451408760809,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8877887725830078,
      "num_tokens": 380998319.0,
      "step": 10949
    },
    {
      "epoch": 2.033426183844011,
      "grad_norm": 1.5078935139265848,
      "learning_rate": 1e-06,
      "loss": 0.2643,
      "mean_token_accuracy": 0.9099732637405396,
      "num_tokens": 381029888.0,
      "step": 10950
    },
    {
      "epoch": 2.033611884865367,
      "grad_norm": 1.4389876833934323,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8923543691635132,
      "num_tokens": 381072354.0,
      "step": 10951
    },
    {
      "epoch": 2.0337975858867225,
      "grad_norm": 1.434285037646439,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8943219184875488,
      "num_tokens": 381108738.0,
      "step": 10952
    },
    {
      "epoch": 2.033983286908078,
      "grad_norm": 1.542751596326758,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8949660062789917,
      "num_tokens": 381141146.0,
      "step": 10953
    },
    {
      "epoch": 2.0341689879294336,
      "grad_norm": 1.7108905813920137,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8805781602859497,
      "num_tokens": 381172157.0,
      "step": 10954
    },
    {
      "epoch": 2.0343546889507893,
      "grad_norm": 1.623685097596731,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8703593015670776,
      "num_tokens": 381209724.0,
      "step": 10955
    },
    {
      "epoch": 2.034540389972145,
      "grad_norm": 1.7305072902874488,
      "learning_rate": 1e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8763192892074585,
      "num_tokens": 381239543.0,
      "step": 10956
    },
    {
      "epoch": 2.0347260909935003,
      "grad_norm": 1.4196555681767977,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8883271217346191,
      "num_tokens": 381280387.0,
      "step": 10957
    },
    {
      "epoch": 2.034911792014856,
      "grad_norm": 1.549478971812332,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8822972178459167,
      "num_tokens": 381316306.0,
      "step": 10958
    },
    {
      "epoch": 2.035097493036212,
      "grad_norm": 1.5017554569421676,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8736987113952637,
      "num_tokens": 381357194.0,
      "step": 10959
    },
    {
      "epoch": 2.0352831940575675,
      "grad_norm": 1.4958920088010688,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8838359713554382,
      "num_tokens": 381396659.0,
      "step": 10960
    },
    {
      "epoch": 2.035468895078923,
      "grad_norm": 1.662102938064444,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8784314393997192,
      "num_tokens": 381430635.0,
      "step": 10961
    },
    {
      "epoch": 2.0356545961002785,
      "grad_norm": 1.6267599185936448,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8867939710617065,
      "num_tokens": 381461495.0,
      "step": 10962
    },
    {
      "epoch": 2.0358402971216343,
      "grad_norm": 1.5801318746883022,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8881278038024902,
      "num_tokens": 381494910.0,
      "step": 10963
    },
    {
      "epoch": 2.0360259981429896,
      "grad_norm": 1.5516605741519671,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8948523998260498,
      "num_tokens": 381527342.0,
      "step": 10964
    },
    {
      "epoch": 2.0362116991643453,
      "grad_norm": 1.5123054655663335,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8775167465209961,
      "num_tokens": 381562695.0,
      "step": 10965
    },
    {
      "epoch": 2.036397400185701,
      "grad_norm": 1.5709604596671218,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8873002529144287,
      "num_tokens": 381598852.0,
      "step": 10966
    },
    {
      "epoch": 2.0365831012070568,
      "grad_norm": 1.651225627793181,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8765797019004822,
      "num_tokens": 381631094.0,
      "step": 10967
    },
    {
      "epoch": 2.036768802228412,
      "grad_norm": 1.6625060476300968,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8843854665756226,
      "num_tokens": 381669115.0,
      "step": 10968
    },
    {
      "epoch": 2.036954503249768,
      "grad_norm": 1.5698024112164144,
      "learning_rate": 1e-06,
      "loss": 0.2876,
      "mean_token_accuracy": 0.9010859727859497,
      "num_tokens": 381703652.0,
      "step": 10969
    },
    {
      "epoch": 2.0371402042711235,
      "grad_norm": 1.6950125640106264,
      "learning_rate": 1e-06,
      "loss": 0.2756,
      "mean_token_accuracy": 0.9043010473251343,
      "num_tokens": 381729655.0,
      "step": 10970
    },
    {
      "epoch": 2.0373259052924793,
      "grad_norm": 1.5459585735403492,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8869364261627197,
      "num_tokens": 381763819.0,
      "step": 10971
    },
    {
      "epoch": 2.0375116063138345,
      "grad_norm": 1.5693186754543742,
      "learning_rate": 1e-06,
      "loss": 0.2901,
      "mean_token_accuracy": 0.8989090919494629,
      "num_tokens": 381794699.0,
      "step": 10972
    },
    {
      "epoch": 2.0376973073351903,
      "grad_norm": 1.476051943153466,
      "learning_rate": 1e-06,
      "loss": 0.2833,
      "mean_token_accuracy": 0.9019213914871216,
      "num_tokens": 381832154.0,
      "step": 10973
    },
    {
      "epoch": 2.037883008356546,
      "grad_norm": 1.7281078786292856,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8864295482635498,
      "num_tokens": 381863805.0,
      "step": 10974
    },
    {
      "epoch": 2.0380687093779017,
      "grad_norm": 1.4746443971400134,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8810156583786011,
      "num_tokens": 381902038.0,
      "step": 10975
    },
    {
      "epoch": 2.038254410399257,
      "grad_norm": 1.8800804393659458,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8828727602958679,
      "num_tokens": 381925307.0,
      "step": 10976
    },
    {
      "epoch": 2.0384401114206128,
      "grad_norm": 1.6618255760914633,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8795924782752991,
      "num_tokens": 381963581.0,
      "step": 10977
    },
    {
      "epoch": 2.0386258124419685,
      "grad_norm": 1.6225241941461022,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.894372820854187,
      "num_tokens": 381997001.0,
      "step": 10978
    },
    {
      "epoch": 2.0388115134633242,
      "grad_norm": 1.72119988145317,
      "learning_rate": 1e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.869677722454071,
      "num_tokens": 382031885.0,
      "step": 10979
    },
    {
      "epoch": 2.0389972144846795,
      "grad_norm": 1.5136884698965476,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8776547908782959,
      "num_tokens": 382068832.0,
      "step": 10980
    },
    {
      "epoch": 2.0391829155060353,
      "grad_norm": 1.5842981933275238,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8874996900558472,
      "num_tokens": 382100200.0,
      "step": 10981
    },
    {
      "epoch": 2.039368616527391,
      "grad_norm": 1.4859815301843569,
      "learning_rate": 1e-06,
      "loss": 0.2729,
      "mean_token_accuracy": 0.9011412262916565,
      "num_tokens": 382132215.0,
      "step": 10982
    },
    {
      "epoch": 2.0395543175487467,
      "grad_norm": 1.613435986310082,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8950796127319336,
      "num_tokens": 382162410.0,
      "step": 10983
    },
    {
      "epoch": 2.039740018570102,
      "grad_norm": 1.5339938074781172,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8945439457893372,
      "num_tokens": 382196686.0,
      "step": 10984
    },
    {
      "epoch": 2.0399257195914577,
      "grad_norm": 1.426704818110976,
      "learning_rate": 1e-06,
      "loss": 0.2755,
      "mean_token_accuracy": 0.9027620553970337,
      "num_tokens": 382233804.0,
      "step": 10985
    },
    {
      "epoch": 2.0401114206128135,
      "grad_norm": 1.5646595367067306,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8867769241333008,
      "num_tokens": 382269980.0,
      "step": 10986
    },
    {
      "epoch": 2.040297121634169,
      "grad_norm": 1.4846409507691813,
      "learning_rate": 1e-06,
      "loss": 0.2831,
      "mean_token_accuracy": 0.9000144004821777,
      "num_tokens": 382304146.0,
      "step": 10987
    },
    {
      "epoch": 2.0404828226555245,
      "grad_norm": 1.616246743903049,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8934953808784485,
      "num_tokens": 382335170.0,
      "step": 10988
    },
    {
      "epoch": 2.0406685236768802,
      "grad_norm": 1.635957361641812,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8744686245918274,
      "num_tokens": 382370413.0,
      "step": 10989
    },
    {
      "epoch": 2.040854224698236,
      "grad_norm": 1.668957713977435,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.870050311088562,
      "num_tokens": 382403580.0,
      "step": 10990
    },
    {
      "epoch": 2.0410399257195913,
      "grad_norm": 1.6410132492394132,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8830288648605347,
      "num_tokens": 382437077.0,
      "step": 10991
    },
    {
      "epoch": 2.041225626740947,
      "grad_norm": 1.5178844094504609,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8945499062538147,
      "num_tokens": 382473388.0,
      "step": 10992
    },
    {
      "epoch": 2.0414113277623027,
      "grad_norm": 1.6411158333852087,
      "learning_rate": 1e-06,
      "loss": 0.3048,
      "mean_token_accuracy": 0.894298255443573,
      "num_tokens": 382505678.0,
      "step": 10993
    },
    {
      "epoch": 2.0415970287836585,
      "grad_norm": 1.7829766728969496,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8627475500106812,
      "num_tokens": 382535326.0,
      "step": 10994
    },
    {
      "epoch": 2.0417827298050137,
      "grad_norm": 1.7061783885784627,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8786081075668335,
      "num_tokens": 382567012.0,
      "step": 10995
    },
    {
      "epoch": 2.0419684308263695,
      "grad_norm": 1.7096856075326743,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8854364156723022,
      "num_tokens": 382597340.0,
      "step": 10996
    },
    {
      "epoch": 2.042154131847725,
      "grad_norm": 1.6595067220912731,
      "learning_rate": 1e-06,
      "loss": 0.3674,
      "mean_token_accuracy": 0.8693957328796387,
      "num_tokens": 382631557.0,
      "step": 10997
    },
    {
      "epoch": 2.042339832869081,
      "grad_norm": 1.659681033251493,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8849048614501953,
      "num_tokens": 382664389.0,
      "step": 10998
    },
    {
      "epoch": 2.0425255338904362,
      "grad_norm": 1.391197022066563,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8917527198791504,
      "num_tokens": 382706868.0,
      "step": 10999
    },
    {
      "epoch": 2.042711234911792,
      "grad_norm": 1.5610919174759004,
      "learning_rate": 1e-06,
      "loss": 0.2776,
      "mean_token_accuracy": 0.9006444811820984,
      "num_tokens": 382740397.0,
      "step": 11000
    },
    {
      "epoch": 2.0428969359331477,
      "grad_norm": 1.7124019927039702,
      "learning_rate": 1e-06,
      "loss": 0.3732,
      "mean_token_accuracy": 0.8698668479919434,
      "num_tokens": 382774498.0,
      "step": 11001
    },
    {
      "epoch": 2.0430826369545034,
      "grad_norm": 1.4469424356890221,
      "learning_rate": 1e-06,
      "loss": 0.272,
      "mean_token_accuracy": 0.9033824801445007,
      "num_tokens": 382812826.0,
      "step": 11002
    },
    {
      "epoch": 2.0432683379758587,
      "grad_norm": 1.6515754229913324,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8847622871398926,
      "num_tokens": 382844445.0,
      "step": 11003
    },
    {
      "epoch": 2.0434540389972145,
      "grad_norm": 1.682074183073453,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8912971019744873,
      "num_tokens": 382873675.0,
      "step": 11004
    },
    {
      "epoch": 2.04363974001857,
      "grad_norm": 1.6068381695411182,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8834320306777954,
      "num_tokens": 382906670.0,
      "step": 11005
    },
    {
      "epoch": 2.043825441039926,
      "grad_norm": 1.5108695428528525,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8740382194519043,
      "num_tokens": 382945305.0,
      "step": 11006
    },
    {
      "epoch": 2.044011142061281,
      "grad_norm": 1.6753524443533303,
      "learning_rate": 1e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8590644598007202,
      "num_tokens": 382981923.0,
      "step": 11007
    },
    {
      "epoch": 2.044196843082637,
      "grad_norm": 1.6100879254421188,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.886062741279602,
      "num_tokens": 383014967.0,
      "step": 11008
    },
    {
      "epoch": 2.0443825441039927,
      "grad_norm": 1.6238807236307442,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8855381011962891,
      "num_tokens": 383047992.0,
      "step": 11009
    },
    {
      "epoch": 2.0445682451253484,
      "grad_norm": 1.378355114519235,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8940566778182983,
      "num_tokens": 383086086.0,
      "step": 11010
    },
    {
      "epoch": 2.0447539461467037,
      "grad_norm": 1.5494596352377128,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8910616636276245,
      "num_tokens": 383120999.0,
      "step": 11011
    },
    {
      "epoch": 2.0449396471680594,
      "grad_norm": 1.5902883686440104,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8763071894645691,
      "num_tokens": 383155319.0,
      "step": 11012
    },
    {
      "epoch": 2.045125348189415,
      "grad_norm": 1.5418378979700906,
      "learning_rate": 1e-06,
      "loss": 0.2914,
      "mean_token_accuracy": 0.8963024616241455,
      "num_tokens": 383186240.0,
      "step": 11013
    },
    {
      "epoch": 2.0453110492107704,
      "grad_norm": 1.7187997930445147,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8886537551879883,
      "num_tokens": 383214002.0,
      "step": 11014
    },
    {
      "epoch": 2.045496750232126,
      "grad_norm": 1.6098855170330386,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8881896734237671,
      "num_tokens": 383247979.0,
      "step": 11015
    },
    {
      "epoch": 2.045682451253482,
      "grad_norm": 1.5674690471112271,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8757724761962891,
      "num_tokens": 383281792.0,
      "step": 11016
    },
    {
      "epoch": 2.0458681522748376,
      "grad_norm": 1.6629028689888699,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.881744384765625,
      "num_tokens": 383315159.0,
      "step": 11017
    },
    {
      "epoch": 2.046053853296193,
      "grad_norm": 1.581221618231081,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8882949948310852,
      "num_tokens": 383349021.0,
      "step": 11018
    },
    {
      "epoch": 2.0462395543175487,
      "grad_norm": 1.5689059885226362,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8963765501976013,
      "num_tokens": 383381717.0,
      "step": 11019
    },
    {
      "epoch": 2.0464252553389044,
      "grad_norm": 1.5427962196830354,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8862926363945007,
      "num_tokens": 383416912.0,
      "step": 11020
    },
    {
      "epoch": 2.04661095636026,
      "grad_norm": 1.5746785934760161,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8793112635612488,
      "num_tokens": 383451649.0,
      "step": 11021
    },
    {
      "epoch": 2.0467966573816154,
      "grad_norm": 1.5656755250701218,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8906410336494446,
      "num_tokens": 383488472.0,
      "step": 11022
    },
    {
      "epoch": 2.046982358402971,
      "grad_norm": 1.519603384503179,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.88411545753479,
      "num_tokens": 383524805.0,
      "step": 11023
    },
    {
      "epoch": 2.047168059424327,
      "grad_norm": 1.6994332690608596,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8844226598739624,
      "num_tokens": 383557659.0,
      "step": 11024
    },
    {
      "epoch": 2.0473537604456826,
      "grad_norm": 1.659344389912768,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8729217052459717,
      "num_tokens": 383593044.0,
      "step": 11025
    },
    {
      "epoch": 2.047539461467038,
      "grad_norm": 1.6737450743515958,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8917562961578369,
      "num_tokens": 383624982.0,
      "step": 11026
    },
    {
      "epoch": 2.0477251624883936,
      "grad_norm": 1.51478074350044,
      "learning_rate": 1e-06,
      "loss": 0.3553,
      "mean_token_accuracy": 0.8765978813171387,
      "num_tokens": 383663539.0,
      "step": 11027
    },
    {
      "epoch": 2.0479108635097494,
      "grad_norm": 1.692051205173449,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8921996355056763,
      "num_tokens": 383691132.0,
      "step": 11028
    },
    {
      "epoch": 2.048096564531105,
      "grad_norm": 1.4741178158091346,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8864985704421997,
      "num_tokens": 383730088.0,
      "step": 11029
    },
    {
      "epoch": 2.0482822655524604,
      "grad_norm": 1.6979347514625498,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8813077807426453,
      "num_tokens": 383762567.0,
      "step": 11030
    },
    {
      "epoch": 2.048467966573816,
      "grad_norm": 1.5134113923087718,
      "learning_rate": 1e-06,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9055246114730835,
      "num_tokens": 383794901.0,
      "step": 11031
    },
    {
      "epoch": 2.048653667595172,
      "grad_norm": 1.5267660937704872,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8816649317741394,
      "num_tokens": 383832594.0,
      "step": 11032
    },
    {
      "epoch": 2.0488393686165276,
      "grad_norm": 1.5788830060395966,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8745362758636475,
      "num_tokens": 383869172.0,
      "step": 11033
    },
    {
      "epoch": 2.049025069637883,
      "grad_norm": 1.5355046911878634,
      "learning_rate": 1e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8807247877120972,
      "num_tokens": 383906403.0,
      "step": 11034
    },
    {
      "epoch": 2.0492107706592386,
      "grad_norm": 1.5043664075658283,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8696513175964355,
      "num_tokens": 383945093.0,
      "step": 11035
    },
    {
      "epoch": 2.0493964716805944,
      "grad_norm": 1.4470644778752784,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8913853168487549,
      "num_tokens": 383982572.0,
      "step": 11036
    },
    {
      "epoch": 2.0495821727019496,
      "grad_norm": 1.5681689837186894,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8897879719734192,
      "num_tokens": 384016170.0,
      "step": 11037
    },
    {
      "epoch": 2.0497678737233054,
      "grad_norm": 1.589211787795371,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8789228200912476,
      "num_tokens": 384051411.0,
      "step": 11038
    },
    {
      "epoch": 2.049953574744661,
      "grad_norm": 1.521497954653133,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.882239580154419,
      "num_tokens": 384090152.0,
      "step": 11039
    },
    {
      "epoch": 2.050139275766017,
      "grad_norm": 1.5866458923165765,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8779263496398926,
      "num_tokens": 384124154.0,
      "step": 11040
    },
    {
      "epoch": 2.050324976787372,
      "grad_norm": 1.5425321602253321,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8893698453903198,
      "num_tokens": 384161495.0,
      "step": 11041
    },
    {
      "epoch": 2.050510677808728,
      "grad_norm": 1.6475299499354032,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.8965858221054077,
      "num_tokens": 384189846.0,
      "step": 11042
    },
    {
      "epoch": 2.0506963788300836,
      "grad_norm": 1.6871454144993256,
      "learning_rate": 1e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8633660078048706,
      "num_tokens": 384226526.0,
      "step": 11043
    },
    {
      "epoch": 2.0508820798514393,
      "grad_norm": 1.5350230404242253,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.882299542427063,
      "num_tokens": 384269975.0,
      "step": 11044
    },
    {
      "epoch": 2.0510677808727946,
      "grad_norm": 1.7661355831784715,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.883130669593811,
      "num_tokens": 384298974.0,
      "step": 11045
    },
    {
      "epoch": 2.0512534818941504,
      "grad_norm": 1.4892009498890477,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8874703645706177,
      "num_tokens": 384338588.0,
      "step": 11046
    },
    {
      "epoch": 2.051439182915506,
      "grad_norm": 1.4967368310933455,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8926022052764893,
      "num_tokens": 384372448.0,
      "step": 11047
    },
    {
      "epoch": 2.051624883936862,
      "grad_norm": 1.4162978251891034,
      "learning_rate": 1e-06,
      "loss": 0.2772,
      "mean_token_accuracy": 0.8984420895576477,
      "num_tokens": 384410097.0,
      "step": 11048
    },
    {
      "epoch": 2.051810584958217,
      "grad_norm": 1.6159995821122124,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8812836408615112,
      "num_tokens": 384441441.0,
      "step": 11049
    },
    {
      "epoch": 2.051996285979573,
      "grad_norm": 1.5062571537596197,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8832650780677795,
      "num_tokens": 384478204.0,
      "step": 11050
    },
    {
      "epoch": 2.0521819870009286,
      "grad_norm": 1.4630498327160442,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8765629529953003,
      "num_tokens": 384516015.0,
      "step": 11051
    },
    {
      "epoch": 2.0523676880222843,
      "grad_norm": 1.6886636930126186,
      "learning_rate": 1e-06,
      "loss": 0.2972,
      "mean_token_accuracy": 0.8939827084541321,
      "num_tokens": 384546364.0,
      "step": 11052
    },
    {
      "epoch": 2.0525533890436396,
      "grad_norm": 1.5795909053554544,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.886978268623352,
      "num_tokens": 384580016.0,
      "step": 11053
    },
    {
      "epoch": 2.0527390900649953,
      "grad_norm": 1.456318585839607,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8941553831100464,
      "num_tokens": 384615539.0,
      "step": 11054
    },
    {
      "epoch": 2.052924791086351,
      "grad_norm": 1.5803394233485752,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8895382881164551,
      "num_tokens": 384647203.0,
      "step": 11055
    },
    {
      "epoch": 2.053110492107707,
      "grad_norm": 1.553783696475382,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8793415427207947,
      "num_tokens": 384685857.0,
      "step": 11056
    },
    {
      "epoch": 2.053296193129062,
      "grad_norm": 1.5297826741674456,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8873360753059387,
      "num_tokens": 384720804.0,
      "step": 11057
    },
    {
      "epoch": 2.053481894150418,
      "grad_norm": 1.488675997222662,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8801141977310181,
      "num_tokens": 384760605.0,
      "step": 11058
    },
    {
      "epoch": 2.0536675951717736,
      "grad_norm": 1.5896206542932172,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8874995708465576,
      "num_tokens": 384793453.0,
      "step": 11059
    },
    {
      "epoch": 2.053853296193129,
      "grad_norm": 1.526253469253286,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.888975203037262,
      "num_tokens": 384828795.0,
      "step": 11060
    },
    {
      "epoch": 2.0540389972144846,
      "grad_norm": 1.7319630295220738,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8802059292793274,
      "num_tokens": 384857940.0,
      "step": 11061
    },
    {
      "epoch": 2.0542246982358403,
      "grad_norm": 1.482164729344953,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8920994400978088,
      "num_tokens": 384897846.0,
      "step": 11062
    },
    {
      "epoch": 2.054410399257196,
      "grad_norm": 1.5274249913558895,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8912224769592285,
      "num_tokens": 384934863.0,
      "step": 11063
    },
    {
      "epoch": 2.0545961002785513,
      "grad_norm": 1.6226435435710032,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8926545977592468,
      "num_tokens": 384966010.0,
      "step": 11064
    },
    {
      "epoch": 2.054781801299907,
      "grad_norm": 1.5850312530100061,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8783977031707764,
      "num_tokens": 385002473.0,
      "step": 11065
    },
    {
      "epoch": 2.054967502321263,
      "grad_norm": 1.6075329282766049,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8881728649139404,
      "num_tokens": 385037522.0,
      "step": 11066
    },
    {
      "epoch": 2.0551532033426185,
      "grad_norm": 1.4865410030535575,
      "learning_rate": 1e-06,
      "loss": 0.2929,
      "mean_token_accuracy": 0.8967502117156982,
      "num_tokens": 385072921.0,
      "step": 11067
    },
    {
      "epoch": 2.055338904363974,
      "grad_norm": 1.5460742477751868,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8856135606765747,
      "num_tokens": 385109879.0,
      "step": 11068
    },
    {
      "epoch": 2.0555246053853296,
      "grad_norm": 1.5590984371974592,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8937326669692993,
      "num_tokens": 385142660.0,
      "step": 11069
    },
    {
      "epoch": 2.0557103064066853,
      "grad_norm": 1.5718373187948478,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8916704654693604,
      "num_tokens": 385175492.0,
      "step": 11070
    },
    {
      "epoch": 2.055896007428041,
      "grad_norm": 1.7219678814278525,
      "learning_rate": 1e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.871839165687561,
      "num_tokens": 385208290.0,
      "step": 11071
    },
    {
      "epoch": 2.0560817084493963,
      "grad_norm": 1.673545844402136,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8820183873176575,
      "num_tokens": 385236801.0,
      "step": 11072
    },
    {
      "epoch": 2.056267409470752,
      "grad_norm": 1.5695796554441495,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.8943193554878235,
      "num_tokens": 385268938.0,
      "step": 11073
    },
    {
      "epoch": 2.0564531104921078,
      "grad_norm": 1.5500169225985063,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8974606990814209,
      "num_tokens": 385301893.0,
      "step": 11074
    },
    {
      "epoch": 2.0566388115134635,
      "grad_norm": 1.6118145640661938,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8873066306114197,
      "num_tokens": 385334041.0,
      "step": 11075
    },
    {
      "epoch": 2.056824512534819,
      "grad_norm": 1.4008812351484692,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8906200528144836,
      "num_tokens": 385376807.0,
      "step": 11076
    },
    {
      "epoch": 2.0570102135561745,
      "grad_norm": 1.5414007007987423,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8904807567596436,
      "num_tokens": 385411821.0,
      "step": 11077
    },
    {
      "epoch": 2.0571959145775303,
      "grad_norm": 1.5958694139302694,
      "learning_rate": 1e-06,
      "loss": 0.3048,
      "mean_token_accuracy": 0.8957756757736206,
      "num_tokens": 385443949.0,
      "step": 11078
    },
    {
      "epoch": 2.057381615598886,
      "grad_norm": 1.536141441167473,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.891160249710083,
      "num_tokens": 385475391.0,
      "step": 11079
    },
    {
      "epoch": 2.0575673166202413,
      "grad_norm": 1.4452109336824217,
      "learning_rate": 1e-06,
      "loss": 0.2703,
      "mean_token_accuracy": 0.907342255115509,
      "num_tokens": 385507714.0,
      "step": 11080
    },
    {
      "epoch": 2.057753017641597,
      "grad_norm": 1.726983097516361,
      "learning_rate": 1e-06,
      "loss": 0.378,
      "mean_token_accuracy": 0.8680486679077148,
      "num_tokens": 385537266.0,
      "step": 11081
    },
    {
      "epoch": 2.0579387186629527,
      "grad_norm": 1.4188967859699053,
      "learning_rate": 1e-06,
      "loss": 0.2933,
      "mean_token_accuracy": 0.8958138823509216,
      "num_tokens": 385574692.0,
      "step": 11082
    },
    {
      "epoch": 2.0581244196843085,
      "grad_norm": 1.5096762114532694,
      "learning_rate": 1e-06,
      "loss": 0.2805,
      "mean_token_accuracy": 0.8994975090026855,
      "num_tokens": 385605652.0,
      "step": 11083
    },
    {
      "epoch": 2.0583101207056638,
      "grad_norm": 1.4955114220239094,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8891794681549072,
      "num_tokens": 385642676.0,
      "step": 11084
    },
    {
      "epoch": 2.0584958217270195,
      "grad_norm": 1.5533713176765218,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8852990865707397,
      "num_tokens": 385675836.0,
      "step": 11085
    },
    {
      "epoch": 2.0586815227483752,
      "grad_norm": 1.5506880244506736,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8910456895828247,
      "num_tokens": 385710484.0,
      "step": 11086
    },
    {
      "epoch": 2.0588672237697305,
      "grad_norm": 1.5787166541799968,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8921476006507874,
      "num_tokens": 385744256.0,
      "step": 11087
    },
    {
      "epoch": 2.0590529247910863,
      "grad_norm": 1.4978737453787545,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8906474113464355,
      "num_tokens": 385780104.0,
      "step": 11088
    },
    {
      "epoch": 2.059238625812442,
      "grad_norm": 1.4381197771090282,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.885317862033844,
      "num_tokens": 385822710.0,
      "step": 11089
    },
    {
      "epoch": 2.0594243268337977,
      "grad_norm": 1.7700228966119838,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8807740807533264,
      "num_tokens": 385855102.0,
      "step": 11090
    },
    {
      "epoch": 2.059610027855153,
      "grad_norm": 1.5012090205754542,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8722571730613708,
      "num_tokens": 385896626.0,
      "step": 11091
    },
    {
      "epoch": 2.0597957288765087,
      "grad_norm": 1.5214212287383468,
      "learning_rate": 1e-06,
      "loss": 0.2712,
      "mean_token_accuracy": 0.9016943573951721,
      "num_tokens": 385931598.0,
      "step": 11092
    },
    {
      "epoch": 2.0599814298978645,
      "grad_norm": 1.4709364517005497,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8875488638877869,
      "num_tokens": 385970948.0,
      "step": 11093
    },
    {
      "epoch": 2.06016713091922,
      "grad_norm": 1.4564041265464531,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8959574103355408,
      "num_tokens": 386008934.0,
      "step": 11094
    },
    {
      "epoch": 2.0603528319405755,
      "grad_norm": 1.5954706688099456,
      "learning_rate": 1e-06,
      "loss": 0.3088,
      "mean_token_accuracy": 0.8911401033401489,
      "num_tokens": 386042829.0,
      "step": 11095
    },
    {
      "epoch": 2.0605385329619312,
      "grad_norm": 1.5076345604749524,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8873542547225952,
      "num_tokens": 386080984.0,
      "step": 11096
    },
    {
      "epoch": 2.060724233983287,
      "grad_norm": 1.4724387219826465,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8996864557266235,
      "num_tokens": 386118591.0,
      "step": 11097
    },
    {
      "epoch": 2.0609099350046427,
      "grad_norm": 1.61702559983188,
      "learning_rate": 1e-06,
      "loss": 0.2945,
      "mean_token_accuracy": 0.8938292264938354,
      "num_tokens": 386149784.0,
      "step": 11098
    },
    {
      "epoch": 2.061095636025998,
      "grad_norm": 1.584672448631015,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.880540132522583,
      "num_tokens": 386184040.0,
      "step": 11099
    },
    {
      "epoch": 2.0612813370473537,
      "grad_norm": 1.467936650953849,
      "learning_rate": 1e-06,
      "loss": 0.2766,
      "mean_token_accuracy": 0.9027755856513977,
      "num_tokens": 386217017.0,
      "step": 11100
    },
    {
      "epoch": 2.0614670380687095,
      "grad_norm": 1.524759065595606,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8900650143623352,
      "num_tokens": 386253956.0,
      "step": 11101
    },
    {
      "epoch": 2.061652739090065,
      "grad_norm": 1.6579770353022427,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8883261680603027,
      "num_tokens": 386288194.0,
      "step": 11102
    },
    {
      "epoch": 2.0618384401114205,
      "grad_norm": 1.4782158257295472,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8907886743545532,
      "num_tokens": 386324811.0,
      "step": 11103
    },
    {
      "epoch": 2.062024141132776,
      "grad_norm": 1.5002424218586956,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8821288347244263,
      "num_tokens": 386363974.0,
      "step": 11104
    },
    {
      "epoch": 2.062209842154132,
      "grad_norm": 1.4465878715442306,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8842893242835999,
      "num_tokens": 386405436.0,
      "step": 11105
    },
    {
      "epoch": 2.0623955431754877,
      "grad_norm": 1.540476074491471,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8839633464813232,
      "num_tokens": 386440291.0,
      "step": 11106
    },
    {
      "epoch": 2.062581244196843,
      "grad_norm": 1.541579431720527,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8877382874488831,
      "num_tokens": 386475817.0,
      "step": 11107
    },
    {
      "epoch": 2.0627669452181987,
      "grad_norm": 1.4333788471057791,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8788732290267944,
      "num_tokens": 386517746.0,
      "step": 11108
    },
    {
      "epoch": 2.0629526462395544,
      "grad_norm": 1.6608045819299677,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8851949572563171,
      "num_tokens": 386547745.0,
      "step": 11109
    },
    {
      "epoch": 2.0631383472609097,
      "grad_norm": 1.6935473046958358,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8929070234298706,
      "num_tokens": 386576391.0,
      "step": 11110
    },
    {
      "epoch": 2.0633240482822655,
      "grad_norm": 1.6417914247779166,
      "learning_rate": 1e-06,
      "loss": 0.2725,
      "mean_token_accuracy": 0.903218686580658,
      "num_tokens": 386604775.0,
      "step": 11111
    },
    {
      "epoch": 2.063509749303621,
      "grad_norm": 1.514399357093865,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8747810125350952,
      "num_tokens": 386645535.0,
      "step": 11112
    },
    {
      "epoch": 2.063695450324977,
      "grad_norm": 1.6385388507993772,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8892330527305603,
      "num_tokens": 386676479.0,
      "step": 11113
    },
    {
      "epoch": 2.063881151346332,
      "grad_norm": 1.636907670804244,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8736562132835388,
      "num_tokens": 386710229.0,
      "step": 11114
    },
    {
      "epoch": 2.064066852367688,
      "grad_norm": 1.6984447983090163,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8871641755104065,
      "num_tokens": 386741036.0,
      "step": 11115
    },
    {
      "epoch": 2.0642525533890437,
      "grad_norm": 1.5273640038580416,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8881272077560425,
      "num_tokens": 386776570.0,
      "step": 11116
    },
    {
      "epoch": 2.0644382544103994,
      "grad_norm": 1.4958852038485535,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8902711868286133,
      "num_tokens": 386813775.0,
      "step": 11117
    },
    {
      "epoch": 2.0646239554317547,
      "grad_norm": 1.4679177554956628,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8903080821037292,
      "num_tokens": 386853475.0,
      "step": 11118
    },
    {
      "epoch": 2.0648096564531104,
      "grad_norm": 1.7345911407709846,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8782720565795898,
      "num_tokens": 386881055.0,
      "step": 11119
    },
    {
      "epoch": 2.064995357474466,
      "grad_norm": 1.574229435626189,
      "learning_rate": 1e-06,
      "loss": 0.2984,
      "mean_token_accuracy": 0.891228973865509,
      "num_tokens": 386913985.0,
      "step": 11120
    },
    {
      "epoch": 2.065181058495822,
      "grad_norm": 1.4866719738006138,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.881558895111084,
      "num_tokens": 386953658.0,
      "step": 11121
    },
    {
      "epoch": 2.065366759517177,
      "grad_norm": 1.7249832177757105,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8810991048812866,
      "num_tokens": 386984592.0,
      "step": 11122
    },
    {
      "epoch": 2.065552460538533,
      "grad_norm": 1.4430542967569941,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8959071636199951,
      "num_tokens": 387020474.0,
      "step": 11123
    },
    {
      "epoch": 2.0657381615598887,
      "grad_norm": 1.6632497232461527,
      "learning_rate": 1e-06,
      "loss": 0.2719,
      "mean_token_accuracy": 0.9012742638587952,
      "num_tokens": 387047781.0,
      "step": 11124
    },
    {
      "epoch": 2.0659238625812444,
      "grad_norm": 1.6163822966981025,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8899046182632446,
      "num_tokens": 387081178.0,
      "step": 11125
    },
    {
      "epoch": 2.0661095636025997,
      "grad_norm": 1.5594962829840928,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8855043649673462,
      "num_tokens": 387114549.0,
      "step": 11126
    },
    {
      "epoch": 2.0662952646239554,
      "grad_norm": 1.429388184706365,
      "learning_rate": 1e-06,
      "loss": 0.2786,
      "mean_token_accuracy": 0.9004614949226379,
      "num_tokens": 387151418.0,
      "step": 11127
    },
    {
      "epoch": 2.066480965645311,
      "grad_norm": 1.4613277202339272,
      "learning_rate": 1e-06,
      "loss": 0.2998,
      "mean_token_accuracy": 0.8926539421081543,
      "num_tokens": 387186893.0,
      "step": 11128
    },
    {
      "epoch": 2.066666666666667,
      "grad_norm": 1.710153836595586,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8964316844940186,
      "num_tokens": 387213542.0,
      "step": 11129
    },
    {
      "epoch": 2.066852367688022,
      "grad_norm": 1.623342484441844,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8825249075889587,
      "num_tokens": 387246903.0,
      "step": 11130
    },
    {
      "epoch": 2.067038068709378,
      "grad_norm": 1.5198835849413885,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8847843408584595,
      "num_tokens": 387283023.0,
      "step": 11131
    },
    {
      "epoch": 2.0672237697307336,
      "grad_norm": 1.4522500913414003,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8965354561805725,
      "num_tokens": 387317857.0,
      "step": 11132
    },
    {
      "epoch": 2.0674094707520894,
      "grad_norm": 1.4453526076529732,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8831458687782288,
      "num_tokens": 387357179.0,
      "step": 11133
    },
    {
      "epoch": 2.0675951717734447,
      "grad_norm": 1.5371783660121192,
      "learning_rate": 1e-06,
      "loss": 0.2726,
      "mean_token_accuracy": 0.9018492102622986,
      "num_tokens": 387389119.0,
      "step": 11134
    },
    {
      "epoch": 2.0677808727948004,
      "grad_norm": 1.5136095775045357,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.8939889073371887,
      "num_tokens": 387421927.0,
      "step": 11135
    },
    {
      "epoch": 2.067966573816156,
      "grad_norm": 1.6614212363181244,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8857803344726562,
      "num_tokens": 387456055.0,
      "step": 11136
    },
    {
      "epoch": 2.0681522748375114,
      "grad_norm": 1.73396731622181,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8881430625915527,
      "num_tokens": 387483393.0,
      "step": 11137
    },
    {
      "epoch": 2.068337975858867,
      "grad_norm": 1.665799911867806,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8793803453445435,
      "num_tokens": 387513130.0,
      "step": 11138
    },
    {
      "epoch": 2.068523676880223,
      "grad_norm": 1.7064224369355243,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8843119740486145,
      "num_tokens": 387541904.0,
      "step": 11139
    },
    {
      "epoch": 2.0687093779015786,
      "grad_norm": 1.6213807296401168,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8803340196609497,
      "num_tokens": 387573252.0,
      "step": 11140
    },
    {
      "epoch": 2.068895078922934,
      "grad_norm": 1.5920525141175959,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8869184255599976,
      "num_tokens": 387607220.0,
      "step": 11141
    },
    {
      "epoch": 2.0690807799442896,
      "grad_norm": 1.7300451285275842,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8872699737548828,
      "num_tokens": 387641691.0,
      "step": 11142
    },
    {
      "epoch": 2.0692664809656454,
      "grad_norm": 1.7215861388191496,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8837707042694092,
      "num_tokens": 387675943.0,
      "step": 11143
    },
    {
      "epoch": 2.069452181987001,
      "grad_norm": 1.5716644168725726,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.8985364437103271,
      "num_tokens": 387708743.0,
      "step": 11144
    },
    {
      "epoch": 2.0696378830083564,
      "grad_norm": 1.6899400327591563,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8805937767028809,
      "num_tokens": 387738790.0,
      "step": 11145
    },
    {
      "epoch": 2.069823584029712,
      "grad_norm": 1.5997071559646407,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.888590931892395,
      "num_tokens": 387770888.0,
      "step": 11146
    },
    {
      "epoch": 2.070009285051068,
      "grad_norm": 1.6041058197070805,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8881582617759705,
      "num_tokens": 387803690.0,
      "step": 11147
    },
    {
      "epoch": 2.0701949860724236,
      "grad_norm": 1.6716536331426088,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8862728476524353,
      "num_tokens": 387833184.0,
      "step": 11148
    },
    {
      "epoch": 2.070380687093779,
      "grad_norm": 1.5978581202242361,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8897364139556885,
      "num_tokens": 387864317.0,
      "step": 11149
    },
    {
      "epoch": 2.0705663881151346,
      "grad_norm": 1.433584091547605,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8964398503303528,
      "num_tokens": 387897883.0,
      "step": 11150
    },
    {
      "epoch": 2.0707520891364903,
      "grad_norm": 1.5804467438484782,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8883338570594788,
      "num_tokens": 387930578.0,
      "step": 11151
    },
    {
      "epoch": 2.070937790157846,
      "grad_norm": 1.5157542464858016,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8827990293502808,
      "num_tokens": 387963911.0,
      "step": 11152
    },
    {
      "epoch": 2.0711234911792014,
      "grad_norm": 1.6225959694251755,
      "learning_rate": 1e-06,
      "loss": 0.2881,
      "mean_token_accuracy": 0.8950940370559692,
      "num_tokens": 387995022.0,
      "step": 11153
    },
    {
      "epoch": 2.071309192200557,
      "grad_norm": 1.6999808118927724,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8832470178604126,
      "num_tokens": 388030558.0,
      "step": 11154
    },
    {
      "epoch": 2.071494893221913,
      "grad_norm": 1.4854985590852454,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.891616702079773,
      "num_tokens": 388067130.0,
      "step": 11155
    },
    {
      "epoch": 2.0716805942432686,
      "grad_norm": 1.6427406968649692,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8841879367828369,
      "num_tokens": 388100269.0,
      "step": 11156
    },
    {
      "epoch": 2.071866295264624,
      "grad_norm": 1.6350516107181938,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8895742893218994,
      "num_tokens": 388132222.0,
      "step": 11157
    },
    {
      "epoch": 2.0720519962859796,
      "grad_norm": 1.59361544041057,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8863392472267151,
      "num_tokens": 388168886.0,
      "step": 11158
    },
    {
      "epoch": 2.0722376973073353,
      "grad_norm": 1.5023866656047333,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.882082998752594,
      "num_tokens": 388208626.0,
      "step": 11159
    },
    {
      "epoch": 2.0724233983286906,
      "grad_norm": 1.8241203863445439,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8794006109237671,
      "num_tokens": 388233169.0,
      "step": 11160
    },
    {
      "epoch": 2.0726090993500463,
      "grad_norm": 1.638691987124303,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8869218230247498,
      "num_tokens": 388266161.0,
      "step": 11161
    },
    {
      "epoch": 2.072794800371402,
      "grad_norm": 1.6099719253963085,
      "learning_rate": 1e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8809289336204529,
      "num_tokens": 388302832.0,
      "step": 11162
    },
    {
      "epoch": 2.072980501392758,
      "grad_norm": 1.5457457539644641,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8859621286392212,
      "num_tokens": 388338550.0,
      "step": 11163
    },
    {
      "epoch": 2.073166202414113,
      "grad_norm": 1.689088978744731,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.8760968446731567,
      "num_tokens": 388372288.0,
      "step": 11164
    },
    {
      "epoch": 2.073351903435469,
      "grad_norm": 1.473117112864367,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8837829232215881,
      "num_tokens": 388413775.0,
      "step": 11165
    },
    {
      "epoch": 2.0735376044568246,
      "grad_norm": 1.5958577378629442,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8858568668365479,
      "num_tokens": 388454800.0,
      "step": 11166
    },
    {
      "epoch": 2.0737233054781803,
      "grad_norm": 1.5501845501142395,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.8935144543647766,
      "num_tokens": 388488940.0,
      "step": 11167
    },
    {
      "epoch": 2.0739090064995356,
      "grad_norm": 1.6519827812867458,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8900048136711121,
      "num_tokens": 388519678.0,
      "step": 11168
    },
    {
      "epoch": 2.0740947075208913,
      "grad_norm": 1.636684922734428,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8746795654296875,
      "num_tokens": 388553472.0,
      "step": 11169
    },
    {
      "epoch": 2.074280408542247,
      "grad_norm": 1.515170542891432,
      "learning_rate": 1e-06,
      "loss": 0.2785,
      "mean_token_accuracy": 0.900192141532898,
      "num_tokens": 388587782.0,
      "step": 11170
    },
    {
      "epoch": 2.074466109563603,
      "grad_norm": 1.5368844036533302,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8896070718765259,
      "num_tokens": 388622442.0,
      "step": 11171
    },
    {
      "epoch": 2.074651810584958,
      "grad_norm": 1.574218441320373,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8866552114486694,
      "num_tokens": 388656333.0,
      "step": 11172
    },
    {
      "epoch": 2.074837511606314,
      "grad_norm": 1.4462040146067536,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8920022249221802,
      "num_tokens": 388696455.0,
      "step": 11173
    },
    {
      "epoch": 2.0750232126276695,
      "grad_norm": 1.6191793482492025,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8843207359313965,
      "num_tokens": 388731868.0,
      "step": 11174
    },
    {
      "epoch": 2.0752089136490253,
      "grad_norm": 1.67161533268626,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8779854774475098,
      "num_tokens": 388765465.0,
      "step": 11175
    },
    {
      "epoch": 2.0753946146703806,
      "grad_norm": 1.4650034306732782,
      "learning_rate": 1e-06,
      "loss": 0.285,
      "mean_token_accuracy": 0.8976691961288452,
      "num_tokens": 388803358.0,
      "step": 11176
    },
    {
      "epoch": 2.0755803156917363,
      "grad_norm": 1.6971016952308533,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8790125846862793,
      "num_tokens": 388835984.0,
      "step": 11177
    },
    {
      "epoch": 2.075766016713092,
      "grad_norm": 1.5340848265806297,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8948545455932617,
      "num_tokens": 388874046.0,
      "step": 11178
    },
    {
      "epoch": 2.0759517177344478,
      "grad_norm": 1.509140748236438,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8957853317260742,
      "num_tokens": 388911973.0,
      "step": 11179
    },
    {
      "epoch": 2.076137418755803,
      "grad_norm": 1.5652753921267886,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8901939392089844,
      "num_tokens": 388943843.0,
      "step": 11180
    },
    {
      "epoch": 2.0763231197771588,
      "grad_norm": 1.6477154053860539,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.882296085357666,
      "num_tokens": 388979375.0,
      "step": 11181
    },
    {
      "epoch": 2.0765088207985145,
      "grad_norm": 1.6391584768465302,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.882461667060852,
      "num_tokens": 389014031.0,
      "step": 11182
    },
    {
      "epoch": 2.07669452181987,
      "grad_norm": 1.6081442817601188,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8860547542572021,
      "num_tokens": 389046478.0,
      "step": 11183
    },
    {
      "epoch": 2.0768802228412255,
      "grad_norm": 1.6434226608024483,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.875015377998352,
      "num_tokens": 389082130.0,
      "step": 11184
    },
    {
      "epoch": 2.0770659238625813,
      "grad_norm": 1.6475840202899188,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8815358877182007,
      "num_tokens": 389115682.0,
      "step": 11185
    },
    {
      "epoch": 2.077251624883937,
      "grad_norm": 1.477981574232701,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8889507055282593,
      "num_tokens": 389156938.0,
      "step": 11186
    },
    {
      "epoch": 2.0774373259052923,
      "grad_norm": 1.6250368210539872,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8839303255081177,
      "num_tokens": 389191646.0,
      "step": 11187
    },
    {
      "epoch": 2.077623026926648,
      "grad_norm": 1.54908018467715,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8898172974586487,
      "num_tokens": 389226192.0,
      "step": 11188
    },
    {
      "epoch": 2.0778087279480038,
      "grad_norm": 1.5472179075313686,
      "learning_rate": 1e-06,
      "loss": 0.2636,
      "mean_token_accuracy": 0.9035972356796265,
      "num_tokens": 389255334.0,
      "step": 11189
    },
    {
      "epoch": 2.0779944289693595,
      "grad_norm": 1.5130371361168795,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.883324146270752,
      "num_tokens": 389293775.0,
      "step": 11190
    },
    {
      "epoch": 2.0781801299907148,
      "grad_norm": 1.5480657245841576,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8937095403671265,
      "num_tokens": 389324575.0,
      "step": 11191
    },
    {
      "epoch": 2.0783658310120705,
      "grad_norm": 1.5753782246819132,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8770003318786621,
      "num_tokens": 389358626.0,
      "step": 11192
    },
    {
      "epoch": 2.0785515320334262,
      "grad_norm": 1.436586668854813,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8769569993019104,
      "num_tokens": 389397165.0,
      "step": 11193
    },
    {
      "epoch": 2.078737233054782,
      "grad_norm": 1.6082200991407882,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8761512637138367,
      "num_tokens": 389435910.0,
      "step": 11194
    },
    {
      "epoch": 2.0789229340761373,
      "grad_norm": 1.5999988713952982,
      "learning_rate": 1e-06,
      "loss": 0.3222,
      "mean_token_accuracy": 0.8872569799423218,
      "num_tokens": 389468898.0,
      "step": 11195
    },
    {
      "epoch": 2.079108635097493,
      "grad_norm": 1.445496243217497,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8854109048843384,
      "num_tokens": 389509901.0,
      "step": 11196
    },
    {
      "epoch": 2.0792943361188487,
      "grad_norm": 1.5147930445267097,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8852159380912781,
      "num_tokens": 389546085.0,
      "step": 11197
    },
    {
      "epoch": 2.0794800371402045,
      "grad_norm": 1.4787614853160385,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8985393047332764,
      "num_tokens": 389582542.0,
      "step": 11198
    },
    {
      "epoch": 2.0796657381615598,
      "grad_norm": 1.622372491082646,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8795002698898315,
      "num_tokens": 389615952.0,
      "step": 11199
    },
    {
      "epoch": 2.0798514391829155,
      "grad_norm": 1.6174718580802916,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.888058066368103,
      "num_tokens": 389648238.0,
      "step": 11200
    },
    {
      "epoch": 2.080037140204271,
      "grad_norm": 1.6509613788081499,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8692688345909119,
      "num_tokens": 389685454.0,
      "step": 11201
    },
    {
      "epoch": 2.080222841225627,
      "grad_norm": 1.5606438071622026,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.881054699420929,
      "num_tokens": 389724360.0,
      "step": 11202
    },
    {
      "epoch": 2.0804085422469822,
      "grad_norm": 1.5580504996886515,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8900196552276611,
      "num_tokens": 389760730.0,
      "step": 11203
    },
    {
      "epoch": 2.080594243268338,
      "grad_norm": 1.6791662498143063,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8940620422363281,
      "num_tokens": 389787476.0,
      "step": 11204
    },
    {
      "epoch": 2.0807799442896937,
      "grad_norm": 1.4676935389643693,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8810482025146484,
      "num_tokens": 389826491.0,
      "step": 11205
    },
    {
      "epoch": 2.080965645311049,
      "grad_norm": 1.60677081522874,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8875722289085388,
      "num_tokens": 389856189.0,
      "step": 11206
    },
    {
      "epoch": 2.0811513463324047,
      "grad_norm": 1.6509723644652414,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8864155411720276,
      "num_tokens": 389887141.0,
      "step": 11207
    },
    {
      "epoch": 2.0813370473537605,
      "grad_norm": 1.6861002398771407,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8855903148651123,
      "num_tokens": 389917367.0,
      "step": 11208
    },
    {
      "epoch": 2.081522748375116,
      "grad_norm": 1.6023647226359172,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8912866711616516,
      "num_tokens": 389947306.0,
      "step": 11209
    },
    {
      "epoch": 2.0817084493964715,
      "grad_norm": 1.6304880021302715,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.879982590675354,
      "num_tokens": 389981705.0,
      "step": 11210
    },
    {
      "epoch": 2.081894150417827,
      "grad_norm": 1.668134122714673,
      "learning_rate": 1e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8702914118766785,
      "num_tokens": 390017008.0,
      "step": 11211
    },
    {
      "epoch": 2.082079851439183,
      "grad_norm": 1.5045750055515792,
      "learning_rate": 1e-06,
      "loss": 0.2811,
      "mean_token_accuracy": 0.8997551798820496,
      "num_tokens": 390050394.0,
      "step": 11212
    },
    {
      "epoch": 2.0822655524605387,
      "grad_norm": 1.5064304217292943,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8910224437713623,
      "num_tokens": 390085653.0,
      "step": 11213
    },
    {
      "epoch": 2.082451253481894,
      "grad_norm": 1.505267111961256,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.886756181716919,
      "num_tokens": 390126343.0,
      "step": 11214
    },
    {
      "epoch": 2.0826369545032497,
      "grad_norm": 1.4032015065700025,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.8912285566329956,
      "num_tokens": 390164009.0,
      "step": 11215
    },
    {
      "epoch": 2.0828226555246054,
      "grad_norm": 1.5772528378859794,
      "learning_rate": 1e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8624265193939209,
      "num_tokens": 390205020.0,
      "step": 11216
    },
    {
      "epoch": 2.083008356545961,
      "grad_norm": 1.5599059062225864,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8916405439376831,
      "num_tokens": 390239958.0,
      "step": 11217
    },
    {
      "epoch": 2.0831940575673165,
      "grad_norm": 1.4294139360156148,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8908162117004395,
      "num_tokens": 390277745.0,
      "step": 11218
    },
    {
      "epoch": 2.083379758588672,
      "grad_norm": 1.7272983242995994,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8855845928192139,
      "num_tokens": 390304701.0,
      "step": 11219
    },
    {
      "epoch": 2.083565459610028,
      "grad_norm": 1.697019774907213,
      "learning_rate": 1e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8604665398597717,
      "num_tokens": 390339709.0,
      "step": 11220
    },
    {
      "epoch": 2.0837511606313837,
      "grad_norm": 1.609191960755367,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8830147981643677,
      "num_tokens": 390371716.0,
      "step": 11221
    },
    {
      "epoch": 2.083936861652739,
      "grad_norm": 1.6768533205088787,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8831630945205688,
      "num_tokens": 390403924.0,
      "step": 11222
    },
    {
      "epoch": 2.0841225626740947,
      "grad_norm": 1.5942981789108903,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8945169448852539,
      "num_tokens": 390440209.0,
      "step": 11223
    },
    {
      "epoch": 2.0843082636954504,
      "grad_norm": 1.5922218260147216,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8851123452186584,
      "num_tokens": 390472555.0,
      "step": 11224
    },
    {
      "epoch": 2.084493964716806,
      "grad_norm": 1.6404164208105225,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8837437629699707,
      "num_tokens": 390506769.0,
      "step": 11225
    },
    {
      "epoch": 2.0846796657381614,
      "grad_norm": 1.4715780494141495,
      "learning_rate": 1e-06,
      "loss": 0.2882,
      "mean_token_accuracy": 0.8978221416473389,
      "num_tokens": 390543253.0,
      "step": 11226
    },
    {
      "epoch": 2.084865366759517,
      "grad_norm": 1.450892626994805,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8808771371841431,
      "num_tokens": 390585404.0,
      "step": 11227
    },
    {
      "epoch": 2.085051067780873,
      "grad_norm": 1.6477862979561653,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8754029273986816,
      "num_tokens": 390616344.0,
      "step": 11228
    },
    {
      "epoch": 2.085236768802228,
      "grad_norm": 1.5457692409203916,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8961190581321716,
      "num_tokens": 390649601.0,
      "step": 11229
    },
    {
      "epoch": 2.085422469823584,
      "grad_norm": 1.6101160242772914,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8904813528060913,
      "num_tokens": 390684588.0,
      "step": 11230
    },
    {
      "epoch": 2.0856081708449397,
      "grad_norm": 1.7554674699742443,
      "learning_rate": 1e-06,
      "loss": 0.3026,
      "mean_token_accuracy": 0.891727089881897,
      "num_tokens": 390719836.0,
      "step": 11231
    },
    {
      "epoch": 2.0857938718662954,
      "grad_norm": 1.6598387291011947,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8895542621612549,
      "num_tokens": 390748233.0,
      "step": 11232
    },
    {
      "epoch": 2.0859795728876507,
      "grad_norm": 1.4590389768274417,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8950506448745728,
      "num_tokens": 390783410.0,
      "step": 11233
    },
    {
      "epoch": 2.0861652739090064,
      "grad_norm": 1.5369710093010966,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8841677904129028,
      "num_tokens": 390821130.0,
      "step": 11234
    },
    {
      "epoch": 2.086350974930362,
      "grad_norm": 1.6986330757465937,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.894207775592804,
      "num_tokens": 390849489.0,
      "step": 11235
    },
    {
      "epoch": 2.086536675951718,
      "grad_norm": 1.5674793428890867,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8886969685554504,
      "num_tokens": 390880016.0,
      "step": 11236
    },
    {
      "epoch": 2.086722376973073,
      "grad_norm": 1.6022902932363763,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8911213874816895,
      "num_tokens": 390914656.0,
      "step": 11237
    },
    {
      "epoch": 2.086908077994429,
      "grad_norm": 1.4418443645977954,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8942626714706421,
      "num_tokens": 390952219.0,
      "step": 11238
    },
    {
      "epoch": 2.0870937790157846,
      "grad_norm": 1.5780439979210157,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8944381475448608,
      "num_tokens": 390984649.0,
      "step": 11239
    },
    {
      "epoch": 2.0872794800371404,
      "grad_norm": 1.6771932754712848,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8909025192260742,
      "num_tokens": 391015516.0,
      "step": 11240
    },
    {
      "epoch": 2.0874651810584957,
      "grad_norm": 1.587202545258093,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8910079598426819,
      "num_tokens": 391050915.0,
      "step": 11241
    },
    {
      "epoch": 2.0876508820798514,
      "grad_norm": 1.5621759053889472,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8907184600830078,
      "num_tokens": 391087441.0,
      "step": 11242
    },
    {
      "epoch": 2.087836583101207,
      "grad_norm": 1.619707250542513,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8715179562568665,
      "num_tokens": 391122952.0,
      "step": 11243
    },
    {
      "epoch": 2.088022284122563,
      "grad_norm": 1.5881864349400994,
      "learning_rate": 1e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8691596984863281,
      "num_tokens": 391160453.0,
      "step": 11244
    },
    {
      "epoch": 2.088207985143918,
      "grad_norm": 1.5628757267877698,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.891912579536438,
      "num_tokens": 391191338.0,
      "step": 11245
    },
    {
      "epoch": 2.088393686165274,
      "grad_norm": 1.5985075528383663,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8831586837768555,
      "num_tokens": 391224321.0,
      "step": 11246
    },
    {
      "epoch": 2.0885793871866296,
      "grad_norm": 1.5370746077467465,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8860045671463013,
      "num_tokens": 391265346.0,
      "step": 11247
    },
    {
      "epoch": 2.0887650882079853,
      "grad_norm": 1.738850819849628,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8782925605773926,
      "num_tokens": 391292497.0,
      "step": 11248
    },
    {
      "epoch": 2.0889507892293406,
      "grad_norm": 1.6438881957749052,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.886624813079834,
      "num_tokens": 391326698.0,
      "step": 11249
    },
    {
      "epoch": 2.0891364902506964,
      "grad_norm": 1.761366049180747,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8981257677078247,
      "num_tokens": 391357836.0,
      "step": 11250
    },
    {
      "epoch": 2.089322191272052,
      "grad_norm": 1.573827871043064,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8921353220939636,
      "num_tokens": 391393185.0,
      "step": 11251
    },
    {
      "epoch": 2.0895078922934074,
      "grad_norm": 1.811430830437212,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8695456981658936,
      "num_tokens": 391421637.0,
      "step": 11252
    },
    {
      "epoch": 2.089693593314763,
      "grad_norm": 1.5197320049801633,
      "learning_rate": 1e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.8910530805587769,
      "num_tokens": 391455808.0,
      "step": 11253
    },
    {
      "epoch": 2.089879294336119,
      "grad_norm": 1.4456348148616578,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8891137838363647,
      "num_tokens": 391495097.0,
      "step": 11254
    },
    {
      "epoch": 2.0900649953574746,
      "grad_norm": 1.5168778549186892,
      "learning_rate": 1e-06,
      "loss": 0.2874,
      "mean_token_accuracy": 0.8973928689956665,
      "num_tokens": 391530762.0,
      "step": 11255
    },
    {
      "epoch": 2.09025069637883,
      "grad_norm": 1.6943849939726867,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8850870132446289,
      "num_tokens": 391561221.0,
      "step": 11256
    },
    {
      "epoch": 2.0904363974001856,
      "grad_norm": 1.5602082105109862,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8765108585357666,
      "num_tokens": 391598947.0,
      "step": 11257
    },
    {
      "epoch": 2.0906220984215413,
      "grad_norm": 1.6162492984275247,
      "learning_rate": 1e-06,
      "loss": 0.2914,
      "mean_token_accuracy": 0.898219883441925,
      "num_tokens": 391628199.0,
      "step": 11258
    },
    {
      "epoch": 2.090807799442897,
      "grad_norm": 1.6982388314495882,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8762636184692383,
      "num_tokens": 391662870.0,
      "step": 11259
    },
    {
      "epoch": 2.0909935004642524,
      "grad_norm": 1.4848208669363754,
      "learning_rate": 1e-06,
      "loss": 0.2926,
      "mean_token_accuracy": 0.8960400223731995,
      "num_tokens": 391700424.0,
      "step": 11260
    },
    {
      "epoch": 2.091179201485608,
      "grad_norm": 1.5601122493235635,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8843542337417603,
      "num_tokens": 391741096.0,
      "step": 11261
    },
    {
      "epoch": 2.091364902506964,
      "grad_norm": 1.5815779221615551,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.894889235496521,
      "num_tokens": 391771662.0,
      "step": 11262
    },
    {
      "epoch": 2.0915506035283196,
      "grad_norm": 1.6689839829159077,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8848503828048706,
      "num_tokens": 391802952.0,
      "step": 11263
    },
    {
      "epoch": 2.091736304549675,
      "grad_norm": 1.5634032191135938,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8713784217834473,
      "num_tokens": 391840470.0,
      "step": 11264
    },
    {
      "epoch": 2.0919220055710306,
      "grad_norm": 1.703126163618808,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8849719762802124,
      "num_tokens": 391867629.0,
      "step": 11265
    },
    {
      "epoch": 2.0921077065923863,
      "grad_norm": 1.6048668217441,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8903186917304993,
      "num_tokens": 391898786.0,
      "step": 11266
    },
    {
      "epoch": 2.092293407613742,
      "grad_norm": 1.4344911156807985,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8869150280952454,
      "num_tokens": 391942251.0,
      "step": 11267
    },
    {
      "epoch": 2.0924791086350973,
      "grad_norm": 1.7138236592003924,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8913091421127319,
      "num_tokens": 391968904.0,
      "step": 11268
    },
    {
      "epoch": 2.092664809656453,
      "grad_norm": 1.5867555526027108,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8876871466636658,
      "num_tokens": 392002699.0,
      "step": 11269
    },
    {
      "epoch": 2.092850510677809,
      "grad_norm": 1.7155903796602292,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8710744380950928,
      "num_tokens": 392037954.0,
      "step": 11270
    },
    {
      "epoch": 2.0930362116991645,
      "grad_norm": 1.5192526772579849,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8909695148468018,
      "num_tokens": 392075788.0,
      "step": 11271
    },
    {
      "epoch": 2.09322191272052,
      "grad_norm": 1.4499765943476308,
      "learning_rate": 1e-06,
      "loss": 0.2968,
      "mean_token_accuracy": 0.895256519317627,
      "num_tokens": 392115564.0,
      "step": 11272
    },
    {
      "epoch": 2.0934076137418756,
      "grad_norm": 1.6032640398812332,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8808181285858154,
      "num_tokens": 392150173.0,
      "step": 11273
    },
    {
      "epoch": 2.0935933147632313,
      "grad_norm": 1.5794186612372738,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8902915716171265,
      "num_tokens": 392185804.0,
      "step": 11274
    },
    {
      "epoch": 2.093779015784587,
      "grad_norm": 1.470327382906036,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8872849941253662,
      "num_tokens": 392224688.0,
      "step": 11275
    },
    {
      "epoch": 2.0939647168059423,
      "grad_norm": 1.887726613032776,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8859889507293701,
      "num_tokens": 392251580.0,
      "step": 11276
    },
    {
      "epoch": 2.094150417827298,
      "grad_norm": 1.8442793551752517,
      "learning_rate": 1e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8651387691497803,
      "num_tokens": 392278952.0,
      "step": 11277
    },
    {
      "epoch": 2.094336118848654,
      "grad_norm": 1.5434598175819787,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8733559846878052,
      "num_tokens": 392317455.0,
      "step": 11278
    },
    {
      "epoch": 2.094521819870009,
      "grad_norm": 1.7111037773997055,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8871732950210571,
      "num_tokens": 392345094.0,
      "step": 11279
    },
    {
      "epoch": 2.094707520891365,
      "grad_norm": 1.6199531338932602,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8844757676124573,
      "num_tokens": 392379965.0,
      "step": 11280
    },
    {
      "epoch": 2.0948932219127205,
      "grad_norm": 1.594669610927039,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8831643462181091,
      "num_tokens": 392411524.0,
      "step": 11281
    },
    {
      "epoch": 2.0950789229340763,
      "grad_norm": 1.591876421504666,
      "learning_rate": 1e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8699997663497925,
      "num_tokens": 392447544.0,
      "step": 11282
    },
    {
      "epoch": 2.0952646239554316,
      "grad_norm": 1.5411405103712301,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8835562467575073,
      "num_tokens": 392482504.0,
      "step": 11283
    },
    {
      "epoch": 2.0954503249767873,
      "grad_norm": 1.5927966154359465,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.884289562702179,
      "num_tokens": 392516777.0,
      "step": 11284
    },
    {
      "epoch": 2.095636025998143,
      "grad_norm": 1.547959971137457,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8895434141159058,
      "num_tokens": 392549045.0,
      "step": 11285
    },
    {
      "epoch": 2.0958217270194988,
      "grad_norm": 1.622261379633913,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8903293609619141,
      "num_tokens": 392580641.0,
      "step": 11286
    },
    {
      "epoch": 2.096007428040854,
      "grad_norm": 1.4658286087405548,
      "learning_rate": 1e-06,
      "loss": 0.2968,
      "mean_token_accuracy": 0.894253134727478,
      "num_tokens": 392618312.0,
      "step": 11287
    },
    {
      "epoch": 2.09619312906221,
      "grad_norm": 1.6589664829560933,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8861052393913269,
      "num_tokens": 392652429.0,
      "step": 11288
    },
    {
      "epoch": 2.0963788300835655,
      "grad_norm": 1.5903929308378082,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8731796145439148,
      "num_tokens": 392687734.0,
      "step": 11289
    },
    {
      "epoch": 2.0965645311049212,
      "grad_norm": 1.5375744764454344,
      "learning_rate": 1e-06,
      "loss": 0.27,
      "mean_token_accuracy": 0.903832197189331,
      "num_tokens": 392720248.0,
      "step": 11290
    },
    {
      "epoch": 2.0967502321262765,
      "grad_norm": 1.839642716335146,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8800482749938965,
      "num_tokens": 392753562.0,
      "step": 11291
    },
    {
      "epoch": 2.0969359331476323,
      "grad_norm": 1.7693770397856674,
      "learning_rate": 1e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8626722097396851,
      "num_tokens": 392787444.0,
      "step": 11292
    },
    {
      "epoch": 2.097121634168988,
      "grad_norm": 1.5302542938368175,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8899612426757812,
      "num_tokens": 392822467.0,
      "step": 11293
    },
    {
      "epoch": 2.0973073351903437,
      "grad_norm": 1.56224363588611,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.889290452003479,
      "num_tokens": 392856278.0,
      "step": 11294
    },
    {
      "epoch": 2.097493036211699,
      "grad_norm": 1.460646976998655,
      "learning_rate": 1e-06,
      "loss": 0.2684,
      "mean_token_accuracy": 0.905386209487915,
      "num_tokens": 392892767.0,
      "step": 11295
    },
    {
      "epoch": 2.0976787372330548,
      "grad_norm": 1.4846164066903182,
      "learning_rate": 1e-06,
      "loss": 0.2815,
      "mean_token_accuracy": 0.8987468481063843,
      "num_tokens": 392928664.0,
      "step": 11296
    },
    {
      "epoch": 2.0978644382544105,
      "grad_norm": 1.882433907994584,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8726622462272644,
      "num_tokens": 392957309.0,
      "step": 11297
    },
    {
      "epoch": 2.0980501392757662,
      "grad_norm": 1.8891014783273536,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8804100751876831,
      "num_tokens": 392987907.0,
      "step": 11298
    },
    {
      "epoch": 2.0982358402971215,
      "grad_norm": 1.8038000067599946,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8821624517440796,
      "num_tokens": 393019036.0,
      "step": 11299
    },
    {
      "epoch": 2.0984215413184772,
      "grad_norm": 1.680160357005945,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8865957260131836,
      "num_tokens": 393053526.0,
      "step": 11300
    },
    {
      "epoch": 2.098607242339833,
      "grad_norm": 1.5868955627963661,
      "learning_rate": 1e-06,
      "loss": 0.2967,
      "mean_token_accuracy": 0.8967866897583008,
      "num_tokens": 393086943.0,
      "step": 11301
    },
    {
      "epoch": 2.0987929433611887,
      "grad_norm": 1.6930181302419547,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8838692307472229,
      "num_tokens": 393115422.0,
      "step": 11302
    },
    {
      "epoch": 2.098978644382544,
      "grad_norm": 1.5104528331156768,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8890107870101929,
      "num_tokens": 393152160.0,
      "step": 11303
    },
    {
      "epoch": 2.0991643454038997,
      "grad_norm": 1.7774613412272318,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8759516477584839,
      "num_tokens": 393184333.0,
      "step": 11304
    },
    {
      "epoch": 2.0993500464252555,
      "grad_norm": 1.6394866795434069,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.88319993019104,
      "num_tokens": 393218559.0,
      "step": 11305
    },
    {
      "epoch": 2.0995357474466108,
      "grad_norm": 1.7251797557294999,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8850570321083069,
      "num_tokens": 393248497.0,
      "step": 11306
    },
    {
      "epoch": 2.0997214484679665,
      "grad_norm": 1.5495523902392716,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8971492648124695,
      "num_tokens": 393281258.0,
      "step": 11307
    },
    {
      "epoch": 2.0999071494893222,
      "grad_norm": 1.4478005918301373,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8862480521202087,
      "num_tokens": 393323966.0,
      "step": 11308
    },
    {
      "epoch": 2.100092850510678,
      "grad_norm": 1.539926605404602,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8784273862838745,
      "num_tokens": 393361868.0,
      "step": 11309
    },
    {
      "epoch": 2.1002785515320332,
      "grad_norm": 1.6399908183276957,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8819188475608826,
      "num_tokens": 393395136.0,
      "step": 11310
    },
    {
      "epoch": 2.100464252553389,
      "grad_norm": 1.6075982778055375,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8866942524909973,
      "num_tokens": 393426478.0,
      "step": 11311
    },
    {
      "epoch": 2.1006499535747447,
      "grad_norm": 1.5000653295266912,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8778635859489441,
      "num_tokens": 393466608.0,
      "step": 11312
    },
    {
      "epoch": 2.1008356545961004,
      "grad_norm": 1.4586482622926564,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8864148855209351,
      "num_tokens": 393507796.0,
      "step": 11313
    },
    {
      "epoch": 2.1010213556174557,
      "grad_norm": 1.6472584838195248,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.892545223236084,
      "num_tokens": 393537667.0,
      "step": 11314
    },
    {
      "epoch": 2.1012070566388115,
      "grad_norm": 1.5526581214773658,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8910738229751587,
      "num_tokens": 393573931.0,
      "step": 11315
    },
    {
      "epoch": 2.101392757660167,
      "grad_norm": 1.7276342095635797,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8974388837814331,
      "num_tokens": 393603668.0,
      "step": 11316
    },
    {
      "epoch": 2.101578458681523,
      "grad_norm": 1.721959439058018,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8933840990066528,
      "num_tokens": 393632577.0,
      "step": 11317
    },
    {
      "epoch": 2.101764159702878,
      "grad_norm": 1.5629481363404707,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.8720207810401917,
      "num_tokens": 393669983.0,
      "step": 11318
    },
    {
      "epoch": 2.101949860724234,
      "grad_norm": 1.5232011741215061,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8831406235694885,
      "num_tokens": 393707462.0,
      "step": 11319
    },
    {
      "epoch": 2.1021355617455897,
      "grad_norm": 1.492547933681137,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8911651372909546,
      "num_tokens": 393744276.0,
      "step": 11320
    },
    {
      "epoch": 2.1023212627669454,
      "grad_norm": 1.578024099211028,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8870871663093567,
      "num_tokens": 393779503.0,
      "step": 11321
    },
    {
      "epoch": 2.1025069637883007,
      "grad_norm": 1.5875494959159457,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8816314935684204,
      "num_tokens": 393816411.0,
      "step": 11322
    },
    {
      "epoch": 2.1026926648096564,
      "grad_norm": 1.6722510720419974,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8838151693344116,
      "num_tokens": 393845988.0,
      "step": 11323
    },
    {
      "epoch": 2.102878365831012,
      "grad_norm": 1.567422604736922,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8879363536834717,
      "num_tokens": 393884838.0,
      "step": 11324
    },
    {
      "epoch": 2.103064066852368,
      "grad_norm": 1.6419616578525025,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8732539415359497,
      "num_tokens": 393918478.0,
      "step": 11325
    },
    {
      "epoch": 2.103249767873723,
      "grad_norm": 1.533489871459747,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8877554535865784,
      "num_tokens": 393953903.0,
      "step": 11326
    },
    {
      "epoch": 2.103435468895079,
      "grad_norm": 1.5382962915631035,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8892903327941895,
      "num_tokens": 393990158.0,
      "step": 11327
    },
    {
      "epoch": 2.1036211699164347,
      "grad_norm": 1.5756521263818484,
      "learning_rate": 1e-06,
      "loss": 0.3002,
      "mean_token_accuracy": 0.8928707838058472,
      "num_tokens": 394022135.0,
      "step": 11328
    },
    {
      "epoch": 2.10380687093779,
      "grad_norm": 1.5783588911155149,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8848286867141724,
      "num_tokens": 394055434.0,
      "step": 11329
    },
    {
      "epoch": 2.1039925719591457,
      "grad_norm": 1.609538241760226,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.8916876316070557,
      "num_tokens": 394085499.0,
      "step": 11330
    },
    {
      "epoch": 2.1041782729805014,
      "grad_norm": 1.397313381233551,
      "learning_rate": 1e-06,
      "loss": 0.2842,
      "mean_token_accuracy": 0.8972463011741638,
      "num_tokens": 394124351.0,
      "step": 11331
    },
    {
      "epoch": 2.104363974001857,
      "grad_norm": 1.5580836326418657,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8732954263687134,
      "num_tokens": 394162757.0,
      "step": 11332
    },
    {
      "epoch": 2.1045496750232124,
      "grad_norm": 1.6438537546701013,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8915886878967285,
      "num_tokens": 394190665.0,
      "step": 11333
    },
    {
      "epoch": 2.104735376044568,
      "grad_norm": 1.5053428207841053,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8879014253616333,
      "num_tokens": 394228037.0,
      "step": 11334
    },
    {
      "epoch": 2.104921077065924,
      "grad_norm": 1.5252511357034768,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8865510821342468,
      "num_tokens": 394261025.0,
      "step": 11335
    },
    {
      "epoch": 2.1051067780872796,
      "grad_norm": 1.5906580324896689,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8672332167625427,
      "num_tokens": 394298661.0,
      "step": 11336
    },
    {
      "epoch": 2.105292479108635,
      "grad_norm": 1.5689115126171809,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8827621936798096,
      "num_tokens": 394335325.0,
      "step": 11337
    },
    {
      "epoch": 2.1054781801299907,
      "grad_norm": 1.5787630516336695,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.891690731048584,
      "num_tokens": 394366266.0,
      "step": 11338
    },
    {
      "epoch": 2.1056638811513464,
      "grad_norm": 1.5793461435294343,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.884966254234314,
      "num_tokens": 394403977.0,
      "step": 11339
    },
    {
      "epoch": 2.105849582172702,
      "grad_norm": 1.700932086290094,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8973863124847412,
      "num_tokens": 394436341.0,
      "step": 11340
    },
    {
      "epoch": 2.1060352831940574,
      "grad_norm": 1.602569593491283,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8923097252845764,
      "num_tokens": 394469167.0,
      "step": 11341
    },
    {
      "epoch": 2.106220984215413,
      "grad_norm": 1.5083192782906378,
      "learning_rate": 1e-06,
      "loss": 0.2715,
      "mean_token_accuracy": 0.9035109281539917,
      "num_tokens": 394502075.0,
      "step": 11342
    },
    {
      "epoch": 2.106406685236769,
      "grad_norm": 1.7313480897789397,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8909786939620972,
      "num_tokens": 394532485.0,
      "step": 11343
    },
    {
      "epoch": 2.1065923862581246,
      "grad_norm": 1.6682702178778002,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8847110271453857,
      "num_tokens": 394565098.0,
      "step": 11344
    },
    {
      "epoch": 2.10677808727948,
      "grad_norm": 1.5904474001485398,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.8933748602867126,
      "num_tokens": 394596084.0,
      "step": 11345
    },
    {
      "epoch": 2.1069637883008356,
      "grad_norm": 1.4084634575804431,
      "learning_rate": 1e-06,
      "loss": 0.2829,
      "mean_token_accuracy": 0.9018688797950745,
      "num_tokens": 394633855.0,
      "step": 11346
    },
    {
      "epoch": 2.1071494893221914,
      "grad_norm": 1.567597077843197,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8896483182907104,
      "num_tokens": 394668230.0,
      "step": 11347
    },
    {
      "epoch": 2.107335190343547,
      "grad_norm": 1.9102077214925786,
      "learning_rate": 1e-06,
      "loss": 0.2778,
      "mean_token_accuracy": 0.8978036642074585,
      "num_tokens": 394704504.0,
      "step": 11348
    },
    {
      "epoch": 2.1075208913649024,
      "grad_norm": 1.605051192403684,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8886370658874512,
      "num_tokens": 394738082.0,
      "step": 11349
    },
    {
      "epoch": 2.107706592386258,
      "grad_norm": 1.6152735770825075,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8883312344551086,
      "num_tokens": 394768736.0,
      "step": 11350
    },
    {
      "epoch": 2.107892293407614,
      "grad_norm": 1.6743468274868394,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8891470432281494,
      "num_tokens": 394799931.0,
      "step": 11351
    },
    {
      "epoch": 2.108077994428969,
      "grad_norm": 1.5966062732332045,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.881294846534729,
      "num_tokens": 394835772.0,
      "step": 11352
    },
    {
      "epoch": 2.108263695450325,
      "grad_norm": 1.6629718649491299,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8785881996154785,
      "num_tokens": 394868141.0,
      "step": 11353
    },
    {
      "epoch": 2.1084493964716806,
      "grad_norm": 1.516361575026346,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.891018271446228,
      "num_tokens": 394903176.0,
      "step": 11354
    },
    {
      "epoch": 2.1086350974930363,
      "grad_norm": 1.5612458842643524,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8811965584754944,
      "num_tokens": 394941083.0,
      "step": 11355
    },
    {
      "epoch": 2.1088207985143916,
      "grad_norm": 1.6089895664145863,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8822242021560669,
      "num_tokens": 394974333.0,
      "step": 11356
    },
    {
      "epoch": 2.1090064995357474,
      "grad_norm": 1.5457992565510374,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.894251823425293,
      "num_tokens": 395009197.0,
      "step": 11357
    },
    {
      "epoch": 2.109192200557103,
      "grad_norm": 1.541342157361898,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8855985403060913,
      "num_tokens": 395041712.0,
      "step": 11358
    },
    {
      "epoch": 2.109377901578459,
      "grad_norm": 1.4821870177394931,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.888291597366333,
      "num_tokens": 395079042.0,
      "step": 11359
    },
    {
      "epoch": 2.109563602599814,
      "grad_norm": 1.516010169745181,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8881130218505859,
      "num_tokens": 395115401.0,
      "step": 11360
    },
    {
      "epoch": 2.10974930362117,
      "grad_norm": 1.6980404782505916,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8810359835624695,
      "num_tokens": 395146799.0,
      "step": 11361
    },
    {
      "epoch": 2.1099350046425256,
      "grad_norm": 1.6885147588977167,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8833366632461548,
      "num_tokens": 395178830.0,
      "step": 11362
    },
    {
      "epoch": 2.1101207056638813,
      "grad_norm": 1.6015506355780398,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8884044289588928,
      "num_tokens": 395209339.0,
      "step": 11363
    },
    {
      "epoch": 2.1103064066852366,
      "grad_norm": 1.5353161184871909,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8952537775039673,
      "num_tokens": 395242388.0,
      "step": 11364
    },
    {
      "epoch": 2.1104921077065923,
      "grad_norm": 1.562582573867797,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8869892358779907,
      "num_tokens": 395277199.0,
      "step": 11365
    },
    {
      "epoch": 2.110677808727948,
      "grad_norm": 1.5484221838625838,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8827095031738281,
      "num_tokens": 395312646.0,
      "step": 11366
    },
    {
      "epoch": 2.110863509749304,
      "grad_norm": 1.4660991305064992,
      "learning_rate": 1e-06,
      "loss": 0.2524,
      "mean_token_accuracy": 0.9076142311096191,
      "num_tokens": 395345353.0,
      "step": 11367
    },
    {
      "epoch": 2.111049210770659,
      "grad_norm": 1.4114115244618473,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8963916897773743,
      "num_tokens": 395382950.0,
      "step": 11368
    },
    {
      "epoch": 2.111234911792015,
      "grad_norm": 1.585378611931461,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8833744525909424,
      "num_tokens": 395416472.0,
      "step": 11369
    },
    {
      "epoch": 2.1114206128133706,
      "grad_norm": 1.4237050530768212,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.882148027420044,
      "num_tokens": 395457959.0,
      "step": 11370
    },
    {
      "epoch": 2.1116063138347263,
      "grad_norm": 1.6184894222240567,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8877600431442261,
      "num_tokens": 395492860.0,
      "step": 11371
    },
    {
      "epoch": 2.1117920148560816,
      "grad_norm": 1.5241656513703106,
      "learning_rate": 1e-06,
      "loss": 0.3222,
      "mean_token_accuracy": 0.8848921656608582,
      "num_tokens": 395529661.0,
      "step": 11372
    },
    {
      "epoch": 2.1119777158774373,
      "grad_norm": 1.6694653526108727,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8818898797035217,
      "num_tokens": 395562172.0,
      "step": 11373
    },
    {
      "epoch": 2.112163416898793,
      "grad_norm": 1.5926639139976222,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8719010949134827,
      "num_tokens": 395597007.0,
      "step": 11374
    },
    {
      "epoch": 2.1123491179201483,
      "grad_norm": 1.4865615667007168,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8939110040664673,
      "num_tokens": 395635966.0,
      "step": 11375
    },
    {
      "epoch": 2.112534818941504,
      "grad_norm": 1.4969262095840838,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8901830315589905,
      "num_tokens": 395675911.0,
      "step": 11376
    },
    {
      "epoch": 2.11272051996286,
      "grad_norm": 1.751836732360701,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8865193128585815,
      "num_tokens": 395703566.0,
      "step": 11377
    },
    {
      "epoch": 2.1129062209842155,
      "grad_norm": 1.6354558013685927,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8844641447067261,
      "num_tokens": 395737438.0,
      "step": 11378
    },
    {
      "epoch": 2.113091922005571,
      "grad_norm": 1.5918549608347121,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8896838426589966,
      "num_tokens": 395770759.0,
      "step": 11379
    },
    {
      "epoch": 2.1132776230269266,
      "grad_norm": 1.4624210650134661,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8754551410675049,
      "num_tokens": 395813962.0,
      "step": 11380
    },
    {
      "epoch": 2.1134633240482823,
      "grad_norm": 1.5390517250871405,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8760933876037598,
      "num_tokens": 395853362.0,
      "step": 11381
    },
    {
      "epoch": 2.113649025069638,
      "grad_norm": 1.554761348744996,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8910724520683289,
      "num_tokens": 395886714.0,
      "step": 11382
    },
    {
      "epoch": 2.1138347260909933,
      "grad_norm": 1.64235244044431,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.8942782878875732,
      "num_tokens": 395915526.0,
      "step": 11383
    },
    {
      "epoch": 2.114020427112349,
      "grad_norm": 1.6448789010272213,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8775637745857239,
      "num_tokens": 395949589.0,
      "step": 11384
    },
    {
      "epoch": 2.114206128133705,
      "grad_norm": 1.6213171280554008,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8805510997772217,
      "num_tokens": 395984050.0,
      "step": 11385
    },
    {
      "epoch": 2.1143918291550605,
      "grad_norm": 1.5186784151672688,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8954545259475708,
      "num_tokens": 396019226.0,
      "step": 11386
    },
    {
      "epoch": 2.114577530176416,
      "grad_norm": 1.4245477525524806,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8955087661743164,
      "num_tokens": 396056906.0,
      "step": 11387
    },
    {
      "epoch": 2.1147632311977715,
      "grad_norm": 1.5059154554872534,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8802580833435059,
      "num_tokens": 396096303.0,
      "step": 11388
    },
    {
      "epoch": 2.1149489322191273,
      "grad_norm": 1.650280592859016,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.8731216192245483,
      "num_tokens": 396132016.0,
      "step": 11389
    },
    {
      "epoch": 2.115134633240483,
      "grad_norm": 1.530604787264653,
      "learning_rate": 1e-06,
      "loss": 0.2858,
      "mean_token_accuracy": 0.9004186987876892,
      "num_tokens": 396164507.0,
      "step": 11390
    },
    {
      "epoch": 2.1153203342618383,
      "grad_norm": 1.782864227138797,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8672912120819092,
      "num_tokens": 396198328.0,
      "step": 11391
    },
    {
      "epoch": 2.115506035283194,
      "grad_norm": 1.7019239698492656,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8965833783149719,
      "num_tokens": 396224894.0,
      "step": 11392
    },
    {
      "epoch": 2.1156917363045498,
      "grad_norm": 1.5856356930390396,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.882965087890625,
      "num_tokens": 396257423.0,
      "step": 11393
    },
    {
      "epoch": 2.1158774373259055,
      "grad_norm": 1.4313628875600823,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8899790048599243,
      "num_tokens": 396296120.0,
      "step": 11394
    },
    {
      "epoch": 2.116063138347261,
      "grad_norm": 1.4116383963690353,
      "learning_rate": 1e-06,
      "loss": 0.2856,
      "mean_token_accuracy": 0.896686315536499,
      "num_tokens": 396335348.0,
      "step": 11395
    },
    {
      "epoch": 2.1162488393686165,
      "grad_norm": 1.4930511399089654,
      "learning_rate": 1e-06,
      "loss": 0.2688,
      "mean_token_accuracy": 0.904861330986023,
      "num_tokens": 396368264.0,
      "step": 11396
    },
    {
      "epoch": 2.1164345403899723,
      "grad_norm": 1.5724762135539034,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8797975778579712,
      "num_tokens": 396402749.0,
      "step": 11397
    },
    {
      "epoch": 2.1166202414113275,
      "grad_norm": 1.5201488338752869,
      "learning_rate": 1e-06,
      "loss": 0.2974,
      "mean_token_accuracy": 0.893412709236145,
      "num_tokens": 396435020.0,
      "step": 11398
    },
    {
      "epoch": 2.1168059424326833,
      "grad_norm": 1.563473012597553,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8795549869537354,
      "num_tokens": 396471446.0,
      "step": 11399
    },
    {
      "epoch": 2.116991643454039,
      "grad_norm": 1.542284147134835,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.8971513509750366,
      "num_tokens": 396503492.0,
      "step": 11400
    },
    {
      "epoch": 2.1171773444753947,
      "grad_norm": 1.4701491075218942,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8861010074615479,
      "num_tokens": 396542274.0,
      "step": 11401
    },
    {
      "epoch": 2.11736304549675,
      "grad_norm": 1.5256988108114118,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8951398730278015,
      "num_tokens": 396575506.0,
      "step": 11402
    },
    {
      "epoch": 2.1175487465181058,
      "grad_norm": 1.5206191364907145,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8907685279846191,
      "num_tokens": 396614249.0,
      "step": 11403
    },
    {
      "epoch": 2.1177344475394615,
      "grad_norm": 1.597420285815399,
      "learning_rate": 1e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8704847097396851,
      "num_tokens": 396651022.0,
      "step": 11404
    },
    {
      "epoch": 2.1179201485608172,
      "grad_norm": 1.6871230844039145,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8894623517990112,
      "num_tokens": 396683504.0,
      "step": 11405
    },
    {
      "epoch": 2.1181058495821725,
      "grad_norm": 1.7323148495714606,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8763567805290222,
      "num_tokens": 396712093.0,
      "step": 11406
    },
    {
      "epoch": 2.1182915506035283,
      "grad_norm": 1.5503065399989917,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8874073624610901,
      "num_tokens": 396746030.0,
      "step": 11407
    },
    {
      "epoch": 2.118477251624884,
      "grad_norm": 1.3930208731739806,
      "learning_rate": 1e-06,
      "loss": 0.2887,
      "mean_token_accuracy": 0.8969777822494507,
      "num_tokens": 396786150.0,
      "step": 11408
    },
    {
      "epoch": 2.1186629526462397,
      "grad_norm": 1.5578075306083439,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.895300030708313,
      "num_tokens": 396817296.0,
      "step": 11409
    },
    {
      "epoch": 2.118848653667595,
      "grad_norm": 1.5634975341530564,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8860617876052856,
      "num_tokens": 396853892.0,
      "step": 11410
    },
    {
      "epoch": 2.1190343546889507,
      "grad_norm": 1.9498547155952415,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8738665580749512,
      "num_tokens": 396881562.0,
      "step": 11411
    },
    {
      "epoch": 2.1192200557103065,
      "grad_norm": 1.5566654278317202,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8868016600608826,
      "num_tokens": 396917758.0,
      "step": 11412
    },
    {
      "epoch": 2.119405756731662,
      "grad_norm": 1.4979732626573958,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8854840397834778,
      "num_tokens": 396955970.0,
      "step": 11413
    },
    {
      "epoch": 2.1195914577530175,
      "grad_norm": 1.7401949090234612,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8777415752410889,
      "num_tokens": 396989503.0,
      "step": 11414
    },
    {
      "epoch": 2.1197771587743732,
      "grad_norm": 1.6029737266116157,
      "learning_rate": 1e-06,
      "loss": 0.2875,
      "mean_token_accuracy": 0.8972123265266418,
      "num_tokens": 397020270.0,
      "step": 11415
    },
    {
      "epoch": 2.119962859795729,
      "grad_norm": 1.6905778338125648,
      "learning_rate": 1e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8632352352142334,
      "num_tokens": 397053965.0,
      "step": 11416
    },
    {
      "epoch": 2.1201485608170847,
      "grad_norm": 1.5535545480547561,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8815399408340454,
      "num_tokens": 397088572.0,
      "step": 11417
    },
    {
      "epoch": 2.12033426183844,
      "grad_norm": 1.671903856690283,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8837289214134216,
      "num_tokens": 397117614.0,
      "step": 11418
    },
    {
      "epoch": 2.1205199628597957,
      "grad_norm": 1.6174577110255943,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8876896500587463,
      "num_tokens": 397150214.0,
      "step": 11419
    },
    {
      "epoch": 2.1207056638811514,
      "grad_norm": 1.490402870407152,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8721193075180054,
      "num_tokens": 397190329.0,
      "step": 11420
    },
    {
      "epoch": 2.1208913649025067,
      "grad_norm": 1.450355215608712,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8948888778686523,
      "num_tokens": 397228923.0,
      "step": 11421
    },
    {
      "epoch": 2.1210770659238625,
      "grad_norm": 1.6682893547634383,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8857311010360718,
      "num_tokens": 397261776.0,
      "step": 11422
    },
    {
      "epoch": 2.121262766945218,
      "grad_norm": 1.581713475689211,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8878281116485596,
      "num_tokens": 397294375.0,
      "step": 11423
    },
    {
      "epoch": 2.121448467966574,
      "grad_norm": 1.7359989171819514,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8890295028686523,
      "num_tokens": 397322158.0,
      "step": 11424
    },
    {
      "epoch": 2.1216341689879292,
      "grad_norm": 1.500519191764243,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8803312182426453,
      "num_tokens": 397362137.0,
      "step": 11425
    },
    {
      "epoch": 2.121819870009285,
      "grad_norm": 1.6043895979642775,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8884356617927551,
      "num_tokens": 397397858.0,
      "step": 11426
    },
    {
      "epoch": 2.1220055710306407,
      "grad_norm": 1.6959542466233577,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8761716485023499,
      "num_tokens": 397433623.0,
      "step": 11427
    },
    {
      "epoch": 2.1221912720519964,
      "grad_norm": 1.4955169440707725,
      "learning_rate": 1e-06,
      "loss": 0.291,
      "mean_token_accuracy": 0.8952832221984863,
      "num_tokens": 397472503.0,
      "step": 11428
    },
    {
      "epoch": 2.1223769730733517,
      "grad_norm": 1.6629967406712918,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.877016007900238,
      "num_tokens": 397506699.0,
      "step": 11429
    },
    {
      "epoch": 2.1225626740947074,
      "grad_norm": 1.6881135075154812,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8829202055931091,
      "num_tokens": 397538029.0,
      "step": 11430
    },
    {
      "epoch": 2.122748375116063,
      "grad_norm": 1.459490260391327,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8844114542007446,
      "num_tokens": 397576505.0,
      "step": 11431
    },
    {
      "epoch": 2.122934076137419,
      "grad_norm": 1.5055253710741647,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8806679248809814,
      "num_tokens": 397614359.0,
      "step": 11432
    },
    {
      "epoch": 2.123119777158774,
      "grad_norm": 1.6134957194321928,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8767760396003723,
      "num_tokens": 397650996.0,
      "step": 11433
    },
    {
      "epoch": 2.12330547818013,
      "grad_norm": 1.5227285391429037,
      "learning_rate": 1e-06,
      "loss": 0.3668,
      "mean_token_accuracy": 0.8786466121673584,
      "num_tokens": 397689344.0,
      "step": 11434
    },
    {
      "epoch": 2.1234911792014857,
      "grad_norm": 1.4718511248651842,
      "learning_rate": 1e-06,
      "loss": 0.274,
      "mean_token_accuracy": 0.9029051661491394,
      "num_tokens": 397725484.0,
      "step": 11435
    },
    {
      "epoch": 2.1236768802228414,
      "grad_norm": 1.6519942072250653,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8923534750938416,
      "num_tokens": 397753713.0,
      "step": 11436
    },
    {
      "epoch": 2.1238625812441967,
      "grad_norm": 1.6883859288957639,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8753782510757446,
      "num_tokens": 397785396.0,
      "step": 11437
    },
    {
      "epoch": 2.1240482822655524,
      "grad_norm": 1.4787376755944304,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8889061212539673,
      "num_tokens": 397821464.0,
      "step": 11438
    },
    {
      "epoch": 2.124233983286908,
      "grad_norm": 1.5541323983434474,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.8913689851760864,
      "num_tokens": 397857568.0,
      "step": 11439
    },
    {
      "epoch": 2.124419684308264,
      "grad_norm": 1.7109143876925303,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.880041241645813,
      "num_tokens": 397887904.0,
      "step": 11440
    },
    {
      "epoch": 2.124605385329619,
      "grad_norm": 1.6266185048833492,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8833796381950378,
      "num_tokens": 397921356.0,
      "step": 11441
    },
    {
      "epoch": 2.124791086350975,
      "grad_norm": 1.6304490018615827,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8712633848190308,
      "num_tokens": 397957909.0,
      "step": 11442
    },
    {
      "epoch": 2.1249767873723306,
      "grad_norm": 1.5004571717925355,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8944346904754639,
      "num_tokens": 397994266.0,
      "step": 11443
    },
    {
      "epoch": 2.125162488393686,
      "grad_norm": 1.7031559154204274,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8915684819221497,
      "num_tokens": 398025352.0,
      "step": 11444
    },
    {
      "epoch": 2.1253481894150417,
      "grad_norm": 1.583452122120727,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8905548453330994,
      "num_tokens": 398058439.0,
      "step": 11445
    },
    {
      "epoch": 2.1255338904363974,
      "grad_norm": 1.6846646629155246,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8802652359008789,
      "num_tokens": 398090229.0,
      "step": 11446
    },
    {
      "epoch": 2.125719591457753,
      "grad_norm": 1.6031822282551649,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8900101184844971,
      "num_tokens": 398123080.0,
      "step": 11447
    },
    {
      "epoch": 2.125905292479109,
      "grad_norm": 1.5737139184779536,
      "learning_rate": 1e-06,
      "loss": 0.3102,
      "mean_token_accuracy": 0.8889895677566528,
      "num_tokens": 398160411.0,
      "step": 11448
    },
    {
      "epoch": 2.126090993500464,
      "grad_norm": 1.6264903730352254,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8921726942062378,
      "num_tokens": 398192348.0,
      "step": 11449
    },
    {
      "epoch": 2.12627669452182,
      "grad_norm": 1.6208787277936496,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.886330246925354,
      "num_tokens": 398228732.0,
      "step": 11450
    },
    {
      "epoch": 2.1264623955431756,
      "grad_norm": 1.5885610397683163,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8940691947937012,
      "num_tokens": 398258457.0,
      "step": 11451
    },
    {
      "epoch": 2.126648096564531,
      "grad_norm": 1.533178015224458,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.889552116394043,
      "num_tokens": 398292771.0,
      "step": 11452
    },
    {
      "epoch": 2.1268337975858866,
      "grad_norm": 1.7088517750344718,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8712875247001648,
      "num_tokens": 398328727.0,
      "step": 11453
    },
    {
      "epoch": 2.1270194986072424,
      "grad_norm": 1.7064388207936196,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8898837566375732,
      "num_tokens": 398359437.0,
      "step": 11454
    },
    {
      "epoch": 2.127205199628598,
      "grad_norm": 1.7393372017718918,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8788362741470337,
      "num_tokens": 398388600.0,
      "step": 11455
    },
    {
      "epoch": 2.1273909006499534,
      "grad_norm": 1.5531339093246912,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8710723519325256,
      "num_tokens": 398427602.0,
      "step": 11456
    },
    {
      "epoch": 2.127576601671309,
      "grad_norm": 1.5044026371413037,
      "learning_rate": 1e-06,
      "loss": 0.2754,
      "mean_token_accuracy": 0.9013375639915466,
      "num_tokens": 398459661.0,
      "step": 11457
    },
    {
      "epoch": 2.127762302692665,
      "grad_norm": 1.483126326661225,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8818120956420898,
      "num_tokens": 398497913.0,
      "step": 11458
    },
    {
      "epoch": 2.1279480037140206,
      "grad_norm": 1.5379565433044278,
      "learning_rate": 1e-06,
      "loss": 0.2928,
      "mean_token_accuracy": 0.8983028531074524,
      "num_tokens": 398533888.0,
      "step": 11459
    },
    {
      "epoch": 2.128133704735376,
      "grad_norm": 1.578815811094074,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8856598734855652,
      "num_tokens": 398564375.0,
      "step": 11460
    },
    {
      "epoch": 2.1283194057567316,
      "grad_norm": 1.5236061280770443,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8877061605453491,
      "num_tokens": 398604397.0,
      "step": 11461
    },
    {
      "epoch": 2.1285051067780874,
      "grad_norm": 1.7054428866799476,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8889704346656799,
      "num_tokens": 398633592.0,
      "step": 11462
    },
    {
      "epoch": 2.128690807799443,
      "grad_norm": 1.5158303970998743,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8775286078453064,
      "num_tokens": 398678713.0,
      "step": 11463
    },
    {
      "epoch": 2.1288765088207984,
      "grad_norm": 1.5769782154725729,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.873188853263855,
      "num_tokens": 398715239.0,
      "step": 11464
    },
    {
      "epoch": 2.129062209842154,
      "grad_norm": 1.6260658730732147,
      "learning_rate": 1e-06,
      "loss": 0.2901,
      "mean_token_accuracy": 0.898347795009613,
      "num_tokens": 398744868.0,
      "step": 11465
    },
    {
      "epoch": 2.12924791086351,
      "grad_norm": 1.4142279407607947,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8892614245414734,
      "num_tokens": 398782773.0,
      "step": 11466
    },
    {
      "epoch": 2.1294336118848656,
      "grad_norm": 1.6424082791796086,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8872988224029541,
      "num_tokens": 398815155.0,
      "step": 11467
    },
    {
      "epoch": 2.129619312906221,
      "grad_norm": 1.5646919293438408,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8915968537330627,
      "num_tokens": 398848072.0,
      "step": 11468
    },
    {
      "epoch": 2.1298050139275766,
      "grad_norm": 1.4693136646802285,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8869615793228149,
      "num_tokens": 398885250.0,
      "step": 11469
    },
    {
      "epoch": 2.1299907149489323,
      "grad_norm": 1.7008328327760447,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8868783712387085,
      "num_tokens": 398922269.0,
      "step": 11470
    },
    {
      "epoch": 2.130176415970288,
      "grad_norm": 1.4845091924788132,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8884591460227966,
      "num_tokens": 398958296.0,
      "step": 11471
    },
    {
      "epoch": 2.1303621169916434,
      "grad_norm": 1.5325508681269129,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8966382741928101,
      "num_tokens": 398989991.0,
      "step": 11472
    },
    {
      "epoch": 2.130547818012999,
      "grad_norm": 1.5467550288192824,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8877002000808716,
      "num_tokens": 399026268.0,
      "step": 11473
    },
    {
      "epoch": 2.130733519034355,
      "grad_norm": 1.7758016160751084,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8755318522453308,
      "num_tokens": 399053744.0,
      "step": 11474
    },
    {
      "epoch": 2.13091922005571,
      "grad_norm": 1.5054677574184838,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8923738598823547,
      "num_tokens": 399087352.0,
      "step": 11475
    },
    {
      "epoch": 2.131104921077066,
      "grad_norm": 1.4793805137651195,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8823059797286987,
      "num_tokens": 399126875.0,
      "step": 11476
    },
    {
      "epoch": 2.1312906220984216,
      "grad_norm": 1.5612264681681567,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8848668336868286,
      "num_tokens": 399162285.0,
      "step": 11477
    },
    {
      "epoch": 2.1314763231197773,
      "grad_norm": 1.6634249749703263,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8774992227554321,
      "num_tokens": 399194237.0,
      "step": 11478
    },
    {
      "epoch": 2.1316620241411326,
      "grad_norm": 1.5347521434272906,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8795719146728516,
      "num_tokens": 399229811.0,
      "step": 11479
    },
    {
      "epoch": 2.1318477251624883,
      "grad_norm": 1.5745898823521427,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8762418031692505,
      "num_tokens": 399263976.0,
      "step": 11480
    },
    {
      "epoch": 2.132033426183844,
      "grad_norm": 1.434175564941826,
      "learning_rate": 1e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8705272078514099,
      "num_tokens": 399306461.0,
      "step": 11481
    },
    {
      "epoch": 2.1322191272052,
      "grad_norm": 1.3217161418257093,
      "learning_rate": 1e-06,
      "loss": 0.2812,
      "mean_token_accuracy": 0.8960092067718506,
      "num_tokens": 399351358.0,
      "step": 11482
    },
    {
      "epoch": 2.132404828226555,
      "grad_norm": 1.6435531272027486,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8827115297317505,
      "num_tokens": 399381943.0,
      "step": 11483
    },
    {
      "epoch": 2.132590529247911,
      "grad_norm": 1.426686628812192,
      "learning_rate": 1e-06,
      "loss": 0.2808,
      "mean_token_accuracy": 0.8994085788726807,
      "num_tokens": 399417804.0,
      "step": 11484
    },
    {
      "epoch": 2.1327762302692665,
      "grad_norm": 1.5805321518947784,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8805840015411377,
      "num_tokens": 399453201.0,
      "step": 11485
    },
    {
      "epoch": 2.1329619312906223,
      "grad_norm": 1.7631383543055725,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8897936344146729,
      "num_tokens": 399480337.0,
      "step": 11486
    },
    {
      "epoch": 2.1331476323119776,
      "grad_norm": 1.5971090615314245,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8852987289428711,
      "num_tokens": 399513963.0,
      "step": 11487
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 1.8381283866625333,
      "learning_rate": 1e-06,
      "loss": 0.2717,
      "mean_token_accuracy": 0.8990898132324219,
      "num_tokens": 399537853.0,
      "step": 11488
    },
    {
      "epoch": 2.133519034354689,
      "grad_norm": 1.414021592829533,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8882941603660583,
      "num_tokens": 399580571.0,
      "step": 11489
    },
    {
      "epoch": 2.1337047353760448,
      "grad_norm": 1.4988275096623591,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8815294504165649,
      "num_tokens": 399623774.0,
      "step": 11490
    },
    {
      "epoch": 2.1338904363974,
      "grad_norm": 1.49210013051069,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8801535964012146,
      "num_tokens": 399669011.0,
      "step": 11491
    },
    {
      "epoch": 2.134076137418756,
      "grad_norm": 1.6255333954847424,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8739145398139954,
      "num_tokens": 399705614.0,
      "step": 11492
    },
    {
      "epoch": 2.1342618384401115,
      "grad_norm": 1.5009199310317332,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8964217901229858,
      "num_tokens": 399743522.0,
      "step": 11493
    },
    {
      "epoch": 2.1344475394614673,
      "grad_norm": 1.6949372675212615,
      "learning_rate": 1e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8623375296592712,
      "num_tokens": 399780415.0,
      "step": 11494
    },
    {
      "epoch": 2.1346332404828225,
      "grad_norm": 1.6093190433785907,
      "learning_rate": 1e-06,
      "loss": 0.2885,
      "mean_token_accuracy": 0.8962991237640381,
      "num_tokens": 399812557.0,
      "step": 11495
    },
    {
      "epoch": 2.1348189415041783,
      "grad_norm": 1.778940379125714,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8803560733795166,
      "num_tokens": 399843493.0,
      "step": 11496
    },
    {
      "epoch": 2.135004642525534,
      "grad_norm": 1.6356337780989352,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8876007795333862,
      "num_tokens": 399876937.0,
      "step": 11497
    },
    {
      "epoch": 2.1351903435468893,
      "grad_norm": 1.441084915592917,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8929803967475891,
      "num_tokens": 399912397.0,
      "step": 11498
    },
    {
      "epoch": 2.135376044568245,
      "grad_norm": 1.5545093712423823,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8943078517913818,
      "num_tokens": 399946474.0,
      "step": 11499
    },
    {
      "epoch": 2.1355617455896008,
      "grad_norm": 1.4978671284330436,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8935233354568481,
      "num_tokens": 399981547.0,
      "step": 11500
    },
    {
      "epoch": 2.1357474466109565,
      "grad_norm": 1.6285706949349978,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.880455493927002,
      "num_tokens": 400015968.0,
      "step": 11501
    },
    {
      "epoch": 2.135933147632312,
      "grad_norm": 1.4172737477434338,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8917492628097534,
      "num_tokens": 400053462.0,
      "step": 11502
    },
    {
      "epoch": 2.1361188486536675,
      "grad_norm": 1.5139130878887794,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8769376277923584,
      "num_tokens": 400092267.0,
      "step": 11503
    },
    {
      "epoch": 2.1363045496750233,
      "grad_norm": 1.5673459446661968,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8882903456687927,
      "num_tokens": 400126800.0,
      "step": 11504
    },
    {
      "epoch": 2.136490250696379,
      "grad_norm": 1.4310187640038055,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.8957188129425049,
      "num_tokens": 400164845.0,
      "step": 11505
    },
    {
      "epoch": 2.1366759517177343,
      "grad_norm": 1.713506263342064,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8738496899604797,
      "num_tokens": 400197945.0,
      "step": 11506
    },
    {
      "epoch": 2.13686165273909,
      "grad_norm": 1.5470472653762066,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8854048252105713,
      "num_tokens": 400232336.0,
      "step": 11507
    },
    {
      "epoch": 2.1370473537604457,
      "grad_norm": 1.5031235122082935,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8826266527175903,
      "num_tokens": 400269089.0,
      "step": 11508
    },
    {
      "epoch": 2.1372330547818015,
      "grad_norm": 1.6120964430744027,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8877245187759399,
      "num_tokens": 400300926.0,
      "step": 11509
    },
    {
      "epoch": 2.1374187558031568,
      "grad_norm": 1.5870240403995295,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.8982700109481812,
      "num_tokens": 400332550.0,
      "step": 11510
    },
    {
      "epoch": 2.1376044568245125,
      "grad_norm": 1.5717753552875335,
      "learning_rate": 1e-06,
      "loss": 0.3401,
      "mean_token_accuracy": 0.8795785903930664,
      "num_tokens": 400369512.0,
      "step": 11511
    },
    {
      "epoch": 2.1377901578458682,
      "grad_norm": 1.6109736775564667,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8886288404464722,
      "num_tokens": 400404185.0,
      "step": 11512
    },
    {
      "epoch": 2.137975858867224,
      "grad_norm": 1.6740944588800624,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8766639232635498,
      "num_tokens": 400436977.0,
      "step": 11513
    },
    {
      "epoch": 2.1381615598885793,
      "grad_norm": 1.5306993305525143,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8842604756355286,
      "num_tokens": 400473800.0,
      "step": 11514
    },
    {
      "epoch": 2.138347260909935,
      "grad_norm": 1.5395094998862833,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8924497365951538,
      "num_tokens": 400509133.0,
      "step": 11515
    },
    {
      "epoch": 2.1385329619312907,
      "grad_norm": 1.4524794855875047,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8816865682601929,
      "num_tokens": 400552884.0,
      "step": 11516
    },
    {
      "epoch": 2.1387186629526465,
      "grad_norm": 1.5253475708404995,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8961138129234314,
      "num_tokens": 400586911.0,
      "step": 11517
    },
    {
      "epoch": 2.1389043639740017,
      "grad_norm": 1.4382417304601716,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8947675228118896,
      "num_tokens": 400627772.0,
      "step": 11518
    },
    {
      "epoch": 2.1390900649953575,
      "grad_norm": 1.589052293341834,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8827071189880371,
      "num_tokens": 400664340.0,
      "step": 11519
    },
    {
      "epoch": 2.139275766016713,
      "grad_norm": 1.592282266687377,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.888801097869873,
      "num_tokens": 400696775.0,
      "step": 11520
    },
    {
      "epoch": 2.1394614670380685,
      "grad_norm": 1.7292516518022274,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8880491256713867,
      "num_tokens": 400727889.0,
      "step": 11521
    },
    {
      "epoch": 2.1396471680594242,
      "grad_norm": 1.5696656641304427,
      "learning_rate": 1e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8727816343307495,
      "num_tokens": 400769259.0,
      "step": 11522
    },
    {
      "epoch": 2.13983286908078,
      "grad_norm": 1.5111868957449406,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8900810480117798,
      "num_tokens": 400802583.0,
      "step": 11523
    },
    {
      "epoch": 2.1400185701021357,
      "grad_norm": 1.5244707307227434,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8862965106964111,
      "num_tokens": 400839523.0,
      "step": 11524
    },
    {
      "epoch": 2.140204271123491,
      "grad_norm": 1.4298421220750803,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.884052574634552,
      "num_tokens": 400880901.0,
      "step": 11525
    },
    {
      "epoch": 2.1403899721448467,
      "grad_norm": 1.6366293679200277,
      "learning_rate": 1e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8619747161865234,
      "num_tokens": 400919957.0,
      "step": 11526
    },
    {
      "epoch": 2.1405756731662025,
      "grad_norm": 1.5541418917876142,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8831382989883423,
      "num_tokens": 400959714.0,
      "step": 11527
    },
    {
      "epoch": 2.140761374187558,
      "grad_norm": 1.48009235123808,
      "learning_rate": 1e-06,
      "loss": 0.2818,
      "mean_token_accuracy": 0.8994205594062805,
      "num_tokens": 400997565.0,
      "step": 11528
    },
    {
      "epoch": 2.1409470752089135,
      "grad_norm": 1.5525938517892215,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.87627774477005,
      "num_tokens": 401034203.0,
      "step": 11529
    },
    {
      "epoch": 2.141132776230269,
      "grad_norm": 1.5670328491280878,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8854269981384277,
      "num_tokens": 401069842.0,
      "step": 11530
    },
    {
      "epoch": 2.141318477251625,
      "grad_norm": 1.5121280824249208,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8825916051864624,
      "num_tokens": 401108830.0,
      "step": 11531
    },
    {
      "epoch": 2.1415041782729807,
      "grad_norm": 1.6274601532808355,
      "learning_rate": 1e-06,
      "loss": 0.2905,
      "mean_token_accuracy": 0.8974478244781494,
      "num_tokens": 401142874.0,
      "step": 11532
    },
    {
      "epoch": 2.141689879294336,
      "grad_norm": 1.5068779003379609,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.895450234413147,
      "num_tokens": 401180912.0,
      "step": 11533
    },
    {
      "epoch": 2.1418755803156917,
      "grad_norm": 1.474511051800219,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.876017153263092,
      "num_tokens": 401221742.0,
      "step": 11534
    },
    {
      "epoch": 2.1420612813370474,
      "grad_norm": 1.464214123921411,
      "learning_rate": 1e-06,
      "loss": 0.2981,
      "mean_token_accuracy": 0.893720269203186,
      "num_tokens": 401263341.0,
      "step": 11535
    },
    {
      "epoch": 2.142246982358403,
      "grad_norm": 1.5953458300218142,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8889093995094299,
      "num_tokens": 401299038.0,
      "step": 11536
    },
    {
      "epoch": 2.1424326833797585,
      "grad_norm": 1.6619134678924914,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8726118206977844,
      "num_tokens": 401332333.0,
      "step": 11537
    },
    {
      "epoch": 2.142618384401114,
      "grad_norm": 1.523391860535343,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8873032927513123,
      "num_tokens": 401370407.0,
      "step": 11538
    },
    {
      "epoch": 2.14280408542247,
      "grad_norm": 1.5149474313561762,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8953189849853516,
      "num_tokens": 401404584.0,
      "step": 11539
    },
    {
      "epoch": 2.1429897864438257,
      "grad_norm": 1.5272300500079194,
      "learning_rate": 1e-06,
      "loss": 0.2834,
      "mean_token_accuracy": 0.8994483351707458,
      "num_tokens": 401435784.0,
      "step": 11540
    },
    {
      "epoch": 2.143175487465181,
      "grad_norm": 1.791098152447645,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8724932670593262,
      "num_tokens": 401465482.0,
      "step": 11541
    },
    {
      "epoch": 2.1433611884865367,
      "grad_norm": 1.5015501956910797,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8881100416183472,
      "num_tokens": 401503523.0,
      "step": 11542
    },
    {
      "epoch": 2.1435468895078924,
      "grad_norm": 1.5019281745742972,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8927826285362244,
      "num_tokens": 401535998.0,
      "step": 11543
    },
    {
      "epoch": 2.1437325905292477,
      "grad_norm": 1.4286455720599651,
      "learning_rate": 1e-06,
      "loss": 0.2713,
      "mean_token_accuracy": 0.9021672010421753,
      "num_tokens": 401573572.0,
      "step": 11544
    },
    {
      "epoch": 2.1439182915506034,
      "grad_norm": 1.5304948196163124,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8877087831497192,
      "num_tokens": 401612110.0,
      "step": 11545
    },
    {
      "epoch": 2.144103992571959,
      "grad_norm": 1.7661864462713819,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8782317638397217,
      "num_tokens": 401642679.0,
      "step": 11546
    },
    {
      "epoch": 2.144289693593315,
      "grad_norm": 1.5152408080738957,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8800061345100403,
      "num_tokens": 401684730.0,
      "step": 11547
    },
    {
      "epoch": 2.14447539461467,
      "grad_norm": 1.6308250758063063,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8843644857406616,
      "num_tokens": 401719158.0,
      "step": 11548
    },
    {
      "epoch": 2.144661095636026,
      "grad_norm": 1.5071161200396945,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8852584362030029,
      "num_tokens": 401759227.0,
      "step": 11549
    },
    {
      "epoch": 2.1448467966573816,
      "grad_norm": 1.7543396351460598,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8773130774497986,
      "num_tokens": 401790457.0,
      "step": 11550
    },
    {
      "epoch": 2.1450324976787374,
      "grad_norm": 1.559457351273411,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8876218795776367,
      "num_tokens": 401826085.0,
      "step": 11551
    },
    {
      "epoch": 2.1452181987000927,
      "grad_norm": 1.753162764884759,
      "learning_rate": 1e-06,
      "loss": 0.3005,
      "mean_token_accuracy": 0.8969089984893799,
      "num_tokens": 401853496.0,
      "step": 11552
    },
    {
      "epoch": 2.1454038997214484,
      "grad_norm": 1.4845698222725017,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8758682012557983,
      "num_tokens": 401898391.0,
      "step": 11553
    },
    {
      "epoch": 2.145589600742804,
      "grad_norm": 1.563041125878914,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.90055251121521,
      "num_tokens": 401930569.0,
      "step": 11554
    },
    {
      "epoch": 2.14577530176416,
      "grad_norm": 1.8307711217082197,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8801732063293457,
      "num_tokens": 401955314.0,
      "step": 11555
    },
    {
      "epoch": 2.145961002785515,
      "grad_norm": 1.5370962880028118,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8773478865623474,
      "num_tokens": 401993808.0,
      "step": 11556
    },
    {
      "epoch": 2.146146703806871,
      "grad_norm": 1.6148821730131304,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8904607892036438,
      "num_tokens": 402026123.0,
      "step": 11557
    },
    {
      "epoch": 2.1463324048282266,
      "grad_norm": 1.5406432020525183,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8792294263839722,
      "num_tokens": 402065940.0,
      "step": 11558
    },
    {
      "epoch": 2.1465181058495824,
      "grad_norm": 1.6897868858477367,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.8940150737762451,
      "num_tokens": 402094801.0,
      "step": 11559
    },
    {
      "epoch": 2.1467038068709376,
      "grad_norm": 1.5742543998956555,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8855528831481934,
      "num_tokens": 402129910.0,
      "step": 11560
    },
    {
      "epoch": 2.1468895078922934,
      "grad_norm": 1.6097127040489507,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8815881013870239,
      "num_tokens": 402163737.0,
      "step": 11561
    },
    {
      "epoch": 2.147075208913649,
      "grad_norm": 1.693824853556793,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8743383288383484,
      "num_tokens": 402202464.0,
      "step": 11562
    },
    {
      "epoch": 2.147260909935005,
      "grad_norm": 1.779480651287565,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8808624744415283,
      "num_tokens": 402230359.0,
      "step": 11563
    },
    {
      "epoch": 2.14744661095636,
      "grad_norm": 1.6473816338535647,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8930574059486389,
      "num_tokens": 402263992.0,
      "step": 11564
    },
    {
      "epoch": 2.147632311977716,
      "grad_norm": 1.7976937582699783,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8779346942901611,
      "num_tokens": 402298328.0,
      "step": 11565
    },
    {
      "epoch": 2.1478180129990716,
      "grad_norm": 1.492090654820023,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8820126056671143,
      "num_tokens": 402336376.0,
      "step": 11566
    },
    {
      "epoch": 2.148003714020427,
      "grad_norm": 1.4725861893297219,
      "learning_rate": 1e-06,
      "loss": 0.2891,
      "mean_token_accuracy": 0.9009969234466553,
      "num_tokens": 402371540.0,
      "step": 11567
    },
    {
      "epoch": 2.1481894150417826,
      "grad_norm": 1.7664701254664121,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8873357772827148,
      "num_tokens": 402401586.0,
      "step": 11568
    },
    {
      "epoch": 2.1483751160631384,
      "grad_norm": 1.688069823432918,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8809951543807983,
      "num_tokens": 402434095.0,
      "step": 11569
    },
    {
      "epoch": 2.148560817084494,
      "grad_norm": 1.662953069310737,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8842588663101196,
      "num_tokens": 402469056.0,
      "step": 11570
    },
    {
      "epoch": 2.1487465181058494,
      "grad_norm": 1.4264503853224815,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.8897790908813477,
      "num_tokens": 402510054.0,
      "step": 11571
    },
    {
      "epoch": 2.148932219127205,
      "grad_norm": 1.478977338692461,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8815339803695679,
      "num_tokens": 402552892.0,
      "step": 11572
    },
    {
      "epoch": 2.149117920148561,
      "grad_norm": 1.4526058358563956,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8866146802902222,
      "num_tokens": 402591233.0,
      "step": 11573
    },
    {
      "epoch": 2.1493036211699166,
      "grad_norm": 1.6041734294179333,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8659310936927795,
      "num_tokens": 402632660.0,
      "step": 11574
    },
    {
      "epoch": 2.149489322191272,
      "grad_norm": 1.61991633193676,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8937565684318542,
      "num_tokens": 402665705.0,
      "step": 11575
    },
    {
      "epoch": 2.1496750232126276,
      "grad_norm": 1.6001400132693995,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.8953529596328735,
      "num_tokens": 402697345.0,
      "step": 11576
    },
    {
      "epoch": 2.1498607242339833,
      "grad_norm": 1.674277840129989,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8803319931030273,
      "num_tokens": 402729370.0,
      "step": 11577
    },
    {
      "epoch": 2.150046425255339,
      "grad_norm": 1.5163208899266083,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8898463249206543,
      "num_tokens": 402763578.0,
      "step": 11578
    },
    {
      "epoch": 2.1502321262766944,
      "grad_norm": 1.4443189196524082,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8974248170852661,
      "num_tokens": 402800704.0,
      "step": 11579
    },
    {
      "epoch": 2.15041782729805,
      "grad_norm": 1.5766055677783721,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8831839561462402,
      "num_tokens": 402836820.0,
      "step": 11580
    },
    {
      "epoch": 2.150603528319406,
      "grad_norm": 1.626196644436628,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8813991546630859,
      "num_tokens": 402868947.0,
      "step": 11581
    },
    {
      "epoch": 2.1507892293407616,
      "grad_norm": 1.6352289129096784,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8800777196884155,
      "num_tokens": 402903964.0,
      "step": 11582
    },
    {
      "epoch": 2.150974930362117,
      "grad_norm": 1.4585490088181206,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8905223608016968,
      "num_tokens": 402949299.0,
      "step": 11583
    },
    {
      "epoch": 2.1511606313834726,
      "grad_norm": 1.4733682192726325,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.892484188079834,
      "num_tokens": 402985587.0,
      "step": 11584
    },
    {
      "epoch": 2.1513463324048283,
      "grad_norm": 1.7242142851135736,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8838280439376831,
      "num_tokens": 403016433.0,
      "step": 11585
    },
    {
      "epoch": 2.151532033426184,
      "grad_norm": 1.6830529690564924,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8870486617088318,
      "num_tokens": 403045588.0,
      "step": 11586
    },
    {
      "epoch": 2.1517177344475393,
      "grad_norm": 1.6285352036635532,
      "learning_rate": 1e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8712242841720581,
      "num_tokens": 403083685.0,
      "step": 11587
    },
    {
      "epoch": 2.151903435468895,
      "grad_norm": 1.4873123892534181,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8836972713470459,
      "num_tokens": 403118746.0,
      "step": 11588
    },
    {
      "epoch": 2.152089136490251,
      "grad_norm": 1.52154783605868,
      "learning_rate": 1e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8763370513916016,
      "num_tokens": 403155863.0,
      "step": 11589
    },
    {
      "epoch": 2.152274837511606,
      "grad_norm": 1.6408965247912721,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8817853927612305,
      "num_tokens": 403190563.0,
      "step": 11590
    },
    {
      "epoch": 2.152460538532962,
      "grad_norm": 1.5179823370955283,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8833592534065247,
      "num_tokens": 403227897.0,
      "step": 11591
    },
    {
      "epoch": 2.1526462395543176,
      "grad_norm": 1.7140398541098343,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8906210660934448,
      "num_tokens": 403259151.0,
      "step": 11592
    },
    {
      "epoch": 2.1528319405756733,
      "grad_norm": 1.58758508007346,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8886661529541016,
      "num_tokens": 403291750.0,
      "step": 11593
    },
    {
      "epoch": 2.153017641597029,
      "grad_norm": 1.3826648092269318,
      "learning_rate": 1e-06,
      "loss": 0.2735,
      "mean_token_accuracy": 0.901627779006958,
      "num_tokens": 403334366.0,
      "step": 11594
    },
    {
      "epoch": 2.1532033426183843,
      "grad_norm": 1.5496423789396137,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8829307556152344,
      "num_tokens": 403374123.0,
      "step": 11595
    },
    {
      "epoch": 2.15338904363974,
      "grad_norm": 1.6311392702985492,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8886876702308655,
      "num_tokens": 403408370.0,
      "step": 11596
    },
    {
      "epoch": 2.1535747446610958,
      "grad_norm": 1.5808958002612188,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8869552612304688,
      "num_tokens": 403444086.0,
      "step": 11597
    },
    {
      "epoch": 2.153760445682451,
      "grad_norm": 1.5495017602516994,
      "learning_rate": 1e-06,
      "loss": 0.2858,
      "mean_token_accuracy": 0.8969326019287109,
      "num_tokens": 403476439.0,
      "step": 11598
    },
    {
      "epoch": 2.153946146703807,
      "grad_norm": 1.5463935289286206,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.889983057975769,
      "num_tokens": 403510227.0,
      "step": 11599
    },
    {
      "epoch": 2.1541318477251625,
      "grad_norm": 1.6423477888658704,
      "learning_rate": 1e-06,
      "loss": 0.2919,
      "mean_token_accuracy": 0.8925657868385315,
      "num_tokens": 403543022.0,
      "step": 11600
    },
    {
      "epoch": 2.1543175487465183,
      "grad_norm": 1.4730792236214532,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8834500908851624,
      "num_tokens": 403582868.0,
      "step": 11601
    },
    {
      "epoch": 2.1545032497678736,
      "grad_norm": 1.7271764219923977,
      "learning_rate": 1e-06,
      "loss": 0.2823,
      "mean_token_accuracy": 0.899768590927124,
      "num_tokens": 403610731.0,
      "step": 11602
    },
    {
      "epoch": 2.1546889507892293,
      "grad_norm": 1.6182228668604777,
      "learning_rate": 1e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.862220048904419,
      "num_tokens": 403645533.0,
      "step": 11603
    },
    {
      "epoch": 2.154874651810585,
      "grad_norm": 1.6716594581313078,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8898760080337524,
      "num_tokens": 403676157.0,
      "step": 11604
    },
    {
      "epoch": 2.1550603528319408,
      "grad_norm": 1.465751284084704,
      "learning_rate": 1e-06,
      "loss": 0.2735,
      "mean_token_accuracy": 0.9017104506492615,
      "num_tokens": 403711720.0,
      "step": 11605
    },
    {
      "epoch": 2.155246053853296,
      "grad_norm": 1.740584089264913,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8824543952941895,
      "num_tokens": 403744596.0,
      "step": 11606
    },
    {
      "epoch": 2.1554317548746518,
      "grad_norm": 1.58712489886943,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8847248554229736,
      "num_tokens": 403780084.0,
      "step": 11607
    },
    {
      "epoch": 2.1556174558960075,
      "grad_norm": 1.6462030548590354,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8873570561408997,
      "num_tokens": 403815057.0,
      "step": 11608
    },
    {
      "epoch": 2.1558031569173632,
      "grad_norm": 1.6092654679767595,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8850382566452026,
      "num_tokens": 403847899.0,
      "step": 11609
    },
    {
      "epoch": 2.1559888579387185,
      "grad_norm": 1.5752900959942278,
      "learning_rate": 1e-06,
      "loss": 0.2972,
      "mean_token_accuracy": 0.8949625492095947,
      "num_tokens": 403879382.0,
      "step": 11610
    },
    {
      "epoch": 2.1561745589600743,
      "grad_norm": 1.5066240383814242,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8855957984924316,
      "num_tokens": 403920720.0,
      "step": 11611
    },
    {
      "epoch": 2.15636025998143,
      "grad_norm": 1.4988091412892246,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8904925584793091,
      "num_tokens": 403955406.0,
      "step": 11612
    },
    {
      "epoch": 2.1565459610027853,
      "grad_norm": 1.5328587385237922,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.8764269351959229,
      "num_tokens": 403997104.0,
      "step": 11613
    },
    {
      "epoch": 2.156731662024141,
      "grad_norm": 1.529818342983563,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8885599970817566,
      "num_tokens": 404032448.0,
      "step": 11614
    },
    {
      "epoch": 2.1569173630454967,
      "grad_norm": 1.6399690177240949,
      "learning_rate": 1e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8638700246810913,
      "num_tokens": 404071818.0,
      "step": 11615
    },
    {
      "epoch": 2.1571030640668525,
      "grad_norm": 1.603577416624565,
      "learning_rate": 1e-06,
      "loss": 0.2769,
      "mean_token_accuracy": 0.9045116901397705,
      "num_tokens": 404099106.0,
      "step": 11616
    },
    {
      "epoch": 2.157288765088208,
      "grad_norm": 1.639843014053207,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8798249959945679,
      "num_tokens": 404131252.0,
      "step": 11617
    },
    {
      "epoch": 2.1574744661095635,
      "grad_norm": 1.7419349901592038,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8983449935913086,
      "num_tokens": 404166960.0,
      "step": 11618
    },
    {
      "epoch": 2.1576601671309192,
      "grad_norm": 1.5603588087385325,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8844302296638489,
      "num_tokens": 404200822.0,
      "step": 11619
    },
    {
      "epoch": 2.157845868152275,
      "grad_norm": 1.5234698508865907,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8856521844863892,
      "num_tokens": 404238241.0,
      "step": 11620
    },
    {
      "epoch": 2.1580315691736303,
      "grad_norm": 1.4507602859890933,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8799423575401306,
      "num_tokens": 404283033.0,
      "step": 11621
    },
    {
      "epoch": 2.158217270194986,
      "grad_norm": 1.5813424215104461,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8959081172943115,
      "num_tokens": 404315188.0,
      "step": 11622
    },
    {
      "epoch": 2.1584029712163417,
      "grad_norm": 1.537359028112004,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8883582353591919,
      "num_tokens": 404349957.0,
      "step": 11623
    },
    {
      "epoch": 2.1585886722376975,
      "grad_norm": 1.4577118080033968,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8756054639816284,
      "num_tokens": 404392419.0,
      "step": 11624
    },
    {
      "epoch": 2.1587743732590527,
      "grad_norm": 1.6241336275176708,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.893293023109436,
      "num_tokens": 404423070.0,
      "step": 11625
    },
    {
      "epoch": 2.1589600742804085,
      "grad_norm": 1.6910055600583798,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8801904916763306,
      "num_tokens": 404453503.0,
      "step": 11626
    },
    {
      "epoch": 2.159145775301764,
      "grad_norm": 1.4530421590647784,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8792096972465515,
      "num_tokens": 404493484.0,
      "step": 11627
    },
    {
      "epoch": 2.15933147632312,
      "grad_norm": 1.5058952263148986,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8849751949310303,
      "num_tokens": 404531668.0,
      "step": 11628
    },
    {
      "epoch": 2.1595171773444752,
      "grad_norm": 1.4422602239437265,
      "learning_rate": 1e-06,
      "loss": 0.2832,
      "mean_token_accuracy": 0.8975895643234253,
      "num_tokens": 404567634.0,
      "step": 11629
    },
    {
      "epoch": 2.159702878365831,
      "grad_norm": 1.6755741858289044,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8705410361289978,
      "num_tokens": 404600654.0,
      "step": 11630
    },
    {
      "epoch": 2.1598885793871867,
      "grad_norm": 1.5457414144057036,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8845773935317993,
      "num_tokens": 404637051.0,
      "step": 11631
    },
    {
      "epoch": 2.1600742804085424,
      "grad_norm": 1.6173750905009983,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8793096542358398,
      "num_tokens": 404670036.0,
      "step": 11632
    },
    {
      "epoch": 2.1602599814298977,
      "grad_norm": 1.52192472079949,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8777991533279419,
      "num_tokens": 404705657.0,
      "step": 11633
    },
    {
      "epoch": 2.1604456824512535,
      "grad_norm": 1.5066102287167211,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8879801034927368,
      "num_tokens": 404741667.0,
      "step": 11634
    },
    {
      "epoch": 2.160631383472609,
      "grad_norm": 1.577152442435562,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8788632750511169,
      "num_tokens": 404775616.0,
      "step": 11635
    },
    {
      "epoch": 2.160817084493965,
      "grad_norm": 1.5901264803834083,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8834545612335205,
      "num_tokens": 404809260.0,
      "step": 11636
    },
    {
      "epoch": 2.16100278551532,
      "grad_norm": 1.5378489481919153,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8836696743965149,
      "num_tokens": 404848561.0,
      "step": 11637
    },
    {
      "epoch": 2.161188486536676,
      "grad_norm": 1.5066625754717724,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.883664608001709,
      "num_tokens": 404886858.0,
      "step": 11638
    },
    {
      "epoch": 2.1613741875580317,
      "grad_norm": 1.6472097062910867,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.88136225938797,
      "num_tokens": 404920438.0,
      "step": 11639
    },
    {
      "epoch": 2.1615598885793874,
      "grad_norm": 1.6986310781557603,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.891517162322998,
      "num_tokens": 404950808.0,
      "step": 11640
    },
    {
      "epoch": 2.1617455896007427,
      "grad_norm": 1.4827109702800907,
      "learning_rate": 1e-06,
      "loss": 0.2794,
      "mean_token_accuracy": 0.9003127813339233,
      "num_tokens": 404987230.0,
      "step": 11641
    },
    {
      "epoch": 2.1619312906220984,
      "grad_norm": 1.473233769962908,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8883503675460815,
      "num_tokens": 405024811.0,
      "step": 11642
    },
    {
      "epoch": 2.162116991643454,
      "grad_norm": 1.4506465669398185,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8925612568855286,
      "num_tokens": 405062848.0,
      "step": 11643
    },
    {
      "epoch": 2.1623026926648095,
      "grad_norm": 1.734026338608741,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.8952740430831909,
      "num_tokens": 405091111.0,
      "step": 11644
    },
    {
      "epoch": 2.162488393686165,
      "grad_norm": 1.4465647913329163,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8830732107162476,
      "num_tokens": 405130106.0,
      "step": 11645
    },
    {
      "epoch": 2.162674094707521,
      "grad_norm": 1.8023478741168566,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8885711431503296,
      "num_tokens": 405159315.0,
      "step": 11646
    },
    {
      "epoch": 2.1628597957288767,
      "grad_norm": 1.603677211071541,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.874884307384491,
      "num_tokens": 405195603.0,
      "step": 11647
    },
    {
      "epoch": 2.163045496750232,
      "grad_norm": 1.6224605834601886,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8944467306137085,
      "num_tokens": 405225906.0,
      "step": 11648
    },
    {
      "epoch": 2.1632311977715877,
      "grad_norm": 1.4632027229136442,
      "learning_rate": 1e-06,
      "loss": 0.2872,
      "mean_token_accuracy": 0.8983588218688965,
      "num_tokens": 405258995.0,
      "step": 11649
    },
    {
      "epoch": 2.1634168987929434,
      "grad_norm": 1.5093017291444557,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8942629098892212,
      "num_tokens": 405296003.0,
      "step": 11650
    },
    {
      "epoch": 2.163602599814299,
      "grad_norm": 1.6200803305679252,
      "learning_rate": 1e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8760121464729309,
      "num_tokens": 405331764.0,
      "step": 11651
    },
    {
      "epoch": 2.1637883008356544,
      "grad_norm": 1.6282627391976525,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.876571774482727,
      "num_tokens": 405366179.0,
      "step": 11652
    },
    {
      "epoch": 2.16397400185701,
      "grad_norm": 1.7005189124896374,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8819717764854431,
      "num_tokens": 405398674.0,
      "step": 11653
    },
    {
      "epoch": 2.164159702878366,
      "grad_norm": 1.5829509995488278,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8901887536048889,
      "num_tokens": 405435739.0,
      "step": 11654
    },
    {
      "epoch": 2.1643454038997216,
      "grad_norm": 1.4924973689429024,
      "learning_rate": 1e-06,
      "loss": 0.2826,
      "mean_token_accuracy": 0.8999420404434204,
      "num_tokens": 405468439.0,
      "step": 11655
    },
    {
      "epoch": 2.164531104921077,
      "grad_norm": 1.5595716063307843,
      "learning_rate": 1e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.871844470500946,
      "num_tokens": 405505272.0,
      "step": 11656
    },
    {
      "epoch": 2.1647168059424327,
      "grad_norm": 1.6529867578439552,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8889447450637817,
      "num_tokens": 405537632.0,
      "step": 11657
    },
    {
      "epoch": 2.1649025069637884,
      "grad_norm": 1.6792200183656023,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.879692554473877,
      "num_tokens": 405567708.0,
      "step": 11658
    },
    {
      "epoch": 2.165088207985144,
      "grad_norm": 1.6609767502006225,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.881998598575592,
      "num_tokens": 405599690.0,
      "step": 11659
    },
    {
      "epoch": 2.1652739090064994,
      "grad_norm": 1.6981319264067236,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8943871259689331,
      "num_tokens": 405632375.0,
      "step": 11660
    },
    {
      "epoch": 2.165459610027855,
      "grad_norm": 1.6904873909889877,
      "learning_rate": 1e-06,
      "loss": 0.3815,
      "mean_token_accuracy": 0.8706902265548706,
      "num_tokens": 405670536.0,
      "step": 11661
    },
    {
      "epoch": 2.165645311049211,
      "grad_norm": 1.5086508026639798,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8868618607521057,
      "num_tokens": 405709986.0,
      "step": 11662
    },
    {
      "epoch": 2.1658310120705666,
      "grad_norm": 1.6189609158848093,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8750041723251343,
      "num_tokens": 405746369.0,
      "step": 11663
    },
    {
      "epoch": 2.166016713091922,
      "grad_norm": 1.5558789509370796,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.9010622501373291,
      "num_tokens": 405779624.0,
      "step": 11664
    },
    {
      "epoch": 2.1662024141132776,
      "grad_norm": 1.529653637862676,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8829967975616455,
      "num_tokens": 405819860.0,
      "step": 11665
    },
    {
      "epoch": 2.1663881151346334,
      "grad_norm": 1.4802855624934614,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.876934826374054,
      "num_tokens": 405860507.0,
      "step": 11666
    },
    {
      "epoch": 2.1665738161559887,
      "grad_norm": 1.6660699247469022,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8823645710945129,
      "num_tokens": 405893573.0,
      "step": 11667
    },
    {
      "epoch": 2.1667595171773444,
      "grad_norm": 1.492419373235323,
      "learning_rate": 1e-06,
      "loss": 0.2833,
      "mean_token_accuracy": 0.9008895754814148,
      "num_tokens": 405930040.0,
      "step": 11668
    },
    {
      "epoch": 2.1669452181987,
      "grad_norm": 1.350776105207982,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8822597861289978,
      "num_tokens": 405977740.0,
      "step": 11669
    },
    {
      "epoch": 2.167130919220056,
      "grad_norm": 1.467628251788487,
      "learning_rate": 1e-06,
      "loss": 0.2846,
      "mean_token_accuracy": 0.8969319462776184,
      "num_tokens": 406012226.0,
      "step": 11670
    },
    {
      "epoch": 2.167316620241411,
      "grad_norm": 1.741453803822538,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8846123814582825,
      "num_tokens": 406040490.0,
      "step": 11671
    },
    {
      "epoch": 2.167502321262767,
      "grad_norm": 1.4982573116214808,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8876293897628784,
      "num_tokens": 406074748.0,
      "step": 11672
    },
    {
      "epoch": 2.1676880222841226,
      "grad_norm": 1.4664203459116028,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8882393836975098,
      "num_tokens": 406111796.0,
      "step": 11673
    },
    {
      "epoch": 2.1678737233054783,
      "grad_norm": 1.817041475910051,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8788217306137085,
      "num_tokens": 406142244.0,
      "step": 11674
    },
    {
      "epoch": 2.1680594243268336,
      "grad_norm": 1.8365176885907175,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.881421685218811,
      "num_tokens": 406177276.0,
      "step": 11675
    },
    {
      "epoch": 2.1682451253481894,
      "grad_norm": 1.6562444922627901,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8821155428886414,
      "num_tokens": 406209794.0,
      "step": 11676
    },
    {
      "epoch": 2.168430826369545,
      "grad_norm": 1.549003119024115,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8926507830619812,
      "num_tokens": 406244309.0,
      "step": 11677
    },
    {
      "epoch": 2.168616527390901,
      "grad_norm": 1.376015531369819,
      "learning_rate": 1e-06,
      "loss": 0.2787,
      "mean_token_accuracy": 0.9023634195327759,
      "num_tokens": 406285012.0,
      "step": 11678
    },
    {
      "epoch": 2.168802228412256,
      "grad_norm": 1.6421277507855827,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8892314434051514,
      "num_tokens": 406316477.0,
      "step": 11679
    },
    {
      "epoch": 2.168987929433612,
      "grad_norm": 1.6019210654060907,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8922992944717407,
      "num_tokens": 406347876.0,
      "step": 11680
    },
    {
      "epoch": 2.1691736304549676,
      "grad_norm": 1.6009712964375475,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8794596791267395,
      "num_tokens": 406384279.0,
      "step": 11681
    },
    {
      "epoch": 2.1693593314763233,
      "grad_norm": 1.6970867919025634,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8847969770431519,
      "num_tokens": 406417476.0,
      "step": 11682
    },
    {
      "epoch": 2.1695450324976786,
      "grad_norm": 1.582193929350581,
      "learning_rate": 1e-06,
      "loss": 0.2775,
      "mean_token_accuracy": 0.9016096591949463,
      "num_tokens": 406449252.0,
      "step": 11683
    },
    {
      "epoch": 2.1697307335190343,
      "grad_norm": 1.7490964290080857,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8870583772659302,
      "num_tokens": 406476588.0,
      "step": 11684
    },
    {
      "epoch": 2.16991643454039,
      "grad_norm": 1.5153412462613962,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8876215815544128,
      "num_tokens": 406515334.0,
      "step": 11685
    },
    {
      "epoch": 2.170102135561746,
      "grad_norm": 1.580231750636459,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8881433010101318,
      "num_tokens": 406547052.0,
      "step": 11686
    },
    {
      "epoch": 2.170287836583101,
      "grad_norm": 1.6149429357487415,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.888941764831543,
      "num_tokens": 406577732.0,
      "step": 11687
    },
    {
      "epoch": 2.170473537604457,
      "grad_norm": 1.618868540368856,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8864215016365051,
      "num_tokens": 406609929.0,
      "step": 11688
    },
    {
      "epoch": 2.1706592386258126,
      "grad_norm": 1.5843431241911579,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8806501626968384,
      "num_tokens": 406645333.0,
      "step": 11689
    },
    {
      "epoch": 2.170844939647168,
      "grad_norm": 1.756267336861905,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.8908711075782776,
      "num_tokens": 406674860.0,
      "step": 11690
    },
    {
      "epoch": 2.1710306406685236,
      "grad_norm": 1.5414618724136109,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8866680860519409,
      "num_tokens": 406712426.0,
      "step": 11691
    },
    {
      "epoch": 2.1712163416898793,
      "grad_norm": 1.577976670925581,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8899382948875427,
      "num_tokens": 406745514.0,
      "step": 11692
    },
    {
      "epoch": 2.171402042711235,
      "grad_norm": 1.6047810399689133,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8798134326934814,
      "num_tokens": 406784022.0,
      "step": 11693
    },
    {
      "epoch": 2.1715877437325903,
      "grad_norm": 1.5669261467166349,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8823909759521484,
      "num_tokens": 406823456.0,
      "step": 11694
    },
    {
      "epoch": 2.171773444753946,
      "grad_norm": 1.7820132119294583,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8780859708786011,
      "num_tokens": 406855263.0,
      "step": 11695
    },
    {
      "epoch": 2.171959145775302,
      "grad_norm": 1.6095200190627226,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8841925859451294,
      "num_tokens": 406890224.0,
      "step": 11696
    },
    {
      "epoch": 2.1721448467966575,
      "grad_norm": 1.495077793506845,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8844853639602661,
      "num_tokens": 406928987.0,
      "step": 11697
    },
    {
      "epoch": 2.172330547818013,
      "grad_norm": 1.4916835494043736,
      "learning_rate": 1e-06,
      "loss": 0.2908,
      "mean_token_accuracy": 0.892615556716919,
      "num_tokens": 406960960.0,
      "step": 11698
    },
    {
      "epoch": 2.1725162488393686,
      "grad_norm": 1.592080944016676,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8900750875473022,
      "num_tokens": 406992145.0,
      "step": 11699
    },
    {
      "epoch": 2.1727019498607243,
      "grad_norm": 1.4653259956835432,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.894980251789093,
      "num_tokens": 407030746.0,
      "step": 11700
    },
    {
      "epoch": 2.17288765088208,
      "grad_norm": 1.691971084761875,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8771306276321411,
      "num_tokens": 407058945.0,
      "step": 11701
    },
    {
      "epoch": 2.1730733519034353,
      "grad_norm": 1.622329202966502,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8797669410705566,
      "num_tokens": 407091285.0,
      "step": 11702
    },
    {
      "epoch": 2.173259052924791,
      "grad_norm": 1.6847957384834755,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8880097270011902,
      "num_tokens": 407122473.0,
      "step": 11703
    },
    {
      "epoch": 2.173444753946147,
      "grad_norm": 1.4730017363694679,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8677788972854614,
      "num_tokens": 407165227.0,
      "step": 11704
    },
    {
      "epoch": 2.1736304549675025,
      "grad_norm": 1.6703887643400217,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8809496164321899,
      "num_tokens": 407195419.0,
      "step": 11705
    },
    {
      "epoch": 2.173816155988858,
      "grad_norm": 1.560302255112577,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8834553360939026,
      "num_tokens": 407230924.0,
      "step": 11706
    },
    {
      "epoch": 2.1740018570102135,
      "grad_norm": 1.6312805373434442,
      "learning_rate": 1e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.8772331476211548,
      "num_tokens": 407263249.0,
      "step": 11707
    },
    {
      "epoch": 2.1741875580315693,
      "grad_norm": 1.4361563616508746,
      "learning_rate": 1e-06,
      "loss": 0.2735,
      "mean_token_accuracy": 0.9046775102615356,
      "num_tokens": 407302266.0,
      "step": 11708
    },
    {
      "epoch": 2.174373259052925,
      "grad_norm": 1.4802843252342461,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8850916624069214,
      "num_tokens": 407345247.0,
      "step": 11709
    },
    {
      "epoch": 2.1745589600742803,
      "grad_norm": 1.6597953648777817,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8923326134681702,
      "num_tokens": 407376796.0,
      "step": 11710
    },
    {
      "epoch": 2.174744661095636,
      "grad_norm": 1.7271982603295393,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8903069496154785,
      "num_tokens": 407407401.0,
      "step": 11711
    },
    {
      "epoch": 2.1749303621169918,
      "grad_norm": 1.6774567755159104,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8721356391906738,
      "num_tokens": 407440280.0,
      "step": 11712
    },
    {
      "epoch": 2.175116063138347,
      "grad_norm": 1.6935618080138672,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8824126720428467,
      "num_tokens": 407471026.0,
      "step": 11713
    },
    {
      "epoch": 2.1753017641597028,
      "grad_norm": 1.6736518737194435,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8810527920722961,
      "num_tokens": 407502402.0,
      "step": 11714
    },
    {
      "epoch": 2.1754874651810585,
      "grad_norm": 1.568132061911594,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8797996044158936,
      "num_tokens": 407536228.0,
      "step": 11715
    },
    {
      "epoch": 2.1756731662024142,
      "grad_norm": 1.8551717512433716,
      "learning_rate": 1e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8575643301010132,
      "num_tokens": 407564601.0,
      "step": 11716
    },
    {
      "epoch": 2.1758588672237695,
      "grad_norm": 1.4609091584918041,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8826881647109985,
      "num_tokens": 407606378.0,
      "step": 11717
    },
    {
      "epoch": 2.1760445682451253,
      "grad_norm": 1.5146283534617857,
      "learning_rate": 1e-06,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9079827070236206,
      "num_tokens": 407636160.0,
      "step": 11718
    },
    {
      "epoch": 2.176230269266481,
      "grad_norm": 1.5906168510028444,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.885375440120697,
      "num_tokens": 407666384.0,
      "step": 11719
    },
    {
      "epoch": 2.1764159702878367,
      "grad_norm": 1.43780772379913,
      "learning_rate": 1e-06,
      "loss": 0.2909,
      "mean_token_accuracy": 0.8979142308235168,
      "num_tokens": 407704064.0,
      "step": 11720
    },
    {
      "epoch": 2.176601671309192,
      "grad_norm": 1.5739786014118788,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8796271085739136,
      "num_tokens": 407741003.0,
      "step": 11721
    },
    {
      "epoch": 2.1767873723305478,
      "grad_norm": 1.6742828254208668,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8811733722686768,
      "num_tokens": 407772787.0,
      "step": 11722
    },
    {
      "epoch": 2.1769730733519035,
      "grad_norm": 1.5019961150193757,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8828333020210266,
      "num_tokens": 407809877.0,
      "step": 11723
    },
    {
      "epoch": 2.177158774373259,
      "grad_norm": 1.6060371623487384,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8936021327972412,
      "num_tokens": 407844019.0,
      "step": 11724
    },
    {
      "epoch": 2.1773444753946145,
      "grad_norm": 1.495448128667933,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8815107941627502,
      "num_tokens": 407883274.0,
      "step": 11725
    },
    {
      "epoch": 2.1775301764159702,
      "grad_norm": 1.596092531671599,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8909286260604858,
      "num_tokens": 407916476.0,
      "step": 11726
    },
    {
      "epoch": 2.177715877437326,
      "grad_norm": 1.5737768671211696,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8922535181045532,
      "num_tokens": 407952590.0,
      "step": 11727
    },
    {
      "epoch": 2.1779015784586817,
      "grad_norm": 1.6498045385811715,
      "learning_rate": 1e-06,
      "loss": 0.384,
      "mean_token_accuracy": 0.8694945573806763,
      "num_tokens": 407990741.0,
      "step": 11728
    },
    {
      "epoch": 2.178087279480037,
      "grad_norm": 1.4497199542832069,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8863305449485779,
      "num_tokens": 408033459.0,
      "step": 11729
    },
    {
      "epoch": 2.1782729805013927,
      "grad_norm": 1.549976935588261,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8913540840148926,
      "num_tokens": 408067986.0,
      "step": 11730
    },
    {
      "epoch": 2.1784586815227485,
      "grad_norm": 1.50901819258204,
      "learning_rate": 1e-06,
      "loss": 0.2871,
      "mean_token_accuracy": 0.8968956470489502,
      "num_tokens": 408103433.0,
      "step": 11731
    },
    {
      "epoch": 2.178644382544104,
      "grad_norm": 1.4855847002661084,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.8791409730911255,
      "num_tokens": 408145062.0,
      "step": 11732
    },
    {
      "epoch": 2.1788300835654595,
      "grad_norm": 1.490959487857966,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8848362565040588,
      "num_tokens": 408182863.0,
      "step": 11733
    },
    {
      "epoch": 2.179015784586815,
      "grad_norm": 1.667292658988168,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8850803375244141,
      "num_tokens": 408213886.0,
      "step": 11734
    },
    {
      "epoch": 2.179201485608171,
      "grad_norm": 1.730087983572203,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8964974880218506,
      "num_tokens": 408242957.0,
      "step": 11735
    },
    {
      "epoch": 2.1793871866295262,
      "grad_norm": 1.451552071875697,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8722068667411804,
      "num_tokens": 408286209.0,
      "step": 11736
    },
    {
      "epoch": 2.179572887650882,
      "grad_norm": 1.541362679259722,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8880661725997925,
      "num_tokens": 408318799.0,
      "step": 11737
    },
    {
      "epoch": 2.1797585886722377,
      "grad_norm": 1.6773554452553405,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8885502815246582,
      "num_tokens": 408350265.0,
      "step": 11738
    },
    {
      "epoch": 2.1799442896935934,
      "grad_norm": 1.732196333548394,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8854386806488037,
      "num_tokens": 408382629.0,
      "step": 11739
    },
    {
      "epoch": 2.1801299907149487,
      "grad_norm": 1.5768532325877633,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8779534697532654,
      "num_tokens": 408419199.0,
      "step": 11740
    },
    {
      "epoch": 2.1803156917363045,
      "grad_norm": 1.6230413922857796,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.879455029964447,
      "num_tokens": 408456218.0,
      "step": 11741
    },
    {
      "epoch": 2.18050139275766,
      "grad_norm": 1.8151417361509898,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8926313519477844,
      "num_tokens": 408478700.0,
      "step": 11742
    },
    {
      "epoch": 2.180687093779016,
      "grad_norm": 1.6471155057819131,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8921184539794922,
      "num_tokens": 408510187.0,
      "step": 11743
    },
    {
      "epoch": 2.180872794800371,
      "grad_norm": 1.6436778261960825,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8884751796722412,
      "num_tokens": 408545689.0,
      "step": 11744
    },
    {
      "epoch": 2.181058495821727,
      "grad_norm": 1.5458407994031942,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8871248364448547,
      "num_tokens": 408582244.0,
      "step": 11745
    },
    {
      "epoch": 2.1812441968430827,
      "grad_norm": 1.588514036035574,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8850130438804626,
      "num_tokens": 408617526.0,
      "step": 11746
    },
    {
      "epoch": 2.1814298978644384,
      "grad_norm": 1.5500911142003182,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8754976987838745,
      "num_tokens": 408655227.0,
      "step": 11747
    },
    {
      "epoch": 2.1816155988857937,
      "grad_norm": 1.5767201616289437,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8891832828521729,
      "num_tokens": 408690746.0,
      "step": 11748
    },
    {
      "epoch": 2.1818012999071494,
      "grad_norm": 1.5050626731821464,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8864777684211731,
      "num_tokens": 408728177.0,
      "step": 11749
    },
    {
      "epoch": 2.181987000928505,
      "grad_norm": 1.7313724387660157,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8857358694076538,
      "num_tokens": 408756719.0,
      "step": 11750
    },
    {
      "epoch": 2.182172701949861,
      "grad_norm": 1.4480760756921789,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8797770738601685,
      "num_tokens": 408799206.0,
      "step": 11751
    },
    {
      "epoch": 2.182358402971216,
      "grad_norm": 1.6029337166773634,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8791242837905884,
      "num_tokens": 408834761.0,
      "step": 11752
    },
    {
      "epoch": 2.182544103992572,
      "grad_norm": 1.654679179905958,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.883325457572937,
      "num_tokens": 408866388.0,
      "step": 11753
    },
    {
      "epoch": 2.1827298050139277,
      "grad_norm": 1.6039141426933317,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8901825547218323,
      "num_tokens": 408899220.0,
      "step": 11754
    },
    {
      "epoch": 2.1829155060352834,
      "grad_norm": 1.5627368992187127,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.889889121055603,
      "num_tokens": 408940425.0,
      "step": 11755
    },
    {
      "epoch": 2.1831012070566387,
      "grad_norm": 1.737986160518385,
      "learning_rate": 1e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.857288122177124,
      "num_tokens": 408976335.0,
      "step": 11756
    },
    {
      "epoch": 2.1832869080779944,
      "grad_norm": 1.5272375589527476,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8924522995948792,
      "num_tokens": 409012543.0,
      "step": 11757
    },
    {
      "epoch": 2.18347260909935,
      "grad_norm": 1.7368138737857237,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8929377794265747,
      "num_tokens": 409043564.0,
      "step": 11758
    },
    {
      "epoch": 2.1836583101207054,
      "grad_norm": 1.4384774176421355,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8860253095626831,
      "num_tokens": 409084305.0,
      "step": 11759
    },
    {
      "epoch": 2.183844011142061,
      "grad_norm": 1.365454931006,
      "learning_rate": 1e-06,
      "loss": 0.2818,
      "mean_token_accuracy": 0.89763343334198,
      "num_tokens": 409123453.0,
      "step": 11760
    },
    {
      "epoch": 2.184029712163417,
      "grad_norm": 1.4938799184217515,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8920440077781677,
      "num_tokens": 409159598.0,
      "step": 11761
    },
    {
      "epoch": 2.1842154131847726,
      "grad_norm": 1.666730974318469,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8755941390991211,
      "num_tokens": 409193907.0,
      "step": 11762
    },
    {
      "epoch": 2.1844011142061284,
      "grad_norm": 1.5411041954164793,
      "learning_rate": 1e-06,
      "loss": 0.2815,
      "mean_token_accuracy": 0.9026957154273987,
      "num_tokens": 409225155.0,
      "step": 11763
    },
    {
      "epoch": 2.1845868152274837,
      "grad_norm": 1.7150149338963436,
      "learning_rate": 1e-06,
      "loss": 0.2992,
      "mean_token_accuracy": 0.8964397311210632,
      "num_tokens": 409256081.0,
      "step": 11764
    },
    {
      "epoch": 2.1847725162488394,
      "grad_norm": 1.4374880231033063,
      "learning_rate": 1e-06,
      "loss": 0.2613,
      "mean_token_accuracy": 0.906947135925293,
      "num_tokens": 409290910.0,
      "step": 11765
    },
    {
      "epoch": 2.184958217270195,
      "grad_norm": 1.5586745206932362,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8879925608634949,
      "num_tokens": 409326304.0,
      "step": 11766
    },
    {
      "epoch": 2.1851439182915504,
      "grad_norm": 1.479482147724449,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8804723024368286,
      "num_tokens": 409365447.0,
      "step": 11767
    },
    {
      "epoch": 2.185329619312906,
      "grad_norm": 1.533262480026759,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.884694516658783,
      "num_tokens": 409402855.0,
      "step": 11768
    },
    {
      "epoch": 2.185515320334262,
      "grad_norm": 1.4837702486822513,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8835959434509277,
      "num_tokens": 409443360.0,
      "step": 11769
    },
    {
      "epoch": 2.1857010213556176,
      "grad_norm": 1.6465299462145668,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.884005069732666,
      "num_tokens": 409477527.0,
      "step": 11770
    },
    {
      "epoch": 2.185886722376973,
      "grad_norm": 1.7252229284247185,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8912485837936401,
      "num_tokens": 409508347.0,
      "step": 11771
    },
    {
      "epoch": 2.1860724233983286,
      "grad_norm": 1.7593819942032194,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8818418979644775,
      "num_tokens": 409538296.0,
      "step": 11772
    },
    {
      "epoch": 2.1862581244196844,
      "grad_norm": 1.5220613708858939,
      "learning_rate": 1e-06,
      "loss": 0.2809,
      "mean_token_accuracy": 0.8994089365005493,
      "num_tokens": 409572886.0,
      "step": 11773
    },
    {
      "epoch": 2.18644382544104,
      "grad_norm": 1.576176528520037,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8878825306892395,
      "num_tokens": 409608914.0,
      "step": 11774
    },
    {
      "epoch": 2.1866295264623954,
      "grad_norm": 1.4981185168608542,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8823946714401245,
      "num_tokens": 409645437.0,
      "step": 11775
    },
    {
      "epoch": 2.186815227483751,
      "grad_norm": 1.5916655175430845,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8839104771614075,
      "num_tokens": 409681321.0,
      "step": 11776
    },
    {
      "epoch": 2.187000928505107,
      "grad_norm": 1.4830185500113313,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8942093253135681,
      "num_tokens": 409720681.0,
      "step": 11777
    },
    {
      "epoch": 2.1871866295264626,
      "grad_norm": 1.5326081079804565,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.8988446593284607,
      "num_tokens": 409755594.0,
      "step": 11778
    },
    {
      "epoch": 2.187372330547818,
      "grad_norm": 1.5210224738991822,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8811759948730469,
      "num_tokens": 409790667.0,
      "step": 11779
    },
    {
      "epoch": 2.1875580315691736,
      "grad_norm": 1.6102985537636145,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8788309097290039,
      "num_tokens": 409823775.0,
      "step": 11780
    },
    {
      "epoch": 2.1877437325905293,
      "grad_norm": 1.6229060935515016,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8765007257461548,
      "num_tokens": 409862394.0,
      "step": 11781
    },
    {
      "epoch": 2.1879294336118846,
      "grad_norm": 1.4389208199115175,
      "learning_rate": 1e-06,
      "loss": 0.2964,
      "mean_token_accuracy": 0.8951059579849243,
      "num_tokens": 409905756.0,
      "step": 11782
    },
    {
      "epoch": 2.1881151346332404,
      "grad_norm": 1.8572814544660605,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8727420568466187,
      "num_tokens": 409933805.0,
      "step": 11783
    },
    {
      "epoch": 2.188300835654596,
      "grad_norm": 1.4992136228809745,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8929364681243896,
      "num_tokens": 409969977.0,
      "step": 11784
    },
    {
      "epoch": 2.188486536675952,
      "grad_norm": 1.5923648671076727,
      "learning_rate": 1e-06,
      "loss": 0.3246,
      "mean_token_accuracy": 0.8839234113693237,
      "num_tokens": 410007963.0,
      "step": 11785
    },
    {
      "epoch": 2.1886722376973076,
      "grad_norm": 1.6739527701141288,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8926092386245728,
      "num_tokens": 410038310.0,
      "step": 11786
    },
    {
      "epoch": 2.188857938718663,
      "grad_norm": 1.7383618917325434,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.8635373115539551,
      "num_tokens": 410072095.0,
      "step": 11787
    },
    {
      "epoch": 2.1890436397400186,
      "grad_norm": 1.5751046696946434,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8805763721466064,
      "num_tokens": 410110180.0,
      "step": 11788
    },
    {
      "epoch": 2.1892293407613743,
      "grad_norm": 1.6260488596213643,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8896538019180298,
      "num_tokens": 410143422.0,
      "step": 11789
    },
    {
      "epoch": 2.1894150417827296,
      "grad_norm": 1.5848733889903146,
      "learning_rate": 1e-06,
      "loss": 0.2972,
      "mean_token_accuracy": 0.8950058221817017,
      "num_tokens": 410174202.0,
      "step": 11790
    },
    {
      "epoch": 2.1896007428040853,
      "grad_norm": 1.5532021233236128,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8869491219520569,
      "num_tokens": 410209747.0,
      "step": 11791
    },
    {
      "epoch": 2.189786443825441,
      "grad_norm": 1.5765303791648682,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8880296349525452,
      "num_tokens": 410244654.0,
      "step": 11792
    },
    {
      "epoch": 2.189972144846797,
      "grad_norm": 1.52241708350552,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.890896201133728,
      "num_tokens": 410283047.0,
      "step": 11793
    },
    {
      "epoch": 2.190157845868152,
      "grad_norm": 1.7172347656860434,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8817886114120483,
      "num_tokens": 410316132.0,
      "step": 11794
    },
    {
      "epoch": 2.190343546889508,
      "grad_norm": 1.5590558650065316,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8949415683746338,
      "num_tokens": 410348547.0,
      "step": 11795
    },
    {
      "epoch": 2.1905292479108636,
      "grad_norm": 1.551013309430462,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8818937540054321,
      "num_tokens": 410385604.0,
      "step": 11796
    },
    {
      "epoch": 2.1907149489322193,
      "grad_norm": 1.4819749194030751,
      "learning_rate": 1e-06,
      "loss": 0.2785,
      "mean_token_accuracy": 0.8993625640869141,
      "num_tokens": 410420340.0,
      "step": 11797
    },
    {
      "epoch": 2.1909006499535746,
      "grad_norm": 1.5972232019242592,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8833217620849609,
      "num_tokens": 410453961.0,
      "step": 11798
    },
    {
      "epoch": 2.1910863509749303,
      "grad_norm": 1.5266980665310783,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8888108134269714,
      "num_tokens": 410490379.0,
      "step": 11799
    },
    {
      "epoch": 2.191272051996286,
      "grad_norm": 1.36267447159112,
      "learning_rate": 1e-06,
      "loss": 0.2795,
      "mean_token_accuracy": 0.902929425239563,
      "num_tokens": 410531521.0,
      "step": 11800
    },
    {
      "epoch": 2.191457753017642,
      "grad_norm": 1.5520923113029035,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8857349157333374,
      "num_tokens": 410565304.0,
      "step": 11801
    },
    {
      "epoch": 2.191643454038997,
      "grad_norm": 1.5491188739263848,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8850210905075073,
      "num_tokens": 410599291.0,
      "step": 11802
    },
    {
      "epoch": 2.191829155060353,
      "grad_norm": 1.6566267346303274,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8897053599357605,
      "num_tokens": 410630285.0,
      "step": 11803
    },
    {
      "epoch": 2.1920148560817085,
      "grad_norm": 1.6017960178351796,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8891953229904175,
      "num_tokens": 410663816.0,
      "step": 11804
    },
    {
      "epoch": 2.1922005571030643,
      "grad_norm": 1.5126912390664262,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8887307643890381,
      "num_tokens": 410703045.0,
      "step": 11805
    },
    {
      "epoch": 2.1923862581244196,
      "grad_norm": 1.6998577226519656,
      "learning_rate": 1e-06,
      "loss": 0.3018,
      "mean_token_accuracy": 0.8927175998687744,
      "num_tokens": 410730575.0,
      "step": 11806
    },
    {
      "epoch": 2.1925719591457753,
      "grad_norm": 1.3708742005982721,
      "learning_rate": 1e-06,
      "loss": 0.3018,
      "mean_token_accuracy": 0.893916666507721,
      "num_tokens": 410769458.0,
      "step": 11807
    },
    {
      "epoch": 2.192757660167131,
      "grad_norm": 1.7198245787356914,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8808802366256714,
      "num_tokens": 410798000.0,
      "step": 11808
    },
    {
      "epoch": 2.1929433611884868,
      "grad_norm": 1.6707352282358392,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.887373685836792,
      "num_tokens": 410829473.0,
      "step": 11809
    },
    {
      "epoch": 2.193129062209842,
      "grad_norm": 1.4393907639664103,
      "learning_rate": 1e-06,
      "loss": 0.2867,
      "mean_token_accuracy": 0.8992553949356079,
      "num_tokens": 410866193.0,
      "step": 11810
    },
    {
      "epoch": 2.193314763231198,
      "grad_norm": 1.7131353467881862,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8880407810211182,
      "num_tokens": 410902403.0,
      "step": 11811
    },
    {
      "epoch": 2.1935004642525535,
      "grad_norm": 1.6075879973303429,
      "learning_rate": 1e-06,
      "loss": 0.2814,
      "mean_token_accuracy": 0.8995245695114136,
      "num_tokens": 410931986.0,
      "step": 11812
    },
    {
      "epoch": 2.193686165273909,
      "grad_norm": 1.5222978227068675,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.894378125667572,
      "num_tokens": 410965854.0,
      "step": 11813
    },
    {
      "epoch": 2.1938718662952645,
      "grad_norm": 1.4572288933327389,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8869491815567017,
      "num_tokens": 411009873.0,
      "step": 11814
    },
    {
      "epoch": 2.1940575673166203,
      "grad_norm": 1.5353945609546624,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8696085810661316,
      "num_tokens": 411048461.0,
      "step": 11815
    },
    {
      "epoch": 2.194243268337976,
      "grad_norm": 1.5241028790765743,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8899143934249878,
      "num_tokens": 411085939.0,
      "step": 11816
    },
    {
      "epoch": 2.1944289693593313,
      "grad_norm": 1.6072079741694933,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8821831345558167,
      "num_tokens": 411119113.0,
      "step": 11817
    },
    {
      "epoch": 2.194614670380687,
      "grad_norm": 1.6245119288929168,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8829861283302307,
      "num_tokens": 411154177.0,
      "step": 11818
    },
    {
      "epoch": 2.1948003714020428,
      "grad_norm": 1.4806736073545588,
      "learning_rate": 1e-06,
      "loss": 0.2882,
      "mean_token_accuracy": 0.8954348564147949,
      "num_tokens": 411190648.0,
      "step": 11819
    },
    {
      "epoch": 2.1949860724233985,
      "grad_norm": 1.5992341443580458,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.889477550983429,
      "num_tokens": 411223895.0,
      "step": 11820
    },
    {
      "epoch": 2.195171773444754,
      "grad_norm": 1.703499462252178,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8771226406097412,
      "num_tokens": 411255298.0,
      "step": 11821
    },
    {
      "epoch": 2.1953574744661095,
      "grad_norm": 1.6031316864845369,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8846049308776855,
      "num_tokens": 411290094.0,
      "step": 11822
    },
    {
      "epoch": 2.1955431754874652,
      "grad_norm": 1.4451062337610316,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8905032277107239,
      "num_tokens": 411330276.0,
      "step": 11823
    },
    {
      "epoch": 2.195728876508821,
      "grad_norm": 1.5405057127801811,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8920755386352539,
      "num_tokens": 411365109.0,
      "step": 11824
    },
    {
      "epoch": 2.1959145775301763,
      "grad_norm": 1.65162874544624,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8791310787200928,
      "num_tokens": 411400111.0,
      "step": 11825
    },
    {
      "epoch": 2.196100278551532,
      "grad_norm": 1.5521300643005138,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8892718553543091,
      "num_tokens": 411436451.0,
      "step": 11826
    },
    {
      "epoch": 2.1962859795728877,
      "grad_norm": 1.5246765938021811,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8894510269165039,
      "num_tokens": 411474785.0,
      "step": 11827
    },
    {
      "epoch": 2.1964716805942435,
      "grad_norm": 1.5857416697205893,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.893621027469635,
      "num_tokens": 411508590.0,
      "step": 11828
    },
    {
      "epoch": 2.1966573816155988,
      "grad_norm": 1.5786495989409473,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8709887266159058,
      "num_tokens": 411544640.0,
      "step": 11829
    },
    {
      "epoch": 2.1968430826369545,
      "grad_norm": 1.582589722651501,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8919587135314941,
      "num_tokens": 411578172.0,
      "step": 11830
    },
    {
      "epoch": 2.1970287836583102,
      "grad_norm": 1.513400868942596,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.887312114238739,
      "num_tokens": 411613488.0,
      "step": 11831
    },
    {
      "epoch": 2.197214484679666,
      "grad_norm": 1.6555379045291694,
      "learning_rate": 1e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8751640319824219,
      "num_tokens": 411645750.0,
      "step": 11832
    },
    {
      "epoch": 2.1974001857010212,
      "grad_norm": 1.4862611404923594,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8863337635993958,
      "num_tokens": 411685662.0,
      "step": 11833
    },
    {
      "epoch": 2.197585886722377,
      "grad_norm": 1.4700891050237714,
      "learning_rate": 1e-06,
      "loss": 0.2724,
      "mean_token_accuracy": 0.90212082862854,
      "num_tokens": 411721967.0,
      "step": 11834
    },
    {
      "epoch": 2.1977715877437327,
      "grad_norm": 1.4972256344997206,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8971428871154785,
      "num_tokens": 411758843.0,
      "step": 11835
    },
    {
      "epoch": 2.197957288765088,
      "grad_norm": 1.7699842583300562,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8772873878479004,
      "num_tokens": 411789906.0,
      "step": 11836
    },
    {
      "epoch": 2.1981429897864437,
      "grad_norm": 1.592425328748173,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8768553137779236,
      "num_tokens": 411823303.0,
      "step": 11837
    },
    {
      "epoch": 2.1983286908077995,
      "grad_norm": 1.5337594229950675,
      "learning_rate": 1e-06,
      "loss": 0.2774,
      "mean_token_accuracy": 0.9015334844589233,
      "num_tokens": 411854851.0,
      "step": 11838
    },
    {
      "epoch": 2.198514391829155,
      "grad_norm": 1.4892405262073454,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8819270730018616,
      "num_tokens": 411896743.0,
      "step": 11839
    },
    {
      "epoch": 2.1987000928505105,
      "grad_norm": 1.6398390911266314,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8864678144454956,
      "num_tokens": 411928220.0,
      "step": 11840
    },
    {
      "epoch": 2.1988857938718662,
      "grad_norm": 1.5436003152337436,
      "learning_rate": 1e-06,
      "loss": 0.2712,
      "mean_token_accuracy": 0.9028885364532471,
      "num_tokens": 411957785.0,
      "step": 11841
    },
    {
      "epoch": 2.199071494893222,
      "grad_norm": 1.5468449628072765,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8917665481567383,
      "num_tokens": 411994159.0,
      "step": 11842
    },
    {
      "epoch": 2.1992571959145777,
      "grad_norm": 1.6573641997854958,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8762117028236389,
      "num_tokens": 412028861.0,
      "step": 11843
    },
    {
      "epoch": 2.199442896935933,
      "grad_norm": 1.6832261899315917,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8882883787155151,
      "num_tokens": 412059403.0,
      "step": 11844
    },
    {
      "epoch": 2.1996285979572887,
      "grad_norm": 1.3964779117769544,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8940787315368652,
      "num_tokens": 412100734.0,
      "step": 11845
    },
    {
      "epoch": 2.1998142989786444,
      "grad_norm": 1.5246153887174445,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8903002142906189,
      "num_tokens": 412135086.0,
      "step": 11846
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.6883092085492681,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8759942054748535,
      "num_tokens": 412169078.0,
      "step": 11847
    },
    {
      "epoch": 2.2001857010213555,
      "grad_norm": 1.618428034290727,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8911639451980591,
      "num_tokens": 412204289.0,
      "step": 11848
    },
    {
      "epoch": 2.200371402042711,
      "grad_norm": 1.6213090205236795,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8771413564682007,
      "num_tokens": 412238404.0,
      "step": 11849
    },
    {
      "epoch": 2.200557103064067,
      "grad_norm": 1.4771860489044182,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8805691003799438,
      "num_tokens": 412282918.0,
      "step": 11850
    },
    {
      "epoch": 2.2007428040854227,
      "grad_norm": 1.5702911932238313,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8916884660720825,
      "num_tokens": 412316218.0,
      "step": 11851
    },
    {
      "epoch": 2.200928505106778,
      "grad_norm": 1.6052592120788622,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8893910646438599,
      "num_tokens": 412350776.0,
      "step": 11852
    },
    {
      "epoch": 2.2011142061281337,
      "grad_norm": 1.5459513223941272,
      "learning_rate": 1e-06,
      "loss": 0.2705,
      "mean_token_accuracy": 0.9031836986541748,
      "num_tokens": 412385285.0,
      "step": 11853
    },
    {
      "epoch": 2.2012999071494894,
      "grad_norm": 1.5905572998194732,
      "learning_rate": 1e-06,
      "loss": 0.3102,
      "mean_token_accuracy": 0.8927498459815979,
      "num_tokens": 412419781.0,
      "step": 11854
    },
    {
      "epoch": 2.201485608170845,
      "grad_norm": 1.6435973508641468,
      "learning_rate": 1e-06,
      "loss": 0.3536,
      "mean_token_accuracy": 0.8746141195297241,
      "num_tokens": 412453429.0,
      "step": 11855
    },
    {
      "epoch": 2.2016713091922004,
      "grad_norm": 1.7058005821693012,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8773471117019653,
      "num_tokens": 412482268.0,
      "step": 11856
    },
    {
      "epoch": 2.201857010213556,
      "grad_norm": 1.5503122712462065,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8928617835044861,
      "num_tokens": 412514140.0,
      "step": 11857
    },
    {
      "epoch": 2.202042711234912,
      "grad_norm": 1.561515290359034,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8912604451179504,
      "num_tokens": 412548582.0,
      "step": 11858
    },
    {
      "epoch": 2.202228412256267,
      "grad_norm": 1.569243099525651,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8930739164352417,
      "num_tokens": 412584176.0,
      "step": 11859
    },
    {
      "epoch": 2.202414113277623,
      "grad_norm": 1.5331074983077142,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8815933465957642,
      "num_tokens": 412621521.0,
      "step": 11860
    },
    {
      "epoch": 2.2025998142989787,
      "grad_norm": 1.624182136549718,
      "learning_rate": 1e-06,
      "loss": 0.2977,
      "mean_token_accuracy": 0.8920589685440063,
      "num_tokens": 412650911.0,
      "step": 11861
    },
    {
      "epoch": 2.2027855153203344,
      "grad_norm": 1.3639091265390155,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.8947495222091675,
      "num_tokens": 412695738.0,
      "step": 11862
    },
    {
      "epoch": 2.2029712163416897,
      "grad_norm": 1.5604688405024383,
      "learning_rate": 1e-06,
      "loss": 0.2924,
      "mean_token_accuracy": 0.8968231678009033,
      "num_tokens": 412728793.0,
      "step": 11863
    },
    {
      "epoch": 2.2031569173630454,
      "grad_norm": 1.6627691515373906,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8798331022262573,
      "num_tokens": 412761356.0,
      "step": 11864
    },
    {
      "epoch": 2.203342618384401,
      "grad_norm": 1.5241031221618737,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8909769058227539,
      "num_tokens": 412800663.0,
      "step": 11865
    },
    {
      "epoch": 2.203528319405757,
      "grad_norm": 1.5575686512992062,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8834983706474304,
      "num_tokens": 412835213.0,
      "step": 11866
    },
    {
      "epoch": 2.203714020427112,
      "grad_norm": 1.629451986524325,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8671119213104248,
      "num_tokens": 412872793.0,
      "step": 11867
    },
    {
      "epoch": 2.203899721448468,
      "grad_norm": 1.6940435239874432,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8770478963851929,
      "num_tokens": 412905682.0,
      "step": 11868
    },
    {
      "epoch": 2.2040854224698236,
      "grad_norm": 1.5710626619371535,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8751705884933472,
      "num_tokens": 412946292.0,
      "step": 11869
    },
    {
      "epoch": 2.2042711234911794,
      "grad_norm": 1.4500542310173175,
      "learning_rate": 1e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8703120946884155,
      "num_tokens": 412990172.0,
      "step": 11870
    },
    {
      "epoch": 2.2044568245125347,
      "grad_norm": 1.6191963110644085,
      "learning_rate": 1e-06,
      "loss": 0.2988,
      "mean_token_accuracy": 0.8944528102874756,
      "num_tokens": 413020009.0,
      "step": 11871
    },
    {
      "epoch": 2.2046425255338904,
      "grad_norm": 1.650119579225218,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8788114190101624,
      "num_tokens": 413053979.0,
      "step": 11872
    },
    {
      "epoch": 2.204828226555246,
      "grad_norm": 1.4619356233503147,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8943678140640259,
      "num_tokens": 413096130.0,
      "step": 11873
    },
    {
      "epoch": 2.205013927576602,
      "grad_norm": 1.7828774832471008,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.892314076423645,
      "num_tokens": 413121881.0,
      "step": 11874
    },
    {
      "epoch": 2.205199628597957,
      "grad_norm": 1.5551751304258306,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8939094543457031,
      "num_tokens": 413157465.0,
      "step": 11875
    },
    {
      "epoch": 2.205385329619313,
      "grad_norm": 1.580136977580787,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8856084942817688,
      "num_tokens": 413191266.0,
      "step": 11876
    },
    {
      "epoch": 2.2055710306406686,
      "grad_norm": 1.5356364843273091,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.897260308265686,
      "num_tokens": 413224946.0,
      "step": 11877
    },
    {
      "epoch": 2.2057567316620244,
      "grad_norm": 1.6611598433795862,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8782877326011658,
      "num_tokens": 413256424.0,
      "step": 11878
    },
    {
      "epoch": 2.2059424326833796,
      "grad_norm": 1.5514971537779993,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8857933878898621,
      "num_tokens": 413289030.0,
      "step": 11879
    },
    {
      "epoch": 2.2061281337047354,
      "grad_norm": 1.6114598155828008,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8679053783416748,
      "num_tokens": 413324815.0,
      "step": 11880
    },
    {
      "epoch": 2.206313834726091,
      "grad_norm": 1.5972830016604234,
      "learning_rate": 1e-06,
      "loss": 0.2933,
      "mean_token_accuracy": 0.8967569470405579,
      "num_tokens": 413357186.0,
      "step": 11881
    },
    {
      "epoch": 2.2064995357474464,
      "grad_norm": 1.5588597377286386,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8928087949752808,
      "num_tokens": 413392501.0,
      "step": 11882
    },
    {
      "epoch": 2.206685236768802,
      "grad_norm": 1.5537909953366869,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8840492367744446,
      "num_tokens": 413428416.0,
      "step": 11883
    },
    {
      "epoch": 2.206870937790158,
      "grad_norm": 1.7168077375136135,
      "learning_rate": 1e-06,
      "loss": 0.2861,
      "mean_token_accuracy": 0.8984674215316772,
      "num_tokens": 413455865.0,
      "step": 11884
    },
    {
      "epoch": 2.2070566388115136,
      "grad_norm": 1.4874688831843197,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8804928064346313,
      "num_tokens": 413500457.0,
      "step": 11885
    },
    {
      "epoch": 2.207242339832869,
      "grad_norm": 1.5918007752654515,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8880117535591125,
      "num_tokens": 413533937.0,
      "step": 11886
    },
    {
      "epoch": 2.2074280408542246,
      "grad_norm": 1.5031671458907057,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8904384970664978,
      "num_tokens": 413573613.0,
      "step": 11887
    },
    {
      "epoch": 2.2076137418755803,
      "grad_norm": 1.5122861457235643,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8884494304656982,
      "num_tokens": 413611141.0,
      "step": 11888
    },
    {
      "epoch": 2.207799442896936,
      "grad_norm": 1.5119726984379394,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8895554542541504,
      "num_tokens": 413646530.0,
      "step": 11889
    },
    {
      "epoch": 2.2079851439182914,
      "grad_norm": 1.6185710958069643,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8841767907142639,
      "num_tokens": 413679952.0,
      "step": 11890
    },
    {
      "epoch": 2.208170844939647,
      "grad_norm": 1.494274210156857,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.877181351184845,
      "num_tokens": 413720176.0,
      "step": 11891
    },
    {
      "epoch": 2.208356545961003,
      "grad_norm": 1.5668731191770084,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8793959617614746,
      "num_tokens": 413757520.0,
      "step": 11892
    },
    {
      "epoch": 2.2085422469823586,
      "grad_norm": 1.6297688389002598,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.882559597492218,
      "num_tokens": 413793858.0,
      "step": 11893
    },
    {
      "epoch": 2.208727948003714,
      "grad_norm": 1.6106369080002372,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8901845812797546,
      "num_tokens": 413828855.0,
      "step": 11894
    },
    {
      "epoch": 2.2089136490250696,
      "grad_norm": 1.5778380617412562,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8839820623397827,
      "num_tokens": 413866071.0,
      "step": 11895
    },
    {
      "epoch": 2.2090993500464253,
      "grad_norm": 1.702838805065592,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8770763278007507,
      "num_tokens": 413897697.0,
      "step": 11896
    },
    {
      "epoch": 2.209285051067781,
      "grad_norm": 1.6189791294710982,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8803350329399109,
      "num_tokens": 413931540.0,
      "step": 11897
    },
    {
      "epoch": 2.2094707520891363,
      "grad_norm": 1.5693189466509398,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8795878887176514,
      "num_tokens": 413965714.0,
      "step": 11898
    },
    {
      "epoch": 2.209656453110492,
      "grad_norm": 1.5699422187997758,
      "learning_rate": 1e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8951551914215088,
      "num_tokens": 414001464.0,
      "step": 11899
    },
    {
      "epoch": 2.209842154131848,
      "grad_norm": 1.6275590275370693,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8871607780456543,
      "num_tokens": 414033722.0,
      "step": 11900
    },
    {
      "epoch": 2.2100278551532035,
      "grad_norm": 1.6127064091031154,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8979911804199219,
      "num_tokens": 414062529.0,
      "step": 11901
    },
    {
      "epoch": 2.210213556174559,
      "grad_norm": 1.5079064582814443,
      "learning_rate": 1e-06,
      "loss": 0.2672,
      "mean_token_accuracy": 0.9047238826751709,
      "num_tokens": 414095203.0,
      "step": 11902
    },
    {
      "epoch": 2.2103992571959146,
      "grad_norm": 1.5026804405073828,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8834214210510254,
      "num_tokens": 414132468.0,
      "step": 11903
    },
    {
      "epoch": 2.2105849582172703,
      "grad_norm": 1.6648234660303134,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8846637606620789,
      "num_tokens": 414167901.0,
      "step": 11904
    },
    {
      "epoch": 2.2107706592386256,
      "grad_norm": 1.6929958211715077,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8822542428970337,
      "num_tokens": 414198326.0,
      "step": 11905
    },
    {
      "epoch": 2.2109563602599813,
      "grad_norm": 1.5209433509434998,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8749845027923584,
      "num_tokens": 414238602.0,
      "step": 11906
    },
    {
      "epoch": 2.211142061281337,
      "grad_norm": 1.7209942923456314,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8810465335845947,
      "num_tokens": 414270713.0,
      "step": 11907
    },
    {
      "epoch": 2.211327762302693,
      "grad_norm": 1.4917722521603227,
      "learning_rate": 1e-06,
      "loss": 0.2937,
      "mean_token_accuracy": 0.8924636840820312,
      "num_tokens": 414305760.0,
      "step": 11908
    },
    {
      "epoch": 2.211513463324048,
      "grad_norm": 1.5227969946620272,
      "learning_rate": 1e-06,
      "loss": 0.288,
      "mean_token_accuracy": 0.8964430093765259,
      "num_tokens": 414341723.0,
      "step": 11909
    },
    {
      "epoch": 2.211699164345404,
      "grad_norm": 1.5505107740342128,
      "learning_rate": 1e-06,
      "loss": 0.304,
      "mean_token_accuracy": 0.891203761100769,
      "num_tokens": 414376449.0,
      "step": 11910
    },
    {
      "epoch": 2.2118848653667595,
      "grad_norm": 1.770263440404212,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8798643350601196,
      "num_tokens": 414408818.0,
      "step": 11911
    },
    {
      "epoch": 2.2120705663881153,
      "grad_norm": 1.5326080498523558,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8806612491607666,
      "num_tokens": 414446996.0,
      "step": 11912
    },
    {
      "epoch": 2.2122562674094706,
      "grad_norm": 1.6556994551171604,
      "learning_rate": 1e-06,
      "loss": 0.3492,
      "mean_token_accuracy": 0.8764419555664062,
      "num_tokens": 414481922.0,
      "step": 11913
    },
    {
      "epoch": 2.2124419684308263,
      "grad_norm": 1.5622663743066945,
      "learning_rate": 1e-06,
      "loss": 0.305,
      "mean_token_accuracy": 0.8880226016044617,
      "num_tokens": 414514073.0,
      "step": 11914
    },
    {
      "epoch": 2.212627669452182,
      "grad_norm": 1.5675721748757545,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8741346597671509,
      "num_tokens": 414550293.0,
      "step": 11915
    },
    {
      "epoch": 2.2128133704735378,
      "grad_norm": 1.562799664085487,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8807065486907959,
      "num_tokens": 414586205.0,
      "step": 11916
    },
    {
      "epoch": 2.212999071494893,
      "grad_norm": 1.4505897268973307,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8940004110336304,
      "num_tokens": 414626575.0,
      "step": 11917
    },
    {
      "epoch": 2.213184772516249,
      "grad_norm": 1.5122534191406625,
      "learning_rate": 1e-06,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8923389911651611,
      "num_tokens": 414664871.0,
      "step": 11918
    },
    {
      "epoch": 2.2133704735376045,
      "grad_norm": 1.5213184634791657,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.8960837721824646,
      "num_tokens": 414701133.0,
      "step": 11919
    },
    {
      "epoch": 2.2135561745589603,
      "grad_norm": 1.727747162641998,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8850337266921997,
      "num_tokens": 414728122.0,
      "step": 11920
    },
    {
      "epoch": 2.2137418755803155,
      "grad_norm": 1.6462876705730323,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8830370903015137,
      "num_tokens": 414760087.0,
      "step": 11921
    },
    {
      "epoch": 2.2139275766016713,
      "grad_norm": 1.4819982281955655,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8926786780357361,
      "num_tokens": 414798045.0,
      "step": 11922
    },
    {
      "epoch": 2.214113277623027,
      "grad_norm": 1.794457570727256,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8832769393920898,
      "num_tokens": 414828811.0,
      "step": 11923
    },
    {
      "epoch": 2.2142989786443827,
      "grad_norm": 1.5355247976665636,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8866496086120605,
      "num_tokens": 414866170.0,
      "step": 11924
    },
    {
      "epoch": 2.214484679665738,
      "grad_norm": 1.528787396193643,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8931220769882202,
      "num_tokens": 414900657.0,
      "step": 11925
    },
    {
      "epoch": 2.2146703806870938,
      "grad_norm": 1.4786276277780235,
      "learning_rate": 1e-06,
      "loss": 0.3035,
      "mean_token_accuracy": 0.8936992883682251,
      "num_tokens": 414937293.0,
      "step": 11926
    },
    {
      "epoch": 2.2148560817084495,
      "grad_norm": 1.51028787035727,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8936903476715088,
      "num_tokens": 414973711.0,
      "step": 11927
    },
    {
      "epoch": 2.215041782729805,
      "grad_norm": 1.5202083702477263,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.8952391743659973,
      "num_tokens": 415008952.0,
      "step": 11928
    },
    {
      "epoch": 2.2152274837511605,
      "grad_norm": 1.617422656741211,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8832569122314453,
      "num_tokens": 415042678.0,
      "step": 11929
    },
    {
      "epoch": 2.2154131847725163,
      "grad_norm": 1.7469489116359271,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8811469078063965,
      "num_tokens": 415076923.0,
      "step": 11930
    },
    {
      "epoch": 2.215598885793872,
      "grad_norm": 1.5505383318691381,
      "learning_rate": 1e-06,
      "loss": 0.3005,
      "mean_token_accuracy": 0.8951881527900696,
      "num_tokens": 415110079.0,
      "step": 11931
    },
    {
      "epoch": 2.2157845868152277,
      "grad_norm": 1.6239309863405544,
      "learning_rate": 1e-06,
      "loss": 0.3026,
      "mean_token_accuracy": 0.8929755687713623,
      "num_tokens": 415143579.0,
      "step": 11932
    },
    {
      "epoch": 2.215970287836583,
      "grad_norm": 1.5061927399915211,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8856949806213379,
      "num_tokens": 415180964.0,
      "step": 11933
    },
    {
      "epoch": 2.2161559888579387,
      "grad_norm": 1.5371893374736831,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8783416748046875,
      "num_tokens": 415218117.0,
      "step": 11934
    },
    {
      "epoch": 2.2163416898792945,
      "grad_norm": 1.568799095025923,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8824652433395386,
      "num_tokens": 415255044.0,
      "step": 11935
    },
    {
      "epoch": 2.2165273909006498,
      "grad_norm": 1.7103708851566748,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8882527351379395,
      "num_tokens": 415285346.0,
      "step": 11936
    },
    {
      "epoch": 2.2167130919220055,
      "grad_norm": 1.754151788969155,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8953951597213745,
      "num_tokens": 415312709.0,
      "step": 11937
    },
    {
      "epoch": 2.2168987929433612,
      "grad_norm": 1.660943601178831,
      "learning_rate": 1e-06,
      "loss": 0.3042,
      "mean_token_accuracy": 0.8878693580627441,
      "num_tokens": 415342280.0,
      "step": 11938
    },
    {
      "epoch": 2.217084493964717,
      "grad_norm": 1.514480189974936,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8951798677444458,
      "num_tokens": 415376919.0,
      "step": 11939
    },
    {
      "epoch": 2.2172701949860723,
      "grad_norm": 1.563117991539672,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8870441913604736,
      "num_tokens": 415413005.0,
      "step": 11940
    },
    {
      "epoch": 2.217455896007428,
      "grad_norm": 1.5309061447774328,
      "learning_rate": 1e-06,
      "loss": 0.2751,
      "mean_token_accuracy": 0.9030542969703674,
      "num_tokens": 415445257.0,
      "step": 11941
    },
    {
      "epoch": 2.2176415970287837,
      "grad_norm": 1.6410651075992675,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8882160782814026,
      "num_tokens": 415479001.0,
      "step": 11942
    },
    {
      "epoch": 2.2178272980501395,
      "grad_norm": 1.5374375499981394,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8883563876152039,
      "num_tokens": 415515063.0,
      "step": 11943
    },
    {
      "epoch": 2.2180129990714947,
      "grad_norm": 1.6672977406727587,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8779184818267822,
      "num_tokens": 415549265.0,
      "step": 11944
    },
    {
      "epoch": 2.2181987000928505,
      "grad_norm": 1.4860368943984086,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8790361285209656,
      "num_tokens": 415587000.0,
      "step": 11945
    },
    {
      "epoch": 2.218384401114206,
      "grad_norm": 1.5216289411641486,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8848301768302917,
      "num_tokens": 415624832.0,
      "step": 11946
    },
    {
      "epoch": 2.218570102135562,
      "grad_norm": 1.6746783781218006,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.878710150718689,
      "num_tokens": 415657097.0,
      "step": 11947
    },
    {
      "epoch": 2.2187558031569172,
      "grad_norm": 1.6936885675566336,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8868615627288818,
      "num_tokens": 415692531.0,
      "step": 11948
    },
    {
      "epoch": 2.218941504178273,
      "grad_norm": 1.6447728306399734,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8774398565292358,
      "num_tokens": 415724260.0,
      "step": 11949
    },
    {
      "epoch": 2.2191272051996287,
      "grad_norm": 1.5224032229837632,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8903416395187378,
      "num_tokens": 415760658.0,
      "step": 11950
    },
    {
      "epoch": 2.219312906220984,
      "grad_norm": 1.6263222719176391,
      "learning_rate": 1e-06,
      "loss": 0.2928,
      "mean_token_accuracy": 0.895902693271637,
      "num_tokens": 415791202.0,
      "step": 11951
    },
    {
      "epoch": 2.2194986072423397,
      "grad_norm": 1.4334662253868498,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8869156241416931,
      "num_tokens": 415830623.0,
      "step": 11952
    },
    {
      "epoch": 2.2196843082636954,
      "grad_norm": 1.630676476536954,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8868200778961182,
      "num_tokens": 415861157.0,
      "step": 11953
    },
    {
      "epoch": 2.219870009285051,
      "grad_norm": 1.388646128690849,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8902890682220459,
      "num_tokens": 415906176.0,
      "step": 11954
    },
    {
      "epoch": 2.220055710306407,
      "grad_norm": 1.6251371104914027,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8871539831161499,
      "num_tokens": 415938947.0,
      "step": 11955
    },
    {
      "epoch": 2.220241411327762,
      "grad_norm": 1.5917208112102155,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8882350921630859,
      "num_tokens": 415973576.0,
      "step": 11956
    },
    {
      "epoch": 2.220427112349118,
      "grad_norm": 1.5137870935005544,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8873751163482666,
      "num_tokens": 416015467.0,
      "step": 11957
    },
    {
      "epoch": 2.2206128133704737,
      "grad_norm": 1.4857729770204673,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8835767507553101,
      "num_tokens": 416052396.0,
      "step": 11958
    },
    {
      "epoch": 2.220798514391829,
      "grad_norm": 1.377489062515008,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.895780622959137,
      "num_tokens": 416094865.0,
      "step": 11959
    },
    {
      "epoch": 2.2209842154131847,
      "grad_norm": 1.5983033536739495,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8859111070632935,
      "num_tokens": 416128688.0,
      "step": 11960
    },
    {
      "epoch": 2.2211699164345404,
      "grad_norm": 1.5428315622522164,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.881308913230896,
      "num_tokens": 416162715.0,
      "step": 11961
    },
    {
      "epoch": 2.221355617455896,
      "grad_norm": 1.575663853443187,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8877219557762146,
      "num_tokens": 416196783.0,
      "step": 11962
    },
    {
      "epoch": 2.2215413184772514,
      "grad_norm": 1.5775914080256763,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8760477304458618,
      "num_tokens": 416234524.0,
      "step": 11963
    },
    {
      "epoch": 2.221727019498607,
      "grad_norm": 1.6337855728482802,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8891757726669312,
      "num_tokens": 416268275.0,
      "step": 11964
    },
    {
      "epoch": 2.221912720519963,
      "grad_norm": 1.5902706434372846,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8908141851425171,
      "num_tokens": 416300249.0,
      "step": 11965
    },
    {
      "epoch": 2.2220984215413186,
      "grad_norm": 1.4295510203073176,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8883563280105591,
      "num_tokens": 416342191.0,
      "step": 11966
    },
    {
      "epoch": 2.222284122562674,
      "grad_norm": 1.5494344154103812,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.8966010808944702,
      "num_tokens": 416376755.0,
      "step": 11967
    },
    {
      "epoch": 2.2224698235840297,
      "grad_norm": 1.5912887515231051,
      "learning_rate": 1e-06,
      "loss": 0.2852,
      "mean_token_accuracy": 0.8974289894104004,
      "num_tokens": 416409208.0,
      "step": 11968
    },
    {
      "epoch": 2.2226555246053854,
      "grad_norm": 1.655056747986465,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8619195222854614,
      "num_tokens": 416449056.0,
      "step": 11969
    },
    {
      "epoch": 2.222841225626741,
      "grad_norm": 1.728803889243137,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8848680257797241,
      "num_tokens": 416480532.0,
      "step": 11970
    },
    {
      "epoch": 2.2230269266480964,
      "grad_norm": 1.5372470924806854,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8880728483200073,
      "num_tokens": 416515706.0,
      "step": 11971
    },
    {
      "epoch": 2.223212627669452,
      "grad_norm": 1.4656413368133028,
      "learning_rate": 1e-06,
      "loss": 0.276,
      "mean_token_accuracy": 0.9026921987533569,
      "num_tokens": 416551047.0,
      "step": 11972
    },
    {
      "epoch": 2.223398328690808,
      "grad_norm": 1.8172859288528067,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.87932950258255,
      "num_tokens": 416574377.0,
      "step": 11973
    },
    {
      "epoch": 2.2235840297121636,
      "grad_norm": 1.5516817096307383,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8782131671905518,
      "num_tokens": 416610696.0,
      "step": 11974
    },
    {
      "epoch": 2.223769730733519,
      "grad_norm": 1.6265938286144648,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8762913942337036,
      "num_tokens": 416645395.0,
      "step": 11975
    },
    {
      "epoch": 2.2239554317548746,
      "grad_norm": 1.568834302535625,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.880059003829956,
      "num_tokens": 416685150.0,
      "step": 11976
    },
    {
      "epoch": 2.2241411327762304,
      "grad_norm": 1.6712472625002992,
      "learning_rate": 1e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.8902806043624878,
      "num_tokens": 416713651.0,
      "step": 11977
    },
    {
      "epoch": 2.224326833797586,
      "grad_norm": 1.66331581745735,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8762401342391968,
      "num_tokens": 416749742.0,
      "step": 11978
    },
    {
      "epoch": 2.2245125348189414,
      "grad_norm": 1.6745010485835634,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.8957996964454651,
      "num_tokens": 416782408.0,
      "step": 11979
    },
    {
      "epoch": 2.224698235840297,
      "grad_norm": 1.590049260397542,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8822249174118042,
      "num_tokens": 416819626.0,
      "step": 11980
    },
    {
      "epoch": 2.224883936861653,
      "grad_norm": 1.728635626978717,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8868564963340759,
      "num_tokens": 416848125.0,
      "step": 11981
    },
    {
      "epoch": 2.225069637883008,
      "grad_norm": 1.6002788774745036,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8820974230766296,
      "num_tokens": 416885638.0,
      "step": 11982
    },
    {
      "epoch": 2.225255338904364,
      "grad_norm": 1.5775799863664899,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8751616477966309,
      "num_tokens": 416923711.0,
      "step": 11983
    },
    {
      "epoch": 2.2254410399257196,
      "grad_norm": 1.561299771850884,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8944928646087646,
      "num_tokens": 416956458.0,
      "step": 11984
    },
    {
      "epoch": 2.2256267409470754,
      "grad_norm": 1.7019373315242665,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.9007464647293091,
      "num_tokens": 416984878.0,
      "step": 11985
    },
    {
      "epoch": 2.2258124419684306,
      "grad_norm": 1.5683774417110903,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8768654465675354,
      "num_tokens": 417019567.0,
      "step": 11986
    },
    {
      "epoch": 2.2259981429897864,
      "grad_norm": 1.6214338383045914,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8748340606689453,
      "num_tokens": 417053507.0,
      "step": 11987
    },
    {
      "epoch": 2.226183844011142,
      "grad_norm": 1.5929736056263362,
      "learning_rate": 1e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8666499853134155,
      "num_tokens": 417089523.0,
      "step": 11988
    },
    {
      "epoch": 2.226369545032498,
      "grad_norm": 1.6098350941697797,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8907356262207031,
      "num_tokens": 417121670.0,
      "step": 11989
    },
    {
      "epoch": 2.226555246053853,
      "grad_norm": 1.4828780919853108,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8873405456542969,
      "num_tokens": 417161351.0,
      "step": 11990
    },
    {
      "epoch": 2.226740947075209,
      "grad_norm": 1.742336415421102,
      "learning_rate": 1e-06,
      "loss": 0.2983,
      "mean_token_accuracy": 0.892271876335144,
      "num_tokens": 417189066.0,
      "step": 11991
    },
    {
      "epoch": 2.2269266480965646,
      "grad_norm": 1.6793892275660112,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8906643390655518,
      "num_tokens": 417222319.0,
      "step": 11992
    },
    {
      "epoch": 2.2271123491179203,
      "grad_norm": 1.5248968111738577,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8899234533309937,
      "num_tokens": 417259140.0,
      "step": 11993
    },
    {
      "epoch": 2.2272980501392756,
      "grad_norm": 1.60225806958206,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.8961672782897949,
      "num_tokens": 417291962.0,
      "step": 11994
    },
    {
      "epoch": 2.2274837511606314,
      "grad_norm": 1.6465326680076653,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8682436943054199,
      "num_tokens": 417329574.0,
      "step": 11995
    },
    {
      "epoch": 2.227669452181987,
      "grad_norm": 1.6800829400728663,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8960860967636108,
      "num_tokens": 417356725.0,
      "step": 11996
    },
    {
      "epoch": 2.227855153203343,
      "grad_norm": 1.7863095820786472,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.888496994972229,
      "num_tokens": 417390207.0,
      "step": 11997
    },
    {
      "epoch": 2.228040854224698,
      "grad_norm": 1.4673561025311133,
      "learning_rate": 1e-06,
      "loss": 0.2659,
      "mean_token_accuracy": 0.9046995639801025,
      "num_tokens": 417429354.0,
      "step": 11998
    },
    {
      "epoch": 2.228226555246054,
      "grad_norm": 1.7034371952396683,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8741099834442139,
      "num_tokens": 417460803.0,
      "step": 11999
    },
    {
      "epoch": 2.2284122562674096,
      "grad_norm": 1.565017157357703,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8813371658325195,
      "num_tokens": 417493477.0,
      "step": 12000
    },
    {
      "epoch": 2.2285979572887653,
      "grad_norm": 1.7266543976775857,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.886702299118042,
      "num_tokens": 417526382.0,
      "step": 12001
    },
    {
      "epoch": 2.2287836583101206,
      "grad_norm": 1.591654053862632,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8826293349266052,
      "num_tokens": 417560418.0,
      "step": 12002
    },
    {
      "epoch": 2.2289693593314763,
      "grad_norm": 1.5817206071519052,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8797655701637268,
      "num_tokens": 417594657.0,
      "step": 12003
    },
    {
      "epoch": 2.229155060352832,
      "grad_norm": 1.4362190702467863,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8948326706886292,
      "num_tokens": 417630087.0,
      "step": 12004
    },
    {
      "epoch": 2.2293407613741874,
      "grad_norm": 1.5857275222781642,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.893432080745697,
      "num_tokens": 417662233.0,
      "step": 12005
    },
    {
      "epoch": 2.229526462395543,
      "grad_norm": 1.6092020100151956,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8905279040336609,
      "num_tokens": 417693688.0,
      "step": 12006
    },
    {
      "epoch": 2.229712163416899,
      "grad_norm": 1.633385848180236,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8813908100128174,
      "num_tokens": 417725745.0,
      "step": 12007
    },
    {
      "epoch": 2.2298978644382546,
      "grad_norm": 1.4641418722453845,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8960134983062744,
      "num_tokens": 417763584.0,
      "step": 12008
    },
    {
      "epoch": 2.23008356545961,
      "grad_norm": 1.598721121689884,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8851522207260132,
      "num_tokens": 417799172.0,
      "step": 12009
    },
    {
      "epoch": 2.2302692664809656,
      "grad_norm": 1.5611536739357723,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8844820261001587,
      "num_tokens": 417831556.0,
      "step": 12010
    },
    {
      "epoch": 2.2304549675023213,
      "grad_norm": 1.6219516229456765,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8814941644668579,
      "num_tokens": 417863997.0,
      "step": 12011
    },
    {
      "epoch": 2.230640668523677,
      "grad_norm": 1.595086804516476,
      "learning_rate": 1e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.8643044233322144,
      "num_tokens": 417902998.0,
      "step": 12012
    },
    {
      "epoch": 2.2308263695450323,
      "grad_norm": 1.6918210657267323,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8851233720779419,
      "num_tokens": 417933077.0,
      "step": 12013
    },
    {
      "epoch": 2.231012070566388,
      "grad_norm": 1.578573203929865,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.885875940322876,
      "num_tokens": 417966018.0,
      "step": 12014
    },
    {
      "epoch": 2.231197771587744,
      "grad_norm": 1.5230264010455683,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.881537139415741,
      "num_tokens": 418000557.0,
      "step": 12015
    },
    {
      "epoch": 2.2313834726090995,
      "grad_norm": 1.65146553215623,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8814195394515991,
      "num_tokens": 418033424.0,
      "step": 12016
    },
    {
      "epoch": 2.231569173630455,
      "grad_norm": 1.678827530953422,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.883328914642334,
      "num_tokens": 418066257.0,
      "step": 12017
    },
    {
      "epoch": 2.2317548746518105,
      "grad_norm": 1.475341319162367,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8805383443832397,
      "num_tokens": 418107027.0,
      "step": 12018
    },
    {
      "epoch": 2.2319405756731663,
      "grad_norm": 1.5447038034827019,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8895128965377808,
      "num_tokens": 418141870.0,
      "step": 12019
    },
    {
      "epoch": 2.232126276694522,
      "grad_norm": 1.500248313396841,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8694460988044739,
      "num_tokens": 418182131.0,
      "step": 12020
    },
    {
      "epoch": 2.2323119777158773,
      "grad_norm": 1.5449599660782023,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8791420459747314,
      "num_tokens": 418218960.0,
      "step": 12021
    },
    {
      "epoch": 2.232497678737233,
      "grad_norm": 1.574567589862106,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8755466938018799,
      "num_tokens": 418254577.0,
      "step": 12022
    },
    {
      "epoch": 2.2326833797585888,
      "grad_norm": 1.5597519607649775,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8904228210449219,
      "num_tokens": 418286094.0,
      "step": 12023
    },
    {
      "epoch": 2.2328690807799445,
      "grad_norm": 1.5297788881565226,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.88495934009552,
      "num_tokens": 418323605.0,
      "step": 12024
    },
    {
      "epoch": 2.2330547818013,
      "grad_norm": 1.6738878266984856,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8822540044784546,
      "num_tokens": 418358254.0,
      "step": 12025
    },
    {
      "epoch": 2.2332404828226555,
      "grad_norm": 1.5601345770991013,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.88414466381073,
      "num_tokens": 418393895.0,
      "step": 12026
    },
    {
      "epoch": 2.2334261838440113,
      "grad_norm": 1.6623299267504432,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8758233785629272,
      "num_tokens": 418427185.0,
      "step": 12027
    },
    {
      "epoch": 2.2336118848653665,
      "grad_norm": 1.5980990071224104,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8772828578948975,
      "num_tokens": 418462234.0,
      "step": 12028
    },
    {
      "epoch": 2.2337975858867223,
      "grad_norm": 1.4493158184507062,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8861857652664185,
      "num_tokens": 418501861.0,
      "step": 12029
    },
    {
      "epoch": 2.233983286908078,
      "grad_norm": 1.5847572409934465,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.871181845664978,
      "num_tokens": 418539393.0,
      "step": 12030
    },
    {
      "epoch": 2.2341689879294337,
      "grad_norm": 1.500042370902546,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8876103758811951,
      "num_tokens": 418578786.0,
      "step": 12031
    },
    {
      "epoch": 2.234354688950789,
      "grad_norm": 1.5921687566311722,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8765970468521118,
      "num_tokens": 418619288.0,
      "step": 12032
    },
    {
      "epoch": 2.2345403899721448,
      "grad_norm": 1.5565264545825668,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8859765529632568,
      "num_tokens": 418657107.0,
      "step": 12033
    },
    {
      "epoch": 2.2347260909935005,
      "grad_norm": 1.5646454142703183,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8781238794326782,
      "num_tokens": 418694104.0,
      "step": 12034
    },
    {
      "epoch": 2.2349117920148562,
      "grad_norm": 1.5591345266468841,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8887090682983398,
      "num_tokens": 418727583.0,
      "step": 12035
    },
    {
      "epoch": 2.2350974930362115,
      "grad_norm": 1.633032485081609,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8766184449195862,
      "num_tokens": 418762220.0,
      "step": 12036
    },
    {
      "epoch": 2.2352831940575673,
      "grad_norm": 1.506620504658586,
      "learning_rate": 1e-06,
      "loss": 0.2808,
      "mean_token_accuracy": 0.8985459804534912,
      "num_tokens": 418798902.0,
      "step": 12037
    },
    {
      "epoch": 2.235468895078923,
      "grad_norm": 1.8424197800271427,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8841862678527832,
      "num_tokens": 418825849.0,
      "step": 12038
    },
    {
      "epoch": 2.2356545961002787,
      "grad_norm": 1.4734804375641641,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8835141062736511,
      "num_tokens": 418863601.0,
      "step": 12039
    },
    {
      "epoch": 2.235840297121634,
      "grad_norm": 1.5491959781825104,
      "learning_rate": 1e-06,
      "loss": 0.2919,
      "mean_token_accuracy": 0.8962886333465576,
      "num_tokens": 418897459.0,
      "step": 12040
    },
    {
      "epoch": 2.2360259981429897,
      "grad_norm": 1.6679771505355558,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8825602531433105,
      "num_tokens": 418930327.0,
      "step": 12041
    },
    {
      "epoch": 2.2362116991643455,
      "grad_norm": 1.618286863560811,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.891985297203064,
      "num_tokens": 418961536.0,
      "step": 12042
    },
    {
      "epoch": 2.236397400185701,
      "grad_norm": 1.6368513497345338,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8826195001602173,
      "num_tokens": 418995031.0,
      "step": 12043
    },
    {
      "epoch": 2.2365831012070565,
      "grad_norm": 1.58509386958334,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8875753283500671,
      "num_tokens": 419029344.0,
      "step": 12044
    },
    {
      "epoch": 2.2367688022284122,
      "grad_norm": 1.5538388133491683,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8901535272598267,
      "num_tokens": 419063492.0,
      "step": 12045
    },
    {
      "epoch": 2.236954503249768,
      "grad_norm": 1.6278122027158846,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8881102800369263,
      "num_tokens": 419095337.0,
      "step": 12046
    },
    {
      "epoch": 2.2371402042711237,
      "grad_norm": 1.4908985974558722,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8890000581741333,
      "num_tokens": 419137409.0,
      "step": 12047
    },
    {
      "epoch": 2.237325905292479,
      "grad_norm": 1.5175886427659269,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8890628218650818,
      "num_tokens": 419176196.0,
      "step": 12048
    },
    {
      "epoch": 2.2375116063138347,
      "grad_norm": 1.8444943962264648,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8831989765167236,
      "num_tokens": 419202134.0,
      "step": 12049
    },
    {
      "epoch": 2.2376973073351905,
      "grad_norm": 1.5562873353443778,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8836064338684082,
      "num_tokens": 419241568.0,
      "step": 12050
    },
    {
      "epoch": 2.2378830083565457,
      "grad_norm": 1.5906497770962622,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8886402249336243,
      "num_tokens": 419273573.0,
      "step": 12051
    },
    {
      "epoch": 2.2380687093779015,
      "grad_norm": 1.7142273826787076,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8752045631408691,
      "num_tokens": 419304108.0,
      "step": 12052
    },
    {
      "epoch": 2.238254410399257,
      "grad_norm": 1.66707563397224,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8804685473442078,
      "num_tokens": 419335634.0,
      "step": 12053
    },
    {
      "epoch": 2.238440111420613,
      "grad_norm": 1.4567343361478209,
      "learning_rate": 1e-06,
      "loss": 0.3022,
      "mean_token_accuracy": 0.8926085829734802,
      "num_tokens": 419372643.0,
      "step": 12054
    },
    {
      "epoch": 2.2386258124419682,
      "grad_norm": 1.5156348737322787,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8958165049552917,
      "num_tokens": 419406104.0,
      "step": 12055
    },
    {
      "epoch": 2.238811513463324,
      "grad_norm": 1.4893059248066298,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8837177753448486,
      "num_tokens": 419443459.0,
      "step": 12056
    },
    {
      "epoch": 2.2389972144846797,
      "grad_norm": 1.457165647611018,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8895412087440491,
      "num_tokens": 419482261.0,
      "step": 12057
    },
    {
      "epoch": 2.2391829155060354,
      "grad_norm": 1.625337279929574,
      "learning_rate": 1e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8676302433013916,
      "num_tokens": 419520643.0,
      "step": 12058
    },
    {
      "epoch": 2.2393686165273907,
      "grad_norm": 1.506673179355715,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.874066174030304,
      "num_tokens": 419560753.0,
      "step": 12059
    },
    {
      "epoch": 2.2395543175487465,
      "grad_norm": 1.5058923185495223,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8889719843864441,
      "num_tokens": 419600188.0,
      "step": 12060
    },
    {
      "epoch": 2.239740018570102,
      "grad_norm": 1.4342973326548354,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8983176946640015,
      "num_tokens": 419635685.0,
      "step": 12061
    },
    {
      "epoch": 2.239925719591458,
      "grad_norm": 1.4587367611456798,
      "learning_rate": 1e-06,
      "loss": 0.299,
      "mean_token_accuracy": 0.8924508690834045,
      "num_tokens": 419672308.0,
      "step": 12062
    },
    {
      "epoch": 2.240111420612813,
      "grad_norm": 1.564806565637324,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8858349919319153,
      "num_tokens": 419705879.0,
      "step": 12063
    },
    {
      "epoch": 2.240297121634169,
      "grad_norm": 1.4491454324177193,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8878400325775146,
      "num_tokens": 419748555.0,
      "step": 12064
    },
    {
      "epoch": 2.2404828226555247,
      "grad_norm": 1.6829555598233745,
      "learning_rate": 1e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8688342571258545,
      "num_tokens": 419784447.0,
      "step": 12065
    },
    {
      "epoch": 2.2406685236768804,
      "grad_norm": 1.416726985711445,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8948162794113159,
      "num_tokens": 419821200.0,
      "step": 12066
    },
    {
      "epoch": 2.2408542246982357,
      "grad_norm": 1.5294265168817587,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8801052570343018,
      "num_tokens": 419856471.0,
      "step": 12067
    },
    {
      "epoch": 2.2410399257195914,
      "grad_norm": 1.547385212740835,
      "learning_rate": 1e-06,
      "loss": 0.364,
      "mean_token_accuracy": 0.8691011667251587,
      "num_tokens": 419892890.0,
      "step": 12068
    },
    {
      "epoch": 2.241225626740947,
      "grad_norm": 1.4491303609718382,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8883893489837646,
      "num_tokens": 419932807.0,
      "step": 12069
    },
    {
      "epoch": 2.241411327762303,
      "grad_norm": 1.716785409876207,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8894494771957397,
      "num_tokens": 419960758.0,
      "step": 12070
    },
    {
      "epoch": 2.241597028783658,
      "grad_norm": 1.613995002190154,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8917591571807861,
      "num_tokens": 419994929.0,
      "step": 12071
    },
    {
      "epoch": 2.241782729805014,
      "grad_norm": 1.658755436396781,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8950158953666687,
      "num_tokens": 420022238.0,
      "step": 12072
    },
    {
      "epoch": 2.2419684308263697,
      "grad_norm": 1.4327028883721609,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8876317739486694,
      "num_tokens": 420063967.0,
      "step": 12073
    },
    {
      "epoch": 2.242154131847725,
      "grad_norm": 1.9291726946539811,
      "learning_rate": 1e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8617486953735352,
      "num_tokens": 420090796.0,
      "step": 12074
    },
    {
      "epoch": 2.2423398328690807,
      "grad_norm": 1.5372152018458345,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.888237476348877,
      "num_tokens": 420125614.0,
      "step": 12075
    },
    {
      "epoch": 2.2425255338904364,
      "grad_norm": 1.538134245517266,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8807247281074524,
      "num_tokens": 420165127.0,
      "step": 12076
    },
    {
      "epoch": 2.242711234911792,
      "grad_norm": 1.5773248330726763,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8758422136306763,
      "num_tokens": 420200963.0,
      "step": 12077
    },
    {
      "epoch": 2.2428969359331474,
      "grad_norm": 1.5672498304381213,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8778949975967407,
      "num_tokens": 420239506.0,
      "step": 12078
    },
    {
      "epoch": 2.243082636954503,
      "grad_norm": 1.6432083910779445,
      "learning_rate": 1e-06,
      "loss": 0.3473,
      "mean_token_accuracy": 0.88182133436203,
      "num_tokens": 420272783.0,
      "step": 12079
    },
    {
      "epoch": 2.243268337975859,
      "grad_norm": 1.5485112511428054,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.882250189781189,
      "num_tokens": 420307026.0,
      "step": 12080
    },
    {
      "epoch": 2.2434540389972146,
      "grad_norm": 1.615756343403121,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8854164481163025,
      "num_tokens": 420339912.0,
      "step": 12081
    },
    {
      "epoch": 2.24363974001857,
      "grad_norm": 1.5846858047918024,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.880750298500061,
      "num_tokens": 420380683.0,
      "step": 12082
    },
    {
      "epoch": 2.2438254410399256,
      "grad_norm": 1.5276771511732608,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8859874606132507,
      "num_tokens": 420420176.0,
      "step": 12083
    },
    {
      "epoch": 2.2440111420612814,
      "grad_norm": 1.4378880733274555,
      "learning_rate": 1e-06,
      "loss": 0.2904,
      "mean_token_accuracy": 0.8993804454803467,
      "num_tokens": 420459206.0,
      "step": 12084
    },
    {
      "epoch": 2.244196843082637,
      "grad_norm": 1.6187851816825995,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8884207010269165,
      "num_tokens": 420492903.0,
      "step": 12085
    },
    {
      "epoch": 2.2443825441039924,
      "grad_norm": 1.7918904516607808,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8847907781600952,
      "num_tokens": 420520683.0,
      "step": 12086
    },
    {
      "epoch": 2.244568245125348,
      "grad_norm": 1.5689872066948678,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8855093717575073,
      "num_tokens": 420556357.0,
      "step": 12087
    },
    {
      "epoch": 2.244753946146704,
      "grad_norm": 1.5097213132007912,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8898370265960693,
      "num_tokens": 420592176.0,
      "step": 12088
    },
    {
      "epoch": 2.2449396471680596,
      "grad_norm": 1.538635974269434,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8891887664794922,
      "num_tokens": 420629979.0,
      "step": 12089
    },
    {
      "epoch": 2.245125348189415,
      "grad_norm": 1.6630405295514827,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8821941614151001,
      "num_tokens": 420663756.0,
      "step": 12090
    },
    {
      "epoch": 2.2453110492107706,
      "grad_norm": 1.603523242372688,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8918919563293457,
      "num_tokens": 420696573.0,
      "step": 12091
    },
    {
      "epoch": 2.2454967502321264,
      "grad_norm": 1.5357717521230174,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8734051585197449,
      "num_tokens": 420735230.0,
      "step": 12092
    },
    {
      "epoch": 2.245682451253482,
      "grad_norm": 1.6252096215634246,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8838533163070679,
      "num_tokens": 420774015.0,
      "step": 12093
    },
    {
      "epoch": 2.2458681522748374,
      "grad_norm": 1.562989280096306,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8959910869598389,
      "num_tokens": 420809124.0,
      "step": 12094
    },
    {
      "epoch": 2.246053853296193,
      "grad_norm": 1.360692794133078,
      "learning_rate": 1e-06,
      "loss": 0.269,
      "mean_token_accuracy": 0.9060333967208862,
      "num_tokens": 420849163.0,
      "step": 12095
    },
    {
      "epoch": 2.246239554317549,
      "grad_norm": 1.4601013928116382,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8911455869674683,
      "num_tokens": 420891111.0,
      "step": 12096
    },
    {
      "epoch": 2.246425255338904,
      "grad_norm": 1.5587863197746565,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8891927599906921,
      "num_tokens": 420924473.0,
      "step": 12097
    },
    {
      "epoch": 2.24661095636026,
      "grad_norm": 1.6090604501435366,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8756798505783081,
      "num_tokens": 420960838.0,
      "step": 12098
    },
    {
      "epoch": 2.2467966573816156,
      "grad_norm": 1.6054679922669346,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8858960866928101,
      "num_tokens": 420992645.0,
      "step": 12099
    },
    {
      "epoch": 2.2469823584029713,
      "grad_norm": 1.5398571049476912,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8860725164413452,
      "num_tokens": 421029969.0,
      "step": 12100
    },
    {
      "epoch": 2.247168059424327,
      "grad_norm": 1.5793536551350493,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8857691287994385,
      "num_tokens": 421066155.0,
      "step": 12101
    },
    {
      "epoch": 2.2473537604456824,
      "grad_norm": 1.5324573797570769,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8776950240135193,
      "num_tokens": 421102472.0,
      "step": 12102
    },
    {
      "epoch": 2.247539461467038,
      "grad_norm": 1.4089774743623964,
      "learning_rate": 1e-06,
      "loss": 0.2862,
      "mean_token_accuracy": 0.8983808159828186,
      "num_tokens": 421140023.0,
      "step": 12103
    },
    {
      "epoch": 2.247725162488394,
      "grad_norm": 1.7673128695275782,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.896248459815979,
      "num_tokens": 421165874.0,
      "step": 12104
    },
    {
      "epoch": 2.247910863509749,
      "grad_norm": 1.5501892899141148,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8887176513671875,
      "num_tokens": 421200266.0,
      "step": 12105
    },
    {
      "epoch": 2.248096564531105,
      "grad_norm": 1.6170574227889927,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8954565525054932,
      "num_tokens": 421231309.0,
      "step": 12106
    },
    {
      "epoch": 2.2482822655524606,
      "grad_norm": 1.653625679896335,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.887261688709259,
      "num_tokens": 421261854.0,
      "step": 12107
    },
    {
      "epoch": 2.2484679665738163,
      "grad_norm": 1.5422647140701533,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8844864368438721,
      "num_tokens": 421296724.0,
      "step": 12108
    },
    {
      "epoch": 2.2486536675951716,
      "grad_norm": 1.5923937172398162,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8810942769050598,
      "num_tokens": 421328915.0,
      "step": 12109
    },
    {
      "epoch": 2.2488393686165273,
      "grad_norm": 1.5426156460187859,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8792585134506226,
      "num_tokens": 421364141.0,
      "step": 12110
    },
    {
      "epoch": 2.249025069637883,
      "grad_norm": 1.4744566273802144,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8887960910797119,
      "num_tokens": 421405788.0,
      "step": 12111
    },
    {
      "epoch": 2.249210770659239,
      "grad_norm": 1.5424277307157244,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8859120011329651,
      "num_tokens": 421442656.0,
      "step": 12112
    },
    {
      "epoch": 2.249396471680594,
      "grad_norm": 1.4622734837886495,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8919856548309326,
      "num_tokens": 421479576.0,
      "step": 12113
    },
    {
      "epoch": 2.24958217270195,
      "grad_norm": 1.6926882929746894,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8870217800140381,
      "num_tokens": 421510404.0,
      "step": 12114
    },
    {
      "epoch": 2.2497678737233056,
      "grad_norm": 1.4733937821787382,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8884158730506897,
      "num_tokens": 421548081.0,
      "step": 12115
    },
    {
      "epoch": 2.2499535747446613,
      "grad_norm": 1.6577075301509874,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8836071491241455,
      "num_tokens": 421579332.0,
      "step": 12116
    },
    {
      "epoch": 2.2501392757660166,
      "grad_norm": 1.6731245651251454,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8837648630142212,
      "num_tokens": 421610974.0,
      "step": 12117
    },
    {
      "epoch": 2.2503249767873723,
      "grad_norm": 1.6112343436394787,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8830689191818237,
      "num_tokens": 421646884.0,
      "step": 12118
    },
    {
      "epoch": 2.250510677808728,
      "grad_norm": 1.5586283903073308,
      "learning_rate": 1e-06,
      "loss": 0.3514,
      "mean_token_accuracy": 0.8822019100189209,
      "num_tokens": 421684169.0,
      "step": 12119
    },
    {
      "epoch": 2.2506963788300833,
      "grad_norm": 1.5067202808548308,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8918981552124023,
      "num_tokens": 421719562.0,
      "step": 12120
    },
    {
      "epoch": 2.250882079851439,
      "grad_norm": 1.694015532727233,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8777428865432739,
      "num_tokens": 421755386.0,
      "step": 12121
    },
    {
      "epoch": 2.251067780872795,
      "grad_norm": 1.6828901350281664,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8899566531181335,
      "num_tokens": 421784051.0,
      "step": 12122
    },
    {
      "epoch": 2.2512534818941505,
      "grad_norm": 1.490927579099924,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8961492776870728,
      "num_tokens": 421817266.0,
      "step": 12123
    },
    {
      "epoch": 2.2514391829155063,
      "grad_norm": 1.699307832908907,
      "learning_rate": 1e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8697665929794312,
      "num_tokens": 421849823.0,
      "step": 12124
    },
    {
      "epoch": 2.2516248839368616,
      "grad_norm": 1.6414962807381306,
      "learning_rate": 1e-06,
      "loss": 0.2748,
      "mean_token_accuracy": 0.9014377593994141,
      "num_tokens": 421882415.0,
      "step": 12125
    },
    {
      "epoch": 2.2518105849582173,
      "grad_norm": 1.4707052497132511,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8873856067657471,
      "num_tokens": 421920354.0,
      "step": 12126
    },
    {
      "epoch": 2.251996285979573,
      "grad_norm": 1.6013808625342216,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8910728096961975,
      "num_tokens": 421950855.0,
      "step": 12127
    },
    {
      "epoch": 2.2521819870009283,
      "grad_norm": 1.7890399436166744,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8746744394302368,
      "num_tokens": 421979490.0,
      "step": 12128
    },
    {
      "epoch": 2.252367688022284,
      "grad_norm": 1.5255136363745359,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8949058055877686,
      "num_tokens": 422014654.0,
      "step": 12129
    },
    {
      "epoch": 2.2525533890436398,
      "grad_norm": 1.467518316262031,
      "learning_rate": 1e-06,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9056564569473267,
      "num_tokens": 422047049.0,
      "step": 12130
    },
    {
      "epoch": 2.2527390900649955,
      "grad_norm": 1.5784515804986845,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8738456964492798,
      "num_tokens": 422085511.0,
      "step": 12131
    },
    {
      "epoch": 2.252924791086351,
      "grad_norm": 1.7020888055882757,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8850656151771545,
      "num_tokens": 422118724.0,
      "step": 12132
    },
    {
      "epoch": 2.2531104921077065,
      "grad_norm": 1.8687139550717853,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8877687454223633,
      "num_tokens": 422148553.0,
      "step": 12133
    },
    {
      "epoch": 2.2532961931290623,
      "grad_norm": 1.7238488905716822,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8964776396751404,
      "num_tokens": 422176211.0,
      "step": 12134
    },
    {
      "epoch": 2.253481894150418,
      "grad_norm": 1.492555171633288,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.8962631225585938,
      "num_tokens": 422211334.0,
      "step": 12135
    },
    {
      "epoch": 2.2536675951717733,
      "grad_norm": 1.4912986135861224,
      "learning_rate": 1e-06,
      "loss": 0.2813,
      "mean_token_accuracy": 0.9011565446853638,
      "num_tokens": 422248972.0,
      "step": 12136
    },
    {
      "epoch": 2.253853296193129,
      "grad_norm": 1.4152018562483304,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8831885457038879,
      "num_tokens": 422290744.0,
      "step": 12137
    },
    {
      "epoch": 2.2540389972144848,
      "grad_norm": 1.6615518790716315,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8783625364303589,
      "num_tokens": 422321678.0,
      "step": 12138
    },
    {
      "epoch": 2.2542246982358405,
      "grad_norm": 1.5082340688907236,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8894332051277161,
      "num_tokens": 422358509.0,
      "step": 12139
    },
    {
      "epoch": 2.2544103992571958,
      "grad_norm": 1.7114871417398465,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8887704610824585,
      "num_tokens": 422387980.0,
      "step": 12140
    },
    {
      "epoch": 2.2545961002785515,
      "grad_norm": 1.6589610663257868,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.889176070690155,
      "num_tokens": 422424059.0,
      "step": 12141
    },
    {
      "epoch": 2.2547818012999072,
      "grad_norm": 1.7284168462692526,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8900041580200195,
      "num_tokens": 422455583.0,
      "step": 12142
    },
    {
      "epoch": 2.2549675023212625,
      "grad_norm": 1.7389700093906735,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8820784091949463,
      "num_tokens": 422486213.0,
      "step": 12143
    },
    {
      "epoch": 2.2551532033426183,
      "grad_norm": 1.5725512882685189,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8901789784431458,
      "num_tokens": 422518301.0,
      "step": 12144
    },
    {
      "epoch": 2.255338904363974,
      "grad_norm": 1.5120531714122825,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8868505954742432,
      "num_tokens": 422555596.0,
      "step": 12145
    },
    {
      "epoch": 2.2555246053853297,
      "grad_norm": 1.528322176370725,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8870684504508972,
      "num_tokens": 422592991.0,
      "step": 12146
    },
    {
      "epoch": 2.2557103064066855,
      "grad_norm": 1.6670523991809787,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8851080536842346,
      "num_tokens": 422626296.0,
      "step": 12147
    },
    {
      "epoch": 2.2558960074280408,
      "grad_norm": 1.6209824044857548,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8881154656410217,
      "num_tokens": 422655482.0,
      "step": 12148
    },
    {
      "epoch": 2.2560817084493965,
      "grad_norm": 1.5775169773069595,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8926855325698853,
      "num_tokens": 422690004.0,
      "step": 12149
    },
    {
      "epoch": 2.256267409470752,
      "grad_norm": 1.7685274700162548,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8807634115219116,
      "num_tokens": 422717929.0,
      "step": 12150
    },
    {
      "epoch": 2.2564531104921075,
      "grad_norm": 1.5698320486454655,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8791494369506836,
      "num_tokens": 422753887.0,
      "step": 12151
    },
    {
      "epoch": 2.2566388115134632,
      "grad_norm": 1.7266888964522613,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.8946782350540161,
      "num_tokens": 422784742.0,
      "step": 12152
    },
    {
      "epoch": 2.256824512534819,
      "grad_norm": 1.610913660588215,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8865649700164795,
      "num_tokens": 422818040.0,
      "step": 12153
    },
    {
      "epoch": 2.2570102135561747,
      "grad_norm": 1.5248338302152062,
      "learning_rate": 1e-06,
      "loss": 0.2901,
      "mean_token_accuracy": 0.8941518068313599,
      "num_tokens": 422850181.0,
      "step": 12154
    },
    {
      "epoch": 2.25719591457753,
      "grad_norm": 1.401706210924083,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8907837271690369,
      "num_tokens": 422889099.0,
      "step": 12155
    },
    {
      "epoch": 2.2573816155988857,
      "grad_norm": 1.59458120373226,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8882012367248535,
      "num_tokens": 422924849.0,
      "step": 12156
    },
    {
      "epoch": 2.2575673166202415,
      "grad_norm": 1.5947806803961293,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.894806444644928,
      "num_tokens": 422960025.0,
      "step": 12157
    },
    {
      "epoch": 2.257753017641597,
      "grad_norm": 1.442601968541303,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8887327313423157,
      "num_tokens": 423001221.0,
      "step": 12158
    },
    {
      "epoch": 2.2579387186629525,
      "grad_norm": 1.614671874505772,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8859652280807495,
      "num_tokens": 423035353.0,
      "step": 12159
    },
    {
      "epoch": 2.258124419684308,
      "grad_norm": 1.5489749378030386,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8867059350013733,
      "num_tokens": 423070185.0,
      "step": 12160
    },
    {
      "epoch": 2.258310120705664,
      "grad_norm": 1.5705167100606743,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8855711221694946,
      "num_tokens": 423105267.0,
      "step": 12161
    },
    {
      "epoch": 2.2584958217270197,
      "grad_norm": 1.6745722967554302,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8916585445404053,
      "num_tokens": 423139235.0,
      "step": 12162
    },
    {
      "epoch": 2.258681522748375,
      "grad_norm": 1.7440353634279147,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8900219202041626,
      "num_tokens": 423167365.0,
      "step": 12163
    },
    {
      "epoch": 2.2588672237697307,
      "grad_norm": 1.7811301454847415,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8914064168930054,
      "num_tokens": 423195407.0,
      "step": 12164
    },
    {
      "epoch": 2.2590529247910864,
      "grad_norm": 1.5632257301619292,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8960150480270386,
      "num_tokens": 423230729.0,
      "step": 12165
    },
    {
      "epoch": 2.2592386258124417,
      "grad_norm": 1.5651187716632156,
      "learning_rate": 1e-06,
      "loss": 0.2828,
      "mean_token_accuracy": 0.8972135782241821,
      "num_tokens": 423264081.0,
      "step": 12166
    },
    {
      "epoch": 2.2594243268337975,
      "grad_norm": 1.6445437929862883,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8719447255134583,
      "num_tokens": 423299546.0,
      "step": 12167
    },
    {
      "epoch": 2.259610027855153,
      "grad_norm": 1.5209090520676887,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8896152973175049,
      "num_tokens": 423335942.0,
      "step": 12168
    },
    {
      "epoch": 2.259795728876509,
      "grad_norm": 1.5682752930902033,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8927768468856812,
      "num_tokens": 423368292.0,
      "step": 12169
    },
    {
      "epoch": 2.2599814298978647,
      "grad_norm": 1.631141769019071,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8766991496086121,
      "num_tokens": 423403629.0,
      "step": 12170
    },
    {
      "epoch": 2.26016713091922,
      "grad_norm": 1.3736440318060812,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8938870429992676,
      "num_tokens": 423445940.0,
      "step": 12171
    },
    {
      "epoch": 2.2603528319405757,
      "grad_norm": 1.6494076210182709,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8783515691757202,
      "num_tokens": 423478463.0,
      "step": 12172
    },
    {
      "epoch": 2.2605385329619314,
      "grad_norm": 1.637547986526774,
      "learning_rate": 1e-06,
      "loss": 0.3005,
      "mean_token_accuracy": 0.8924853205680847,
      "num_tokens": 423507955.0,
      "step": 12173
    },
    {
      "epoch": 2.2607242339832867,
      "grad_norm": 1.4507171531985732,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8911577463150024,
      "num_tokens": 423544442.0,
      "step": 12174
    },
    {
      "epoch": 2.2609099350046424,
      "grad_norm": 1.6788291911986564,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8922204971313477,
      "num_tokens": 423573983.0,
      "step": 12175
    },
    {
      "epoch": 2.261095636025998,
      "grad_norm": 1.5985196200597922,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8868732452392578,
      "num_tokens": 423610043.0,
      "step": 12176
    },
    {
      "epoch": 2.261281337047354,
      "grad_norm": 1.5078257832653577,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8967430591583252,
      "num_tokens": 423643438.0,
      "step": 12177
    },
    {
      "epoch": 2.261467038068709,
      "grad_norm": 1.4538755096648672,
      "learning_rate": 1e-06,
      "loss": 0.2809,
      "mean_token_accuracy": 0.8998208045959473,
      "num_tokens": 423680263.0,
      "step": 12178
    },
    {
      "epoch": 2.261652739090065,
      "grad_norm": 1.6134479870011595,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8858106136322021,
      "num_tokens": 423712974.0,
      "step": 12179
    },
    {
      "epoch": 2.2618384401114207,
      "grad_norm": 1.574593357222092,
      "learning_rate": 1e-06,
      "loss": 0.3027,
      "mean_token_accuracy": 0.8951018452644348,
      "num_tokens": 423748253.0,
      "step": 12180
    },
    {
      "epoch": 2.2620241411327764,
      "grad_norm": 1.6430564257725437,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8901859521865845,
      "num_tokens": 423778385.0,
      "step": 12181
    },
    {
      "epoch": 2.2622098421541317,
      "grad_norm": 1.723124115906211,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8948689699172974,
      "num_tokens": 423810410.0,
      "step": 12182
    },
    {
      "epoch": 2.2623955431754874,
      "grad_norm": 1.621569089226634,
      "learning_rate": 1e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8680049180984497,
      "num_tokens": 423847335.0,
      "step": 12183
    },
    {
      "epoch": 2.262581244196843,
      "grad_norm": 1.8015494526241023,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8863158226013184,
      "num_tokens": 423878406.0,
      "step": 12184
    },
    {
      "epoch": 2.262766945218199,
      "grad_norm": 1.7470292849910576,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.890132486820221,
      "num_tokens": 423906632.0,
      "step": 12185
    },
    {
      "epoch": 2.262952646239554,
      "grad_norm": 1.4806472257696828,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8954184651374817,
      "num_tokens": 423940691.0,
      "step": 12186
    },
    {
      "epoch": 2.26313834726091,
      "grad_norm": 1.7408289780646014,
      "learning_rate": 1e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8692636489868164,
      "num_tokens": 423971682.0,
      "step": 12187
    },
    {
      "epoch": 2.2633240482822656,
      "grad_norm": 1.7725178588685315,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8752306699752808,
      "num_tokens": 424002763.0,
      "step": 12188
    },
    {
      "epoch": 2.263509749303621,
      "grad_norm": 1.631664793694966,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8880397081375122,
      "num_tokens": 424037173.0,
      "step": 12189
    },
    {
      "epoch": 2.2636954503249767,
      "grad_norm": 1.7146281901547022,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.879691481590271,
      "num_tokens": 424069113.0,
      "step": 12190
    },
    {
      "epoch": 2.2638811513463324,
      "grad_norm": 1.6575212336007836,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8916842341423035,
      "num_tokens": 424102211.0,
      "step": 12191
    },
    {
      "epoch": 2.264066852367688,
      "grad_norm": 1.507699730579279,
      "learning_rate": 1e-06,
      "loss": 0.294,
      "mean_token_accuracy": 0.8951829075813293,
      "num_tokens": 424140579.0,
      "step": 12192
    },
    {
      "epoch": 2.264252553389044,
      "grad_norm": 1.5498373972770545,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.880607545375824,
      "num_tokens": 424177617.0,
      "step": 12193
    },
    {
      "epoch": 2.264438254410399,
      "grad_norm": 1.4964508903794167,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8790644407272339,
      "num_tokens": 424218111.0,
      "step": 12194
    },
    {
      "epoch": 2.264623955431755,
      "grad_norm": 1.647391128844723,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8874306082725525,
      "num_tokens": 424253120.0,
      "step": 12195
    },
    {
      "epoch": 2.2648096564531106,
      "grad_norm": 1.5176329898836265,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8797526359558105,
      "num_tokens": 424292989.0,
      "step": 12196
    },
    {
      "epoch": 2.264995357474466,
      "grad_norm": 1.6503080669955896,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8761930465698242,
      "num_tokens": 424331610.0,
      "step": 12197
    },
    {
      "epoch": 2.2651810584958216,
      "grad_norm": 1.5470945822792637,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8778201341629028,
      "num_tokens": 424371355.0,
      "step": 12198
    },
    {
      "epoch": 2.2653667595171774,
      "grad_norm": 1.5049024300729728,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8892830610275269,
      "num_tokens": 424410306.0,
      "step": 12199
    },
    {
      "epoch": 2.265552460538533,
      "grad_norm": 1.57786864158888,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8857773542404175,
      "num_tokens": 424449048.0,
      "step": 12200
    },
    {
      "epoch": 2.2657381615598884,
      "grad_norm": 1.524297977388505,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8890296816825867,
      "num_tokens": 424482792.0,
      "step": 12201
    },
    {
      "epoch": 2.265923862581244,
      "grad_norm": 1.5964317543445037,
      "learning_rate": 1e-06,
      "loss": 0.2921,
      "mean_token_accuracy": 0.8984788656234741,
      "num_tokens": 424514875.0,
      "step": 12202
    },
    {
      "epoch": 2.2661095636026,
      "grad_norm": 1.5887280420628842,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.885167121887207,
      "num_tokens": 424550765.0,
      "step": 12203
    },
    {
      "epoch": 2.2662952646239556,
      "grad_norm": 1.5132155175563273,
      "learning_rate": 1e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8735224008560181,
      "num_tokens": 424589288.0,
      "step": 12204
    },
    {
      "epoch": 2.266480965645311,
      "grad_norm": 1.518662349899644,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8934168815612793,
      "num_tokens": 424625501.0,
      "step": 12205
    },
    {
      "epoch": 2.2666666666666666,
      "grad_norm": 1.5219667717322727,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8940320014953613,
      "num_tokens": 424659868.0,
      "step": 12206
    },
    {
      "epoch": 2.2668523676880223,
      "grad_norm": 1.559948006599922,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8774871230125427,
      "num_tokens": 424694528.0,
      "step": 12207
    },
    {
      "epoch": 2.267038068709378,
      "grad_norm": 1.6696640398050782,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8822964429855347,
      "num_tokens": 424728967.0,
      "step": 12208
    },
    {
      "epoch": 2.2672237697307334,
      "grad_norm": 1.7066216800353171,
      "learning_rate": 1e-06,
      "loss": 0.3084,
      "mean_token_accuracy": 0.8948710560798645,
      "num_tokens": 424759920.0,
      "step": 12209
    },
    {
      "epoch": 2.267409470752089,
      "grad_norm": 1.5010762298801008,
      "learning_rate": 1e-06,
      "loss": 0.2925,
      "mean_token_accuracy": 0.8937310576438904,
      "num_tokens": 424793034.0,
      "step": 12210
    },
    {
      "epoch": 2.267595171773445,
      "grad_norm": 1.667337813041368,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8809348344802856,
      "num_tokens": 424822725.0,
      "step": 12211
    },
    {
      "epoch": 2.2677808727948006,
      "grad_norm": 1.463336014199761,
      "learning_rate": 1e-06,
      "loss": 0.2767,
      "mean_token_accuracy": 0.9002333879470825,
      "num_tokens": 424859544.0,
      "step": 12212
    },
    {
      "epoch": 2.267966573816156,
      "grad_norm": 1.5311608491437592,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8838225603103638,
      "num_tokens": 424897749.0,
      "step": 12213
    },
    {
      "epoch": 2.2681522748375116,
      "grad_norm": 1.6212532223893292,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8930557370185852,
      "num_tokens": 424927169.0,
      "step": 12214
    },
    {
      "epoch": 2.2683379758588673,
      "grad_norm": 1.5089268653376373,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8908610343933105,
      "num_tokens": 424963153.0,
      "step": 12215
    },
    {
      "epoch": 2.268523676880223,
      "grad_norm": 1.5086869121384618,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8871418237686157,
      "num_tokens": 424997704.0,
      "step": 12216
    },
    {
      "epoch": 2.2687093779015783,
      "grad_norm": 1.675607179760332,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8848036527633667,
      "num_tokens": 425031203.0,
      "step": 12217
    },
    {
      "epoch": 2.268895078922934,
      "grad_norm": 1.7762082463429116,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8811361193656921,
      "num_tokens": 425061056.0,
      "step": 12218
    },
    {
      "epoch": 2.26908077994429,
      "grad_norm": 1.4777470706577793,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8897454738616943,
      "num_tokens": 425099861.0,
      "step": 12219
    },
    {
      "epoch": 2.269266480965645,
      "grad_norm": 1.4371142766245928,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8866416215896606,
      "num_tokens": 425137469.0,
      "step": 12220
    },
    {
      "epoch": 2.269452181987001,
      "grad_norm": 1.446545260322117,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8846898078918457,
      "num_tokens": 425177263.0,
      "step": 12221
    },
    {
      "epoch": 2.2696378830083566,
      "grad_norm": 1.6535255836209817,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8832920789718628,
      "num_tokens": 425210060.0,
      "step": 12222
    },
    {
      "epoch": 2.2698235840297123,
      "grad_norm": 1.5370672652541033,
      "learning_rate": 1e-06,
      "loss": 0.2695,
      "mean_token_accuracy": 0.9035404920578003,
      "num_tokens": 425243461.0,
      "step": 12223
    },
    {
      "epoch": 2.270009285051068,
      "grad_norm": 1.6058569843424828,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8757915496826172,
      "num_tokens": 425278967.0,
      "step": 12224
    },
    {
      "epoch": 2.2701949860724233,
      "grad_norm": 1.6960039700300118,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8900041580200195,
      "num_tokens": 425310820.0,
      "step": 12225
    },
    {
      "epoch": 2.270380687093779,
      "grad_norm": 1.5488493893899697,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8829292058944702,
      "num_tokens": 425347978.0,
      "step": 12226
    },
    {
      "epoch": 2.270566388115135,
      "grad_norm": 1.6174136053366936,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8784892559051514,
      "num_tokens": 425382849.0,
      "step": 12227
    },
    {
      "epoch": 2.27075208913649,
      "grad_norm": 1.4806048588756227,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8881983757019043,
      "num_tokens": 425419821.0,
      "step": 12228
    },
    {
      "epoch": 2.270937790157846,
      "grad_norm": 1.7657994003617863,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8818410038948059,
      "num_tokens": 425451485.0,
      "step": 12229
    },
    {
      "epoch": 2.2711234911792015,
      "grad_norm": 1.7685414352436473,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8813709616661072,
      "num_tokens": 425480559.0,
      "step": 12230
    },
    {
      "epoch": 2.2713091922005573,
      "grad_norm": 1.6031791714126564,
      "learning_rate": 1e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8662073016166687,
      "num_tokens": 425517966.0,
      "step": 12231
    },
    {
      "epoch": 2.2714948932219126,
      "grad_norm": 1.5662495258085363,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8833268880844116,
      "num_tokens": 425552414.0,
      "step": 12232
    },
    {
      "epoch": 2.2716805942432683,
      "grad_norm": 1.523386815488197,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.895751953125,
      "num_tokens": 425588395.0,
      "step": 12233
    },
    {
      "epoch": 2.271866295264624,
      "grad_norm": 1.4802254992235366,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8926005363464355,
      "num_tokens": 425627143.0,
      "step": 12234
    },
    {
      "epoch": 2.2720519962859798,
      "grad_norm": 1.698908993990015,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8864713907241821,
      "num_tokens": 425662459.0,
      "step": 12235
    },
    {
      "epoch": 2.272237697307335,
      "grad_norm": 1.43925939549496,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8943945169448853,
      "num_tokens": 425699398.0,
      "step": 12236
    },
    {
      "epoch": 2.272423398328691,
      "grad_norm": 1.5392217425767776,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8799149394035339,
      "num_tokens": 425736188.0,
      "step": 12237
    },
    {
      "epoch": 2.2726090993500465,
      "grad_norm": 1.5592985458851165,
      "learning_rate": 1e-06,
      "loss": 0.3048,
      "mean_token_accuracy": 0.8901454210281372,
      "num_tokens": 425771968.0,
      "step": 12238
    },
    {
      "epoch": 2.2727948003714022,
      "grad_norm": 1.4790889216854601,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8850860595703125,
      "num_tokens": 425808812.0,
      "step": 12239
    },
    {
      "epoch": 2.2729805013927575,
      "grad_norm": 1.4475664282181482,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8833640813827515,
      "num_tokens": 425849067.0,
      "step": 12240
    },
    {
      "epoch": 2.2731662024141133,
      "grad_norm": 1.5497117160701788,
      "learning_rate": 1e-06,
      "loss": 0.3026,
      "mean_token_accuracy": 0.8905348777770996,
      "num_tokens": 425884333.0,
      "step": 12241
    },
    {
      "epoch": 2.273351903435469,
      "grad_norm": 1.445668390293771,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8921625018119812,
      "num_tokens": 425923052.0,
      "step": 12242
    },
    {
      "epoch": 2.2735376044568243,
      "grad_norm": 1.5844223017839347,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8855066299438477,
      "num_tokens": 425961470.0,
      "step": 12243
    },
    {
      "epoch": 2.27372330547818,
      "grad_norm": 1.4122402981365483,
      "learning_rate": 1e-06,
      "loss": 0.323,
      "mean_token_accuracy": 0.8862792253494263,
      "num_tokens": 426002973.0,
      "step": 12244
    },
    {
      "epoch": 2.2739090064995358,
      "grad_norm": 1.5990591825585347,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8905586004257202,
      "num_tokens": 426033978.0,
      "step": 12245
    },
    {
      "epoch": 2.2740947075208915,
      "grad_norm": 1.554915536361527,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8985136151313782,
      "num_tokens": 426065226.0,
      "step": 12246
    },
    {
      "epoch": 2.2742804085422472,
      "grad_norm": 1.6120715190275197,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8829103112220764,
      "num_tokens": 426103041.0,
      "step": 12247
    },
    {
      "epoch": 2.2744661095636025,
      "grad_norm": 1.7040232846715273,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8875139951705933,
      "num_tokens": 426130461.0,
      "step": 12248
    },
    {
      "epoch": 2.2746518105849582,
      "grad_norm": 1.5086749889114208,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.894705593585968,
      "num_tokens": 426166560.0,
      "step": 12249
    },
    {
      "epoch": 2.274837511606314,
      "grad_norm": 1.6152761907353532,
      "learning_rate": 1e-06,
      "loss": 0.3838,
      "mean_token_accuracy": 0.8721059560775757,
      "num_tokens": 426203483.0,
      "step": 12250
    },
    {
      "epoch": 2.2750232126276693,
      "grad_norm": 1.6329525631300088,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8795100450515747,
      "num_tokens": 426234615.0,
      "step": 12251
    },
    {
      "epoch": 2.275208913649025,
      "grad_norm": 1.6981388144262,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8830532431602478,
      "num_tokens": 426265851.0,
      "step": 12252
    },
    {
      "epoch": 2.2753946146703807,
      "grad_norm": 1.6308596852725126,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8920711278915405,
      "num_tokens": 426299399.0,
      "step": 12253
    },
    {
      "epoch": 2.2755803156917365,
      "grad_norm": 1.540174643782669,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8873481750488281,
      "num_tokens": 426333024.0,
      "step": 12254
    },
    {
      "epoch": 2.2757660167130918,
      "grad_norm": 1.4999248316003226,
      "learning_rate": 1e-06,
      "loss": 0.271,
      "mean_token_accuracy": 0.9033341407775879,
      "num_tokens": 426369282.0,
      "step": 12255
    },
    {
      "epoch": 2.2759517177344475,
      "grad_norm": 1.505200051938119,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8972644209861755,
      "num_tokens": 426404332.0,
      "step": 12256
    },
    {
      "epoch": 2.276137418755803,
      "grad_norm": 1.695295682747994,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.882435142993927,
      "num_tokens": 426433548.0,
      "step": 12257
    },
    {
      "epoch": 2.276323119777159,
      "grad_norm": 1.7363771574606441,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8921173810958862,
      "num_tokens": 426460632.0,
      "step": 12258
    },
    {
      "epoch": 2.2765088207985142,
      "grad_norm": 1.706273292999246,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8744696378707886,
      "num_tokens": 426498723.0,
      "step": 12259
    },
    {
      "epoch": 2.27669452181987,
      "grad_norm": 1.5356329576060521,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8857457637786865,
      "num_tokens": 426534609.0,
      "step": 12260
    },
    {
      "epoch": 2.2768802228412257,
      "grad_norm": 1.6998944646602085,
      "learning_rate": 1e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8645750880241394,
      "num_tokens": 426566426.0,
      "step": 12261
    },
    {
      "epoch": 2.2770659238625814,
      "grad_norm": 1.7214300806105556,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8820818066596985,
      "num_tokens": 426596598.0,
      "step": 12262
    },
    {
      "epoch": 2.2772516248839367,
      "grad_norm": 1.6196164590002409,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.88935786485672,
      "num_tokens": 426628362.0,
      "step": 12263
    },
    {
      "epoch": 2.2774373259052925,
      "grad_norm": 1.6574078567712525,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8868955373764038,
      "num_tokens": 426659096.0,
      "step": 12264
    },
    {
      "epoch": 2.277623026926648,
      "grad_norm": 1.6662482095958537,
      "learning_rate": 1e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8642597794532776,
      "num_tokens": 426694343.0,
      "step": 12265
    },
    {
      "epoch": 2.2778087279480035,
      "grad_norm": 1.6778047428987262,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8800389170646667,
      "num_tokens": 426728828.0,
      "step": 12266
    },
    {
      "epoch": 2.277994428969359,
      "grad_norm": 1.5219757530418097,
      "learning_rate": 1e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8715687394142151,
      "num_tokens": 426766000.0,
      "step": 12267
    },
    {
      "epoch": 2.278180129990715,
      "grad_norm": 1.5667744188408577,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8750193119049072,
      "num_tokens": 426802878.0,
      "step": 12268
    },
    {
      "epoch": 2.2783658310120707,
      "grad_norm": 1.5221064973016993,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8864749670028687,
      "num_tokens": 426840310.0,
      "step": 12269
    },
    {
      "epoch": 2.2785515320334264,
      "grad_norm": 1.7536516775448683,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8823765516281128,
      "num_tokens": 426872483.0,
      "step": 12270
    },
    {
      "epoch": 2.2787372330547817,
      "grad_norm": 1.5425638943745206,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8937671184539795,
      "num_tokens": 426911179.0,
      "step": 12271
    },
    {
      "epoch": 2.2789229340761374,
      "grad_norm": 1.41985793295674,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8844192624092102,
      "num_tokens": 426952700.0,
      "step": 12272
    },
    {
      "epoch": 2.279108635097493,
      "grad_norm": 1.565993453595827,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8785762786865234,
      "num_tokens": 426993266.0,
      "step": 12273
    },
    {
      "epoch": 2.2792943361188485,
      "grad_norm": 1.4961425192070272,
      "learning_rate": 1e-06,
      "loss": 0.281,
      "mean_token_accuracy": 0.9019961357116699,
      "num_tokens": 427027799.0,
      "step": 12274
    },
    {
      "epoch": 2.279480037140204,
      "grad_norm": 1.5698994860272653,
      "learning_rate": 1e-06,
      "loss": 0.3056,
      "mean_token_accuracy": 0.8917255401611328,
      "num_tokens": 427060117.0,
      "step": 12275
    },
    {
      "epoch": 2.27966573816156,
      "grad_norm": 1.6553633155919638,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8832074403762817,
      "num_tokens": 427091932.0,
      "step": 12276
    },
    {
      "epoch": 2.2798514391829157,
      "grad_norm": 1.6576179219270315,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8855299949645996,
      "num_tokens": 427123719.0,
      "step": 12277
    },
    {
      "epoch": 2.280037140204271,
      "grad_norm": 1.4729673677804542,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8951259255409241,
      "num_tokens": 427162109.0,
      "step": 12278
    },
    {
      "epoch": 2.2802228412256267,
      "grad_norm": 1.547090228345458,
      "learning_rate": 1e-06,
      "loss": 0.2711,
      "mean_token_accuracy": 0.9039972424507141,
      "num_tokens": 427193311.0,
      "step": 12279
    },
    {
      "epoch": 2.2804085422469824,
      "grad_norm": 1.496545741775196,
      "learning_rate": 1e-06,
      "loss": 0.2887,
      "mean_token_accuracy": 0.8998801112174988,
      "num_tokens": 427227301.0,
      "step": 12280
    },
    {
      "epoch": 2.280594243268338,
      "grad_norm": 1.5377300810649686,
      "learning_rate": 1e-06,
      "loss": 0.3145,
      "mean_token_accuracy": 0.892279863357544,
      "num_tokens": 427263816.0,
      "step": 12281
    },
    {
      "epoch": 2.2807799442896934,
      "grad_norm": 1.6075618214230158,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8819670081138611,
      "num_tokens": 427297746.0,
      "step": 12282
    },
    {
      "epoch": 2.280965645311049,
      "grad_norm": 1.6266757955336255,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8928524255752563,
      "num_tokens": 427329228.0,
      "step": 12283
    },
    {
      "epoch": 2.281151346332405,
      "grad_norm": 1.5767306044233207,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8754271268844604,
      "num_tokens": 427365074.0,
      "step": 12284
    },
    {
      "epoch": 2.2813370473537606,
      "grad_norm": 1.6142451505475224,
      "learning_rate": 1e-06,
      "loss": 0.3102,
      "mean_token_accuracy": 0.8908183574676514,
      "num_tokens": 427397806.0,
      "step": 12285
    },
    {
      "epoch": 2.281522748375116,
      "grad_norm": 1.4725951007560307,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8892754316329956,
      "num_tokens": 427435764.0,
      "step": 12286
    },
    {
      "epoch": 2.2817084493964717,
      "grad_norm": 1.5418355533909311,
      "learning_rate": 1e-06,
      "loss": 0.2965,
      "mean_token_accuracy": 0.8945199251174927,
      "num_tokens": 427471467.0,
      "step": 12287
    },
    {
      "epoch": 2.2818941504178274,
      "grad_norm": 1.5480049736877421,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8829579949378967,
      "num_tokens": 427512877.0,
      "step": 12288
    },
    {
      "epoch": 2.2820798514391827,
      "grad_norm": 1.690222128973761,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8755865097045898,
      "num_tokens": 427545054.0,
      "step": 12289
    },
    {
      "epoch": 2.2822655524605384,
      "grad_norm": 1.5547445749916582,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8884883522987366,
      "num_tokens": 427583300.0,
      "step": 12290
    },
    {
      "epoch": 2.282451253481894,
      "grad_norm": 1.5884871189948615,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8938809633255005,
      "num_tokens": 427613276.0,
      "step": 12291
    },
    {
      "epoch": 2.28263695450325,
      "grad_norm": 1.4123399364263192,
      "learning_rate": 1e-06,
      "loss": 0.2911,
      "mean_token_accuracy": 0.8951535224914551,
      "num_tokens": 427651117.0,
      "step": 12292
    },
    {
      "epoch": 2.2828226555246056,
      "grad_norm": 1.4836932872630701,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8842402696609497,
      "num_tokens": 427688608.0,
      "step": 12293
    },
    {
      "epoch": 2.283008356545961,
      "grad_norm": 1.5404207667443452,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8910123109817505,
      "num_tokens": 427726355.0,
      "step": 12294
    },
    {
      "epoch": 2.2831940575673166,
      "grad_norm": 1.5428301326178153,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8884682655334473,
      "num_tokens": 427763477.0,
      "step": 12295
    },
    {
      "epoch": 2.2833797585886724,
      "grad_norm": 1.5195562925980057,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8836897611618042,
      "num_tokens": 427800267.0,
      "step": 12296
    },
    {
      "epoch": 2.2835654596100277,
      "grad_norm": 1.6882243185903947,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8841665983200073,
      "num_tokens": 427835042.0,
      "step": 12297
    },
    {
      "epoch": 2.2837511606313834,
      "grad_norm": 1.5905994014997191,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.87123703956604,
      "num_tokens": 427874049.0,
      "step": 12298
    },
    {
      "epoch": 2.283936861652739,
      "grad_norm": 1.5509400095609633,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8958204984664917,
      "num_tokens": 427907232.0,
      "step": 12299
    },
    {
      "epoch": 2.284122562674095,
      "grad_norm": 1.6180390653005174,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8879309892654419,
      "num_tokens": 427941761.0,
      "step": 12300
    },
    {
      "epoch": 2.28430826369545,
      "grad_norm": 1.533985769310548,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8908635377883911,
      "num_tokens": 427974510.0,
      "step": 12301
    },
    {
      "epoch": 2.284493964716806,
      "grad_norm": 1.6266147674679265,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8712258338928223,
      "num_tokens": 428012497.0,
      "step": 12302
    },
    {
      "epoch": 2.2846796657381616,
      "grad_norm": 1.5165216897742335,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8805235624313354,
      "num_tokens": 428051103.0,
      "step": 12303
    },
    {
      "epoch": 2.2848653667595173,
      "grad_norm": 1.7198743082367418,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8790228366851807,
      "num_tokens": 428080004.0,
      "step": 12304
    },
    {
      "epoch": 2.2850510677808726,
      "grad_norm": 1.7285259950717926,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8867719173431396,
      "num_tokens": 428114084.0,
      "step": 12305
    },
    {
      "epoch": 2.2852367688022284,
      "grad_norm": 1.522074466026944,
      "learning_rate": 1e-06,
      "loss": 0.3023,
      "mean_token_accuracy": 0.8946176767349243,
      "num_tokens": 428148925.0,
      "step": 12306
    },
    {
      "epoch": 2.285422469823584,
      "grad_norm": 1.6121154438169423,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8812583684921265,
      "num_tokens": 428184785.0,
      "step": 12307
    },
    {
      "epoch": 2.28560817084494,
      "grad_norm": 1.4910175038087,
      "learning_rate": 1e-06,
      "loss": 0.288,
      "mean_token_accuracy": 0.8970145583152771,
      "num_tokens": 428222831.0,
      "step": 12308
    },
    {
      "epoch": 2.285793871866295,
      "grad_norm": 1.6385746356130848,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.889836311340332,
      "num_tokens": 428253520.0,
      "step": 12309
    },
    {
      "epoch": 2.285979572887651,
      "grad_norm": 1.5196692798864602,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8922421336174011,
      "num_tokens": 428290755.0,
      "step": 12310
    },
    {
      "epoch": 2.2861652739090066,
      "grad_norm": 1.6115030377948263,
      "learning_rate": 1e-06,
      "loss": 0.2993,
      "mean_token_accuracy": 0.8930627107620239,
      "num_tokens": 428320774.0,
      "step": 12311
    },
    {
      "epoch": 2.286350974930362,
      "grad_norm": 1.4740612798744837,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8791953325271606,
      "num_tokens": 428357703.0,
      "step": 12312
    },
    {
      "epoch": 2.2865366759517176,
      "grad_norm": 1.5044123007562378,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8864376544952393,
      "num_tokens": 428395384.0,
      "step": 12313
    },
    {
      "epoch": 2.2867223769730733,
      "grad_norm": 1.4308424499871577,
      "learning_rate": 1e-06,
      "loss": 0.2846,
      "mean_token_accuracy": 0.9007558226585388,
      "num_tokens": 428431637.0,
      "step": 12314
    },
    {
      "epoch": 2.286908077994429,
      "grad_norm": 1.529476846979947,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8763707876205444,
      "num_tokens": 428467009.0,
      "step": 12315
    },
    {
      "epoch": 2.287093779015785,
      "grad_norm": 1.4681528570574844,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8921306133270264,
      "num_tokens": 428505161.0,
      "step": 12316
    },
    {
      "epoch": 2.28727948003714,
      "grad_norm": 1.6037875310148721,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8730263710021973,
      "num_tokens": 428541042.0,
      "step": 12317
    },
    {
      "epoch": 2.287465181058496,
      "grad_norm": 1.530157865502646,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8811511993408203,
      "num_tokens": 428579787.0,
      "step": 12318
    },
    {
      "epoch": 2.2876508820798516,
      "grad_norm": 1.6139508031702379,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8723312616348267,
      "num_tokens": 428616569.0,
      "step": 12319
    },
    {
      "epoch": 2.287836583101207,
      "grad_norm": 1.5599314354700928,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8823387622833252,
      "num_tokens": 428650807.0,
      "step": 12320
    },
    {
      "epoch": 2.2880222841225626,
      "grad_norm": 1.5031843251783286,
      "learning_rate": 1e-06,
      "loss": 0.3114,
      "mean_token_accuracy": 0.8879550695419312,
      "num_tokens": 428685973.0,
      "step": 12321
    },
    {
      "epoch": 2.2882079851439183,
      "grad_norm": 1.6509911864819036,
      "learning_rate": 1e-06,
      "loss": 0.2871,
      "mean_token_accuracy": 0.9033843874931335,
      "num_tokens": 428715200.0,
      "step": 12322
    },
    {
      "epoch": 2.288393686165274,
      "grad_norm": 1.5447476856433782,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8804446458816528,
      "num_tokens": 428751910.0,
      "step": 12323
    },
    {
      "epoch": 2.2885793871866293,
      "grad_norm": 1.8981428514123606,
      "learning_rate": 1e-06,
      "loss": 0.2815,
      "mean_token_accuracy": 0.9025066494941711,
      "num_tokens": 428774525.0,
      "step": 12324
    },
    {
      "epoch": 2.288765088207985,
      "grad_norm": 1.5823154891411522,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8834413290023804,
      "num_tokens": 428806970.0,
      "step": 12325
    },
    {
      "epoch": 2.288950789229341,
      "grad_norm": 1.6275594346517004,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8980627059936523,
      "num_tokens": 428835407.0,
      "step": 12326
    },
    {
      "epoch": 2.2891364902506965,
      "grad_norm": 1.5664229507436902,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8770512938499451,
      "num_tokens": 428871498.0,
      "step": 12327
    },
    {
      "epoch": 2.289322191272052,
      "grad_norm": 1.5143259327123717,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8823621273040771,
      "num_tokens": 428912369.0,
      "step": 12328
    },
    {
      "epoch": 2.2895078922934076,
      "grad_norm": 1.4926337327195984,
      "learning_rate": 1e-06,
      "loss": 0.2866,
      "mean_token_accuracy": 0.8964172601699829,
      "num_tokens": 428948629.0,
      "step": 12329
    },
    {
      "epoch": 2.2896935933147633,
      "grad_norm": 1.5125525943608669,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8793537020683289,
      "num_tokens": 428985926.0,
      "step": 12330
    },
    {
      "epoch": 2.289879294336119,
      "grad_norm": 1.5025238069970879,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.883552610874176,
      "num_tokens": 429026435.0,
      "step": 12331
    },
    {
      "epoch": 2.2900649953574743,
      "grad_norm": 1.5973448299226611,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8871113061904907,
      "num_tokens": 429062857.0,
      "step": 12332
    },
    {
      "epoch": 2.29025069637883,
      "grad_norm": 1.5473289207354262,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8877224922180176,
      "num_tokens": 429103751.0,
      "step": 12333
    },
    {
      "epoch": 2.290436397400186,
      "grad_norm": 1.6338377219504885,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8813294768333435,
      "num_tokens": 429140302.0,
      "step": 12334
    },
    {
      "epoch": 2.290622098421541,
      "grad_norm": 1.549346612355894,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.887712836265564,
      "num_tokens": 429179006.0,
      "step": 12335
    },
    {
      "epoch": 2.290807799442897,
      "grad_norm": 1.5339635604221964,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8783423900604248,
      "num_tokens": 429216644.0,
      "step": 12336
    },
    {
      "epoch": 2.2909935004642525,
      "grad_norm": 1.519388524040644,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8912856578826904,
      "num_tokens": 429252380.0,
      "step": 12337
    },
    {
      "epoch": 2.2911792014856083,
      "grad_norm": 1.5083919728422333,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8904159069061279,
      "num_tokens": 429286866.0,
      "step": 12338
    },
    {
      "epoch": 2.291364902506964,
      "grad_norm": 1.512868394780242,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8792716264724731,
      "num_tokens": 429325684.0,
      "step": 12339
    },
    {
      "epoch": 2.2915506035283193,
      "grad_norm": 1.7372262756591403,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8706647157669067,
      "num_tokens": 429359572.0,
      "step": 12340
    },
    {
      "epoch": 2.291736304549675,
      "grad_norm": 1.4942985924952132,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.885897159576416,
      "num_tokens": 429400803.0,
      "step": 12341
    },
    {
      "epoch": 2.2919220055710308,
      "grad_norm": 1.6903512576326132,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8888918161392212,
      "num_tokens": 429430354.0,
      "step": 12342
    },
    {
      "epoch": 2.292107706592386,
      "grad_norm": 1.55842082624507,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8845717906951904,
      "num_tokens": 429465329.0,
      "step": 12343
    },
    {
      "epoch": 2.292293407613742,
      "grad_norm": 1.528033234210405,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8878426551818848,
      "num_tokens": 429503846.0,
      "step": 12344
    },
    {
      "epoch": 2.2924791086350975,
      "grad_norm": 1.5016077057014465,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8793175220489502,
      "num_tokens": 429542962.0,
      "step": 12345
    },
    {
      "epoch": 2.2926648096564533,
      "grad_norm": 1.5073498489676056,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8877602219581604,
      "num_tokens": 429578043.0,
      "step": 12346
    },
    {
      "epoch": 2.2928505106778085,
      "grad_norm": 1.559582811973556,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.89274001121521,
      "num_tokens": 429612440.0,
      "step": 12347
    },
    {
      "epoch": 2.2930362116991643,
      "grad_norm": 1.5760694928529548,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.884678304195404,
      "num_tokens": 429646166.0,
      "step": 12348
    },
    {
      "epoch": 2.29322191272052,
      "grad_norm": 1.5123674797512598,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8917468190193176,
      "num_tokens": 429682174.0,
      "step": 12349
    },
    {
      "epoch": 2.2934076137418757,
      "grad_norm": 1.4828066337935981,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8895300626754761,
      "num_tokens": 429719614.0,
      "step": 12350
    },
    {
      "epoch": 2.293593314763231,
      "grad_norm": 1.6688180858632538,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8843669891357422,
      "num_tokens": 429752068.0,
      "step": 12351
    },
    {
      "epoch": 2.2937790157845868,
      "grad_norm": 1.471652432107437,
      "learning_rate": 1e-06,
      "loss": 0.2736,
      "mean_token_accuracy": 0.9039424061775208,
      "num_tokens": 429785589.0,
      "step": 12352
    },
    {
      "epoch": 2.2939647168059425,
      "grad_norm": 1.4892702772924082,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8870741724967957,
      "num_tokens": 429822969.0,
      "step": 12353
    },
    {
      "epoch": 2.2941504178272982,
      "grad_norm": 1.687292090150495,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8907836675643921,
      "num_tokens": 429850812.0,
      "step": 12354
    },
    {
      "epoch": 2.2943361188486535,
      "grad_norm": 1.6287390582262447,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8874129056930542,
      "num_tokens": 429881722.0,
      "step": 12355
    },
    {
      "epoch": 2.2945218198700092,
      "grad_norm": 1.571647604787711,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8915976285934448,
      "num_tokens": 429914475.0,
      "step": 12356
    },
    {
      "epoch": 2.294707520891365,
      "grad_norm": 1.6017913936834176,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8736569285392761,
      "num_tokens": 429951653.0,
      "step": 12357
    },
    {
      "epoch": 2.2948932219127203,
      "grad_norm": 1.5137982316924403,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8896769285202026,
      "num_tokens": 429989106.0,
      "step": 12358
    },
    {
      "epoch": 2.295078922934076,
      "grad_norm": 1.4819628447999862,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8918440341949463,
      "num_tokens": 430026686.0,
      "step": 12359
    },
    {
      "epoch": 2.2952646239554317,
      "grad_norm": 1.6050541142859112,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8874180316925049,
      "num_tokens": 430064638.0,
      "step": 12360
    },
    {
      "epoch": 2.2954503249767875,
      "grad_norm": 8.487365406200057,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8895103931427002,
      "num_tokens": 430106771.0,
      "step": 12361
    },
    {
      "epoch": 2.295636025998143,
      "grad_norm": 1.6873801671264903,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8897138237953186,
      "num_tokens": 430139827.0,
      "step": 12362
    },
    {
      "epoch": 2.2958217270194985,
      "grad_norm": 1.5892059778706689,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8806684017181396,
      "num_tokens": 430175642.0,
      "step": 12363
    },
    {
      "epoch": 2.2960074280408542,
      "grad_norm": 1.4825402619274317,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8877220153808594,
      "num_tokens": 430214386.0,
      "step": 12364
    },
    {
      "epoch": 2.29619312906221,
      "grad_norm": 1.8424623486654887,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8735983967781067,
      "num_tokens": 430239344.0,
      "step": 12365
    },
    {
      "epoch": 2.2963788300835652,
      "grad_norm": 1.6274005406918484,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8813707828521729,
      "num_tokens": 430271940.0,
      "step": 12366
    },
    {
      "epoch": 2.296564531104921,
      "grad_norm": 1.4086133580221538,
      "learning_rate": 1e-06,
      "loss": 0.2716,
      "mean_token_accuracy": 0.9047752618789673,
      "num_tokens": 430307725.0,
      "step": 12367
    },
    {
      "epoch": 2.2967502321262767,
      "grad_norm": 1.4624111041255101,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8871239423751831,
      "num_tokens": 430348696.0,
      "step": 12368
    },
    {
      "epoch": 2.2969359331476324,
      "grad_norm": 1.6810100167820676,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8719663619995117,
      "num_tokens": 430383623.0,
      "step": 12369
    },
    {
      "epoch": 2.2971216341689877,
      "grad_norm": 1.5299259571389736,
      "learning_rate": 1e-06,
      "loss": 0.2795,
      "mean_token_accuracy": 0.9051126837730408,
      "num_tokens": 430414008.0,
      "step": 12370
    },
    {
      "epoch": 2.2973073351903435,
      "grad_norm": 1.6839751074011093,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8685038685798645,
      "num_tokens": 430447865.0,
      "step": 12371
    },
    {
      "epoch": 2.297493036211699,
      "grad_norm": 1.6347341647482552,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.887553334236145,
      "num_tokens": 430482708.0,
      "step": 12372
    },
    {
      "epoch": 2.297678737233055,
      "grad_norm": 1.454200152567897,
      "learning_rate": 1e-06,
      "loss": 0.2973,
      "mean_token_accuracy": 0.8955596685409546,
      "num_tokens": 430522229.0,
      "step": 12373
    },
    {
      "epoch": 2.2978644382544102,
      "grad_norm": 1.4881664158241266,
      "learning_rate": 1e-06,
      "loss": 0.299,
      "mean_token_accuracy": 0.8965099453926086,
      "num_tokens": 430559689.0,
      "step": 12374
    },
    {
      "epoch": 2.298050139275766,
      "grad_norm": 1.6221040975523426,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8800786137580872,
      "num_tokens": 430595365.0,
      "step": 12375
    },
    {
      "epoch": 2.2982358402971217,
      "grad_norm": 1.4574154926521612,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8914840817451477,
      "num_tokens": 430640824.0,
      "step": 12376
    },
    {
      "epoch": 2.2984215413184774,
      "grad_norm": 1.7360637000253967,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8749839663505554,
      "num_tokens": 430673458.0,
      "step": 12377
    },
    {
      "epoch": 2.2986072423398327,
      "grad_norm": 1.5423485326932769,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8845367431640625,
      "num_tokens": 430707534.0,
      "step": 12378
    },
    {
      "epoch": 2.2987929433611884,
      "grad_norm": 1.7211155159044587,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8854855298995972,
      "num_tokens": 430737534.0,
      "step": 12379
    },
    {
      "epoch": 2.298978644382544,
      "grad_norm": 1.7117937067396265,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8772037029266357,
      "num_tokens": 430769139.0,
      "step": 12380
    },
    {
      "epoch": 2.2991643454039,
      "grad_norm": 1.7092972834672588,
      "learning_rate": 1e-06,
      "loss": 0.3476,
      "mean_token_accuracy": 0.8800951838493347,
      "num_tokens": 430799534.0,
      "step": 12381
    },
    {
      "epoch": 2.299350046425255,
      "grad_norm": 1.5680571590444448,
      "learning_rate": 1e-06,
      "loss": 0.3001,
      "mean_token_accuracy": 0.8937257528305054,
      "num_tokens": 430831228.0,
      "step": 12382
    },
    {
      "epoch": 2.299535747446611,
      "grad_norm": 1.5317402598967298,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8834327459335327,
      "num_tokens": 430872331.0,
      "step": 12383
    },
    {
      "epoch": 2.2997214484679667,
      "grad_norm": 1.4265500687496386,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8965864181518555,
      "num_tokens": 430911971.0,
      "step": 12384
    },
    {
      "epoch": 2.2999071494893224,
      "grad_norm": 1.4978720102052383,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8775342106819153,
      "num_tokens": 430953296.0,
      "step": 12385
    },
    {
      "epoch": 2.3000928505106777,
      "grad_norm": 1.5709238138379058,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8762115240097046,
      "num_tokens": 430992906.0,
      "step": 12386
    },
    {
      "epoch": 2.3002785515320334,
      "grad_norm": 1.541284706451945,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8904868960380554,
      "num_tokens": 431028615.0,
      "step": 12387
    },
    {
      "epoch": 2.300464252553389,
      "grad_norm": 1.6469996537611,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8887267112731934,
      "num_tokens": 431060667.0,
      "step": 12388
    },
    {
      "epoch": 2.3006499535747444,
      "grad_norm": 1.6763449287996033,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8865842819213867,
      "num_tokens": 431094060.0,
      "step": 12389
    },
    {
      "epoch": 2.3008356545961,
      "grad_norm": 1.5468446289933773,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8929064273834229,
      "num_tokens": 431127508.0,
      "step": 12390
    },
    {
      "epoch": 2.301021355617456,
      "grad_norm": 1.6425377480720258,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8823139667510986,
      "num_tokens": 431158040.0,
      "step": 12391
    },
    {
      "epoch": 2.3012070566388116,
      "grad_norm": 1.5406337115331734,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8805776834487915,
      "num_tokens": 431190942.0,
      "step": 12392
    },
    {
      "epoch": 2.3013927576601674,
      "grad_norm": 1.498915480454061,
      "learning_rate": 1e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.8747832775115967,
      "num_tokens": 431230317.0,
      "step": 12393
    },
    {
      "epoch": 2.3015784586815227,
      "grad_norm": 1.5776689080972035,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8950543999671936,
      "num_tokens": 431267016.0,
      "step": 12394
    },
    {
      "epoch": 2.3017641597028784,
      "grad_norm": 1.4898480015998634,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8828955888748169,
      "num_tokens": 431307661.0,
      "step": 12395
    },
    {
      "epoch": 2.301949860724234,
      "grad_norm": 1.7406507768040118,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8896864056587219,
      "num_tokens": 431334658.0,
      "step": 12396
    },
    {
      "epoch": 2.3021355617455894,
      "grad_norm": 1.519688911572282,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8895745873451233,
      "num_tokens": 431369804.0,
      "step": 12397
    },
    {
      "epoch": 2.302321262766945,
      "grad_norm": 1.7115505355036005,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8928092122077942,
      "num_tokens": 431399265.0,
      "step": 12398
    },
    {
      "epoch": 2.302506963788301,
      "grad_norm": 1.4606358673563629,
      "learning_rate": 1e-06,
      "loss": 0.2907,
      "mean_token_accuracy": 0.8968239426612854,
      "num_tokens": 431436891.0,
      "step": 12399
    },
    {
      "epoch": 2.3026926648096566,
      "grad_norm": 1.541421787037956,
      "learning_rate": 1e-06,
      "loss": 0.2915,
      "mean_token_accuracy": 0.8973079919815063,
      "num_tokens": 431474420.0,
      "step": 12400
    },
    {
      "epoch": 2.302878365831012,
      "grad_norm": 1.6019067163961338,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8863480091094971,
      "num_tokens": 431507016.0,
      "step": 12401
    },
    {
      "epoch": 2.3030640668523676,
      "grad_norm": 1.5200561057879844,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8800308108329773,
      "num_tokens": 431549421.0,
      "step": 12402
    },
    {
      "epoch": 2.3032497678737234,
      "grad_norm": 1.4447688878655482,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.8930017948150635,
      "num_tokens": 431590990.0,
      "step": 12403
    },
    {
      "epoch": 2.303435468895079,
      "grad_norm": 1.5690871987999497,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8899904489517212,
      "num_tokens": 431625706.0,
      "step": 12404
    },
    {
      "epoch": 2.3036211699164344,
      "grad_norm": 1.7685907825978997,
      "learning_rate": 1e-06,
      "loss": 0.3851,
      "mean_token_accuracy": 0.8682617545127869,
      "num_tokens": 431659174.0,
      "step": 12405
    },
    {
      "epoch": 2.30380687093779,
      "grad_norm": 1.6283452769684077,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.88844895362854,
      "num_tokens": 431691012.0,
      "step": 12406
    },
    {
      "epoch": 2.303992571959146,
      "grad_norm": 1.7615424484245301,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.882352888584137,
      "num_tokens": 431720799.0,
      "step": 12407
    },
    {
      "epoch": 2.3041782729805016,
      "grad_norm": 1.6677840927986884,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8865841627120972,
      "num_tokens": 431752297.0,
      "step": 12408
    },
    {
      "epoch": 2.304363974001857,
      "grad_norm": 1.6417116296998644,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8787073493003845,
      "num_tokens": 431786447.0,
      "step": 12409
    },
    {
      "epoch": 2.3045496750232126,
      "grad_norm": 1.6836966184747275,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.869411826133728,
      "num_tokens": 431821123.0,
      "step": 12410
    },
    {
      "epoch": 2.3047353760445684,
      "grad_norm": 1.5235037718251647,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8851234912872314,
      "num_tokens": 431855640.0,
      "step": 12411
    },
    {
      "epoch": 2.3049210770659236,
      "grad_norm": 1.5220110811953496,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8882769346237183,
      "num_tokens": 431891608.0,
      "step": 12412
    },
    {
      "epoch": 2.3051067780872794,
      "grad_norm": 1.6905452342331035,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8916386961936951,
      "num_tokens": 431919138.0,
      "step": 12413
    },
    {
      "epoch": 2.305292479108635,
      "grad_norm": 1.5650210085347032,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8794362545013428,
      "num_tokens": 431954426.0,
      "step": 12414
    },
    {
      "epoch": 2.305478180129991,
      "grad_norm": 1.634506278056301,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.8791157007217407,
      "num_tokens": 431989323.0,
      "step": 12415
    },
    {
      "epoch": 2.3056638811513466,
      "grad_norm": 1.516044691741867,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8864209055900574,
      "num_tokens": 432029072.0,
      "step": 12416
    },
    {
      "epoch": 2.305849582172702,
      "grad_norm": 1.4330352910248967,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8895951509475708,
      "num_tokens": 432066838.0,
      "step": 12417
    },
    {
      "epoch": 2.3060352831940576,
      "grad_norm": 1.5285407572540213,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8872412443161011,
      "num_tokens": 432105286.0,
      "step": 12418
    },
    {
      "epoch": 2.3062209842154133,
      "grad_norm": 1.5594911898762343,
      "learning_rate": 1e-06,
      "loss": 0.3022,
      "mean_token_accuracy": 0.8932262063026428,
      "num_tokens": 432141126.0,
      "step": 12419
    },
    {
      "epoch": 2.3064066852367686,
      "grad_norm": 1.5662032210436374,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8803594708442688,
      "num_tokens": 432179027.0,
      "step": 12420
    },
    {
      "epoch": 2.3065923862581243,
      "grad_norm": 1.9726972147073267,
      "learning_rate": 1e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8525789976119995,
      "num_tokens": 432205734.0,
      "step": 12421
    },
    {
      "epoch": 2.30677808727948,
      "grad_norm": 1.712779343040693,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8776564598083496,
      "num_tokens": 432240603.0,
      "step": 12422
    },
    {
      "epoch": 2.306963788300836,
      "grad_norm": 1.464055236889131,
      "learning_rate": 1e-06,
      "loss": 0.3056,
      "mean_token_accuracy": 0.8919917941093445,
      "num_tokens": 432279370.0,
      "step": 12423
    },
    {
      "epoch": 2.307149489322191,
      "grad_norm": 1.4431381232661362,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8870775103569031,
      "num_tokens": 432316657.0,
      "step": 12424
    },
    {
      "epoch": 2.307335190343547,
      "grad_norm": 1.5634637906085433,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8923377990722656,
      "num_tokens": 432353609.0,
      "step": 12425
    },
    {
      "epoch": 2.3075208913649026,
      "grad_norm": 1.5210578721468009,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8836737871170044,
      "num_tokens": 432394757.0,
      "step": 12426
    },
    {
      "epoch": 2.3077065923862583,
      "grad_norm": 1.5888133026351934,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8771868348121643,
      "num_tokens": 432431949.0,
      "step": 12427
    },
    {
      "epoch": 2.3078922934076136,
      "grad_norm": 1.4796318550563798,
      "learning_rate": 1e-06,
      "loss": 0.2852,
      "mean_token_accuracy": 0.8992867469787598,
      "num_tokens": 432471687.0,
      "step": 12428
    },
    {
      "epoch": 2.3080779944289693,
      "grad_norm": 1.716528722198663,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8914520740509033,
      "num_tokens": 432500477.0,
      "step": 12429
    },
    {
      "epoch": 2.308263695450325,
      "grad_norm": 1.490175243016477,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8891273736953735,
      "num_tokens": 432535438.0,
      "step": 12430
    },
    {
      "epoch": 2.308449396471681,
      "grad_norm": 1.705217100642202,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8944360017776489,
      "num_tokens": 432563886.0,
      "step": 12431
    },
    {
      "epoch": 2.308635097493036,
      "grad_norm": 1.7605454454763918,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8873074054718018,
      "num_tokens": 432593762.0,
      "step": 12432
    },
    {
      "epoch": 2.308820798514392,
      "grad_norm": 1.5755700160305286,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.8965091705322266,
      "num_tokens": 432629491.0,
      "step": 12433
    },
    {
      "epoch": 2.3090064995357475,
      "grad_norm": 1.5411905460190582,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8902349472045898,
      "num_tokens": 432663917.0,
      "step": 12434
    },
    {
      "epoch": 2.309192200557103,
      "grad_norm": 1.713005903527186,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8872735500335693,
      "num_tokens": 432697186.0,
      "step": 12435
    },
    {
      "epoch": 2.3093779015784586,
      "grad_norm": 1.5742835722017952,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8886395692825317,
      "num_tokens": 432733215.0,
      "step": 12436
    },
    {
      "epoch": 2.3095636025998143,
      "grad_norm": 1.5128119134422564,
      "learning_rate": 1e-06,
      "loss": 0.2785,
      "mean_token_accuracy": 0.9011542797088623,
      "num_tokens": 432769207.0,
      "step": 12437
    },
    {
      "epoch": 2.30974930362117,
      "grad_norm": 1.689989640479335,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8834211230278015,
      "num_tokens": 432798906.0,
      "step": 12438
    },
    {
      "epoch": 2.3099350046425258,
      "grad_norm": 1.4691206788978242,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8875117301940918,
      "num_tokens": 432836206.0,
      "step": 12439
    },
    {
      "epoch": 2.310120705663881,
      "grad_norm": 1.550938971738782,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8829021453857422,
      "num_tokens": 432871916.0,
      "step": 12440
    },
    {
      "epoch": 2.310306406685237,
      "grad_norm": 1.7765327785803502,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8797754049301147,
      "num_tokens": 432900653.0,
      "step": 12441
    },
    {
      "epoch": 2.3104921077065925,
      "grad_norm": 1.5834671907430993,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8857598304748535,
      "num_tokens": 432934641.0,
      "step": 12442
    },
    {
      "epoch": 2.310677808727948,
      "grad_norm": 1.5387523234598188,
      "learning_rate": 1e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8717820644378662,
      "num_tokens": 432971827.0,
      "step": 12443
    },
    {
      "epoch": 2.3108635097493035,
      "grad_norm": 1.4440137930488772,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8961009979248047,
      "num_tokens": 433011911.0,
      "step": 12444
    },
    {
      "epoch": 2.3110492107706593,
      "grad_norm": 1.4307955691141179,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.893979549407959,
      "num_tokens": 433050257.0,
      "step": 12445
    },
    {
      "epoch": 2.311234911792015,
      "grad_norm": 1.538848020865748,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8870619535446167,
      "num_tokens": 433084894.0,
      "step": 12446
    },
    {
      "epoch": 2.3114206128133703,
      "grad_norm": 1.5811290991529825,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8856436014175415,
      "num_tokens": 433120931.0,
      "step": 12447
    },
    {
      "epoch": 2.311606313834726,
      "grad_norm": 1.687453483422525,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8736242055892944,
      "num_tokens": 433150845.0,
      "step": 12448
    },
    {
      "epoch": 2.3117920148560818,
      "grad_norm": 1.5701973329203158,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8805999755859375,
      "num_tokens": 433186474.0,
      "step": 12449
    },
    {
      "epoch": 2.3119777158774375,
      "grad_norm": 1.5210076414593183,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8875098824501038,
      "num_tokens": 433224410.0,
      "step": 12450
    },
    {
      "epoch": 2.312163416898793,
      "grad_norm": 1.6728571648922794,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8871981501579285,
      "num_tokens": 433254880.0,
      "step": 12451
    },
    {
      "epoch": 2.3123491179201485,
      "grad_norm": 1.52597069874801,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8940849900245667,
      "num_tokens": 433290737.0,
      "step": 12452
    },
    {
      "epoch": 2.3125348189415043,
      "grad_norm": 1.5770955081702565,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8922818899154663,
      "num_tokens": 433321700.0,
      "step": 12453
    },
    {
      "epoch": 2.31272051996286,
      "grad_norm": 1.5212345915801422,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8883501291275024,
      "num_tokens": 433357981.0,
      "step": 12454
    },
    {
      "epoch": 2.3129062209842153,
      "grad_norm": 1.552107844718583,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.890893816947937,
      "num_tokens": 433390821.0,
      "step": 12455
    },
    {
      "epoch": 2.313091922005571,
      "grad_norm": 1.4276778021092664,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.893114447593689,
      "num_tokens": 433429918.0,
      "step": 12456
    },
    {
      "epoch": 2.3132776230269267,
      "grad_norm": 1.6392522769928597,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8900182843208313,
      "num_tokens": 433461555.0,
      "step": 12457
    },
    {
      "epoch": 2.313463324048282,
      "grad_norm": 1.5330740985508615,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.887855052947998,
      "num_tokens": 433498299.0,
      "step": 12458
    },
    {
      "epoch": 2.3136490250696378,
      "grad_norm": 1.6082888198688017,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8944944739341736,
      "num_tokens": 433528499.0,
      "step": 12459
    },
    {
      "epoch": 2.3138347260909935,
      "grad_norm": 1.6470234130341788,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.88096684217453,
      "num_tokens": 433562052.0,
      "step": 12460
    },
    {
      "epoch": 2.3140204271123492,
      "grad_norm": 1.3372386835957393,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8942732214927673,
      "num_tokens": 433605472.0,
      "step": 12461
    },
    {
      "epoch": 2.314206128133705,
      "grad_norm": 1.5379573945003584,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8752546310424805,
      "num_tokens": 433642927.0,
      "step": 12462
    },
    {
      "epoch": 2.3143918291550603,
      "grad_norm": 1.5240755509472967,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8888295888900757,
      "num_tokens": 433678413.0,
      "step": 12463
    },
    {
      "epoch": 2.314577530176416,
      "grad_norm": 1.656330552355765,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8889286518096924,
      "num_tokens": 433710644.0,
      "step": 12464
    },
    {
      "epoch": 2.3147632311977717,
      "grad_norm": 1.5166585401676376,
      "learning_rate": 1e-06,
      "loss": 0.2934,
      "mean_token_accuracy": 0.8966206908226013,
      "num_tokens": 433745546.0,
      "step": 12465
    },
    {
      "epoch": 2.314948932219127,
      "grad_norm": 1.4713788068908926,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.9010193943977356,
      "num_tokens": 433780068.0,
      "step": 12466
    },
    {
      "epoch": 2.3151346332404827,
      "grad_norm": 1.7257847066442418,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8732637166976929,
      "num_tokens": 433809617.0,
      "step": 12467
    },
    {
      "epoch": 2.3153203342618385,
      "grad_norm": 1.585531183611798,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8775522112846375,
      "num_tokens": 433847261.0,
      "step": 12468
    },
    {
      "epoch": 2.315506035283194,
      "grad_norm": 1.4572331888316397,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8891252279281616,
      "num_tokens": 433886270.0,
      "step": 12469
    },
    {
      "epoch": 2.3156917363045495,
      "grad_norm": 1.5099787940831473,
      "learning_rate": 1e-06,
      "loss": 0.2804,
      "mean_token_accuracy": 0.9024131298065186,
      "num_tokens": 433918195.0,
      "step": 12470
    },
    {
      "epoch": 2.3158774373259052,
      "grad_norm": 1.54817918263992,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8937230110168457,
      "num_tokens": 433954858.0,
      "step": 12471
    },
    {
      "epoch": 2.316063138347261,
      "grad_norm": 1.5568164461337803,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.890089750289917,
      "num_tokens": 433985508.0,
      "step": 12472
    },
    {
      "epoch": 2.3162488393686167,
      "grad_norm": 1.7364862971299782,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8816369771957397,
      "num_tokens": 434015086.0,
      "step": 12473
    },
    {
      "epoch": 2.316434540389972,
      "grad_norm": 1.5713934135489906,
      "learning_rate": 1e-06,
      "loss": 0.2964,
      "mean_token_accuracy": 0.8962558507919312,
      "num_tokens": 434051222.0,
      "step": 12474
    },
    {
      "epoch": 2.3166202414113277,
      "grad_norm": 1.4532188101696524,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.892815113067627,
      "num_tokens": 434090728.0,
      "step": 12475
    },
    {
      "epoch": 2.3168059424326835,
      "grad_norm": 1.5848826460349759,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8954945802688599,
      "num_tokens": 434125382.0,
      "step": 12476
    },
    {
      "epoch": 2.316991643454039,
      "grad_norm": 1.8062457116273554,
      "learning_rate": 1e-06,
      "loss": 0.3563,
      "mean_token_accuracy": 0.8785003423690796,
      "num_tokens": 434153422.0,
      "step": 12477
    },
    {
      "epoch": 2.3171773444753945,
      "grad_norm": 1.7013386402702761,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8893131017684937,
      "num_tokens": 434187053.0,
      "step": 12478
    },
    {
      "epoch": 2.31736304549675,
      "grad_norm": 1.5113038451427723,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8772826194763184,
      "num_tokens": 434226592.0,
      "step": 12479
    },
    {
      "epoch": 2.317548746518106,
      "grad_norm": 1.6131252269305127,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.890355110168457,
      "num_tokens": 434259479.0,
      "step": 12480
    },
    {
      "epoch": 2.3177344475394612,
      "grad_norm": 1.52496642820251,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8750592470169067,
      "num_tokens": 434299137.0,
      "step": 12481
    },
    {
      "epoch": 2.317920148560817,
      "grad_norm": 1.5679214572657076,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8842809796333313,
      "num_tokens": 434335589.0,
      "step": 12482
    },
    {
      "epoch": 2.3181058495821727,
      "grad_norm": 1.6899127822378701,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.894970715045929,
      "num_tokens": 434366974.0,
      "step": 12483
    },
    {
      "epoch": 2.3182915506035284,
      "grad_norm": 1.7059064625908065,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8820574283599854,
      "num_tokens": 434400562.0,
      "step": 12484
    },
    {
      "epoch": 2.318477251624884,
      "grad_norm": 1.7397159933783855,
      "learning_rate": 1e-06,
      "loss": 0.3145,
      "mean_token_accuracy": 0.8874786496162415,
      "num_tokens": 434432208.0,
      "step": 12485
    },
    {
      "epoch": 2.3186629526462395,
      "grad_norm": 1.6699891035366292,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8933825492858887,
      "num_tokens": 434460650.0,
      "step": 12486
    },
    {
      "epoch": 2.318848653667595,
      "grad_norm": 1.55910501164878,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8861775994300842,
      "num_tokens": 434494322.0,
      "step": 12487
    },
    {
      "epoch": 2.319034354688951,
      "grad_norm": 1.6845922159278532,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8918229937553406,
      "num_tokens": 434526154.0,
      "step": 12488
    },
    {
      "epoch": 2.319220055710306,
      "grad_norm": 1.7616279396185175,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8897930383682251,
      "num_tokens": 434554922.0,
      "step": 12489
    },
    {
      "epoch": 2.319405756731662,
      "grad_norm": 1.5755335099019039,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8941466808319092,
      "num_tokens": 434587941.0,
      "step": 12490
    },
    {
      "epoch": 2.3195914577530177,
      "grad_norm": 1.7568348233232312,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8874915838241577,
      "num_tokens": 434621426.0,
      "step": 12491
    },
    {
      "epoch": 2.3197771587743734,
      "grad_norm": 1.5286656097270943,
      "learning_rate": 1e-06,
      "loss": 0.288,
      "mean_token_accuracy": 0.8996967077255249,
      "num_tokens": 434652340.0,
      "step": 12492
    },
    {
      "epoch": 2.3199628597957287,
      "grad_norm": 1.6315150503544071,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8874369859695435,
      "num_tokens": 434684174.0,
      "step": 12493
    },
    {
      "epoch": 2.3201485608170844,
      "grad_norm": 1.4369604691664473,
      "learning_rate": 1e-06,
      "loss": 0.2787,
      "mean_token_accuracy": 0.9040627479553223,
      "num_tokens": 434721971.0,
      "step": 12494
    },
    {
      "epoch": 2.32033426183844,
      "grad_norm": 1.5272051490784524,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8807013034820557,
      "num_tokens": 434759175.0,
      "step": 12495
    },
    {
      "epoch": 2.320519962859796,
      "grad_norm": 1.490167355983472,
      "learning_rate": 1e-06,
      "loss": 0.2916,
      "mean_token_accuracy": 0.8934670686721802,
      "num_tokens": 434797090.0,
      "step": 12496
    },
    {
      "epoch": 2.320705663881151,
      "grad_norm": 1.524278108009253,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8863732814788818,
      "num_tokens": 434836932.0,
      "step": 12497
    },
    {
      "epoch": 2.320891364902507,
      "grad_norm": 1.5209750205970722,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8897276520729065,
      "num_tokens": 434876009.0,
      "step": 12498
    },
    {
      "epoch": 2.3210770659238626,
      "grad_norm": 1.5220854372137236,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8872900009155273,
      "num_tokens": 434912488.0,
      "step": 12499
    },
    {
      "epoch": 2.3212627669452184,
      "grad_norm": 1.5748985916103788,
      "learning_rate": 1e-06,
      "loss": 0.2694,
      "mean_token_accuracy": 0.9032082557678223,
      "num_tokens": 434944494.0,
      "step": 12500
    },
    {
      "epoch": 2.3214484679665737,
      "grad_norm": 1.4086717006878913,
      "learning_rate": 1e-06,
      "loss": 0.2861,
      "mean_token_accuracy": 0.8978350162506104,
      "num_tokens": 434984063.0,
      "step": 12501
    },
    {
      "epoch": 2.3216341689879294,
      "grad_norm": 1.5243721678652178,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.885620653629303,
      "num_tokens": 435026245.0,
      "step": 12502
    },
    {
      "epoch": 2.321819870009285,
      "grad_norm": 1.5580047457735986,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8795457482337952,
      "num_tokens": 435066497.0,
      "step": 12503
    },
    {
      "epoch": 2.3220055710306404,
      "grad_norm": 1.63760492368333,
      "learning_rate": 1e-06,
      "loss": 0.3035,
      "mean_token_accuracy": 0.8932480812072754,
      "num_tokens": 435099146.0,
      "step": 12504
    },
    {
      "epoch": 2.322191272051996,
      "grad_norm": 1.4831184761314067,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8922678828239441,
      "num_tokens": 435138902.0,
      "step": 12505
    },
    {
      "epoch": 2.322376973073352,
      "grad_norm": 1.7351673839391184,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8831582069396973,
      "num_tokens": 435169432.0,
      "step": 12506
    },
    {
      "epoch": 2.3225626740947076,
      "grad_norm": 1.4348017903066892,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8947343826293945,
      "num_tokens": 435206722.0,
      "step": 12507
    },
    {
      "epoch": 2.3227483751160634,
      "grad_norm": 1.5162941430735013,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8853530287742615,
      "num_tokens": 435241473.0,
      "step": 12508
    },
    {
      "epoch": 2.3229340761374186,
      "grad_norm": 1.6308803418665052,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.8937363624572754,
      "num_tokens": 435270827.0,
      "step": 12509
    },
    {
      "epoch": 2.3231197771587744,
      "grad_norm": 1.5062057961743154,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8972395658493042,
      "num_tokens": 435303910.0,
      "step": 12510
    },
    {
      "epoch": 2.32330547818013,
      "grad_norm": 1.690720501594749,
      "learning_rate": 1e-06,
      "loss": 0.2973,
      "mean_token_accuracy": 0.8916839957237244,
      "num_tokens": 435336326.0,
      "step": 12511
    },
    {
      "epoch": 2.3234911792014854,
      "grad_norm": 1.541089591227348,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8845207691192627,
      "num_tokens": 435374464.0,
      "step": 12512
    },
    {
      "epoch": 2.323676880222841,
      "grad_norm": 1.5096839934563586,
      "learning_rate": 1e-06,
      "loss": 0.3205,
      "mean_token_accuracy": 0.8882594704627991,
      "num_tokens": 435411389.0,
      "step": 12513
    },
    {
      "epoch": 2.323862581244197,
      "grad_norm": 1.5371137516385456,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8880609273910522,
      "num_tokens": 435447785.0,
      "step": 12514
    },
    {
      "epoch": 2.3240482822655526,
      "grad_norm": 1.692544491984756,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8795506954193115,
      "num_tokens": 435482686.0,
      "step": 12515
    },
    {
      "epoch": 2.324233983286908,
      "grad_norm": 1.5359788283015752,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8856534957885742,
      "num_tokens": 435521528.0,
      "step": 12516
    },
    {
      "epoch": 2.3244196843082636,
      "grad_norm": 1.5659944493009803,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8964920043945312,
      "num_tokens": 435556335.0,
      "step": 12517
    },
    {
      "epoch": 2.3246053853296194,
      "grad_norm": 1.6826681743700578,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.873684823513031,
      "num_tokens": 435592276.0,
      "step": 12518
    },
    {
      "epoch": 2.324791086350975,
      "grad_norm": 1.4849295187026381,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8923245668411255,
      "num_tokens": 435628182.0,
      "step": 12519
    },
    {
      "epoch": 2.3249767873723304,
      "grad_norm": 1.6954903125571206,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8913372755050659,
      "num_tokens": 435657547.0,
      "step": 12520
    },
    {
      "epoch": 2.325162488393686,
      "grad_norm": 1.445624070347903,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8879101872444153,
      "num_tokens": 435700025.0,
      "step": 12521
    },
    {
      "epoch": 2.325348189415042,
      "grad_norm": 1.5501923217400453,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.877957820892334,
      "num_tokens": 435737266.0,
      "step": 12522
    },
    {
      "epoch": 2.3255338904363976,
      "grad_norm": 1.6845662461924524,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8895418643951416,
      "num_tokens": 435766778.0,
      "step": 12523
    },
    {
      "epoch": 2.325719591457753,
      "grad_norm": 1.5345518632482322,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8792999982833862,
      "num_tokens": 435802916.0,
      "step": 12524
    },
    {
      "epoch": 2.3259052924791086,
      "grad_norm": 1.3708346181821902,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8815898895263672,
      "num_tokens": 435846334.0,
      "step": 12525
    },
    {
      "epoch": 2.3260909935004643,
      "grad_norm": 1.608260884454223,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8857611417770386,
      "num_tokens": 435877315.0,
      "step": 12526
    },
    {
      "epoch": 2.3262766945218196,
      "grad_norm": 1.4831131308083343,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8945617079734802,
      "num_tokens": 435913249.0,
      "step": 12527
    },
    {
      "epoch": 2.3264623955431754,
      "grad_norm": 1.6378486906412286,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.88172847032547,
      "num_tokens": 435944884.0,
      "step": 12528
    },
    {
      "epoch": 2.326648096564531,
      "grad_norm": 1.4852745791214879,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8890548944473267,
      "num_tokens": 435985331.0,
      "step": 12529
    },
    {
      "epoch": 2.326833797585887,
      "grad_norm": 1.7416529194126797,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.877829372882843,
      "num_tokens": 436018781.0,
      "step": 12530
    },
    {
      "epoch": 2.3270194986072426,
      "grad_norm": 1.7540693101067342,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8822466731071472,
      "num_tokens": 436046171.0,
      "step": 12531
    },
    {
      "epoch": 2.327205199628598,
      "grad_norm": 1.6077972337906896,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8741188049316406,
      "num_tokens": 436084189.0,
      "step": 12532
    },
    {
      "epoch": 2.3273909006499536,
      "grad_norm": 1.41360922796248,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8936825394630432,
      "num_tokens": 436127622.0,
      "step": 12533
    },
    {
      "epoch": 2.3275766016713093,
      "grad_norm": 1.5237189988770505,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8908838629722595,
      "num_tokens": 436166102.0,
      "step": 12534
    },
    {
      "epoch": 2.3277623026926646,
      "grad_norm": 1.758361837161511,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8832507133483887,
      "num_tokens": 436197531.0,
      "step": 12535
    },
    {
      "epoch": 2.3279480037140203,
      "grad_norm": 1.6682707221463233,
      "learning_rate": 1e-06,
      "loss": 0.2711,
      "mean_token_accuracy": 0.9036338925361633,
      "num_tokens": 436225296.0,
      "step": 12536
    },
    {
      "epoch": 2.328133704735376,
      "grad_norm": 1.5472719694655803,
      "learning_rate": 1e-06,
      "loss": 0.2984,
      "mean_token_accuracy": 0.8930006623268127,
      "num_tokens": 436258442.0,
      "step": 12537
    },
    {
      "epoch": 2.328319405756732,
      "grad_norm": 1.7589389277790957,
      "learning_rate": 1e-06,
      "loss": 0.2804,
      "mean_token_accuracy": 0.9011706113815308,
      "num_tokens": 436290400.0,
      "step": 12538
    },
    {
      "epoch": 2.328505106778087,
      "grad_norm": 1.4739367237148513,
      "learning_rate": 1e-06,
      "loss": 0.2862,
      "mean_token_accuracy": 0.8972648978233337,
      "num_tokens": 436328354.0,
      "step": 12539
    },
    {
      "epoch": 2.328690807799443,
      "grad_norm": 1.6108456502901918,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8932241797447205,
      "num_tokens": 436362588.0,
      "step": 12540
    },
    {
      "epoch": 2.3288765088207986,
      "grad_norm": 1.4994082530822002,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8866217136383057,
      "num_tokens": 436402953.0,
      "step": 12541
    },
    {
      "epoch": 2.3290622098421543,
      "grad_norm": 1.6405479413698245,
      "learning_rate": 1e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8761661052703857,
      "num_tokens": 436433994.0,
      "step": 12542
    },
    {
      "epoch": 2.3292479108635096,
      "grad_norm": 1.5054958364830804,
      "learning_rate": 1e-06,
      "loss": 0.2858,
      "mean_token_accuracy": 0.8980263471603394,
      "num_tokens": 436466618.0,
      "step": 12543
    },
    {
      "epoch": 2.3294336118848653,
      "grad_norm": 1.6734817920948277,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.881617546081543,
      "num_tokens": 436499027.0,
      "step": 12544
    },
    {
      "epoch": 2.329619312906221,
      "grad_norm": 1.463926159802149,
      "learning_rate": 1e-06,
      "loss": 0.2868,
      "mean_token_accuracy": 0.8968799114227295,
      "num_tokens": 436537239.0,
      "step": 12545
    },
    {
      "epoch": 2.3298050139275768,
      "grad_norm": 1.4824300957459855,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8861461877822876,
      "num_tokens": 436574617.0,
      "step": 12546
    },
    {
      "epoch": 2.329990714948932,
      "grad_norm": 1.5943111674666226,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8878343105316162,
      "num_tokens": 436607425.0,
      "step": 12547
    },
    {
      "epoch": 2.330176415970288,
      "grad_norm": 1.5821903356024434,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8809075355529785,
      "num_tokens": 436642080.0,
      "step": 12548
    },
    {
      "epoch": 2.3303621169916435,
      "grad_norm": 1.4539314681742834,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8904502391815186,
      "num_tokens": 436679014.0,
      "step": 12549
    },
    {
      "epoch": 2.3305478180129993,
      "grad_norm": 1.6699236488469456,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8957152962684631,
      "num_tokens": 436712486.0,
      "step": 12550
    },
    {
      "epoch": 2.3307335190343546,
      "grad_norm": 1.7150532475389442,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.874118983745575,
      "num_tokens": 436747365.0,
      "step": 12551
    },
    {
      "epoch": 2.3309192200557103,
      "grad_norm": 1.7266115032181868,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8817049264907837,
      "num_tokens": 436782503.0,
      "step": 12552
    },
    {
      "epoch": 2.331104921077066,
      "grad_norm": 1.5332757141483033,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.883148193359375,
      "num_tokens": 436820791.0,
      "step": 12553
    },
    {
      "epoch": 2.3312906220984217,
      "grad_norm": 1.618212596705645,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8817034959793091,
      "num_tokens": 436853629.0,
      "step": 12554
    },
    {
      "epoch": 2.331476323119777,
      "grad_norm": 1.562769057338548,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8892225623130798,
      "num_tokens": 436887272.0,
      "step": 12555
    },
    {
      "epoch": 2.3316620241411328,
      "grad_norm": 1.6575002933377214,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8913081884384155,
      "num_tokens": 436920587.0,
      "step": 12556
    },
    {
      "epoch": 2.3318477251624885,
      "grad_norm": 1.5672703243481636,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.8908419609069824,
      "num_tokens": 436956741.0,
      "step": 12557
    },
    {
      "epoch": 2.332033426183844,
      "grad_norm": 1.7379236510854996,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.876797616481781,
      "num_tokens": 436988026.0,
      "step": 12558
    },
    {
      "epoch": 2.3322191272051995,
      "grad_norm": 1.5598665851942646,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8795492053031921,
      "num_tokens": 437023545.0,
      "step": 12559
    },
    {
      "epoch": 2.3324048282265553,
      "grad_norm": 1.5793139534413831,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8930743932723999,
      "num_tokens": 437056159.0,
      "step": 12560
    },
    {
      "epoch": 2.332590529247911,
      "grad_norm": 1.6222784759840019,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8939582705497742,
      "num_tokens": 437088048.0,
      "step": 12561
    },
    {
      "epoch": 2.3327762302692667,
      "grad_norm": 1.5925150339811702,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8774400949478149,
      "num_tokens": 437126361.0,
      "step": 12562
    },
    {
      "epoch": 2.332961931290622,
      "grad_norm": 1.7010782981533148,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8900128602981567,
      "num_tokens": 437155724.0,
      "step": 12563
    },
    {
      "epoch": 2.3331476323119777,
      "grad_norm": 1.5093480586669556,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8955312967300415,
      "num_tokens": 437192730.0,
      "step": 12564
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 1.497571513858365,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8882834315299988,
      "num_tokens": 437232693.0,
      "step": 12565
    },
    {
      "epoch": 2.3335190343546888,
      "grad_norm": 1.621371652953503,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8862128257751465,
      "num_tokens": 437263377.0,
      "step": 12566
    },
    {
      "epoch": 2.3337047353760445,
      "grad_norm": 1.4229114169949957,
      "learning_rate": 1e-06,
      "loss": 0.2903,
      "mean_token_accuracy": 0.8961464166641235,
      "num_tokens": 437302537.0,
      "step": 12567
    },
    {
      "epoch": 2.3338904363974002,
      "grad_norm": 1.6733124418746101,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8946985006332397,
      "num_tokens": 437329719.0,
      "step": 12568
    },
    {
      "epoch": 2.334076137418756,
      "grad_norm": 1.6314977668646597,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.885697603225708,
      "num_tokens": 437365400.0,
      "step": 12569
    },
    {
      "epoch": 2.3342618384401113,
      "grad_norm": 1.547988999318584,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.894335150718689,
      "num_tokens": 437399905.0,
      "step": 12570
    },
    {
      "epoch": 2.334447539461467,
      "grad_norm": 1.6540182664144674,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8842208981513977,
      "num_tokens": 437432760.0,
      "step": 12571
    },
    {
      "epoch": 2.3346332404828227,
      "grad_norm": 1.5755570978762967,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8891406059265137,
      "num_tokens": 437466930.0,
      "step": 12572
    },
    {
      "epoch": 2.3348189415041785,
      "grad_norm": 1.9277180331599773,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8837597966194153,
      "num_tokens": 437490130.0,
      "step": 12573
    },
    {
      "epoch": 2.3350046425255337,
      "grad_norm": 1.550594000346753,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8889772891998291,
      "num_tokens": 437523425.0,
      "step": 12574
    },
    {
      "epoch": 2.3351903435468895,
      "grad_norm": 1.5323526801577736,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8832849264144897,
      "num_tokens": 437560285.0,
      "step": 12575
    },
    {
      "epoch": 2.335376044568245,
      "grad_norm": 1.5176070116282918,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8914657831192017,
      "num_tokens": 437597230.0,
      "step": 12576
    },
    {
      "epoch": 2.335561745589601,
      "grad_norm": 1.583808841180886,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.890967607498169,
      "num_tokens": 437629653.0,
      "step": 12577
    },
    {
      "epoch": 2.3357474466109562,
      "grad_norm": 1.6154297379525007,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.882575273513794,
      "num_tokens": 437665537.0,
      "step": 12578
    },
    {
      "epoch": 2.335933147632312,
      "grad_norm": 1.494710499536805,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.876572847366333,
      "num_tokens": 437705079.0,
      "step": 12579
    },
    {
      "epoch": 2.3361188486536677,
      "grad_norm": 1.6431865515653268,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8771193027496338,
      "num_tokens": 437740488.0,
      "step": 12580
    },
    {
      "epoch": 2.336304549675023,
      "grad_norm": 1.5246043440806702,
      "learning_rate": 1e-06,
      "loss": 0.2916,
      "mean_token_accuracy": 0.8933175802230835,
      "num_tokens": 437778511.0,
      "step": 12581
    },
    {
      "epoch": 2.3364902506963787,
      "grad_norm": 1.6582033224009813,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.887274980545044,
      "num_tokens": 437809756.0,
      "step": 12582
    },
    {
      "epoch": 2.3366759517177345,
      "grad_norm": 1.7541277364373966,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8809660077095032,
      "num_tokens": 437836207.0,
      "step": 12583
    },
    {
      "epoch": 2.33686165273909,
      "grad_norm": 1.5360097228802703,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8924872279167175,
      "num_tokens": 437871671.0,
      "step": 12584
    },
    {
      "epoch": 2.337047353760446,
      "grad_norm": 1.5815840389624485,
      "learning_rate": 1e-06,
      "loss": 0.287,
      "mean_token_accuracy": 0.8953440189361572,
      "num_tokens": 437904022.0,
      "step": 12585
    },
    {
      "epoch": 2.337233054781801,
      "grad_norm": 1.6427485905401982,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8853353261947632,
      "num_tokens": 437937295.0,
      "step": 12586
    },
    {
      "epoch": 2.337418755803157,
      "grad_norm": 1.519121013040696,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8971248865127563,
      "num_tokens": 437974340.0,
      "step": 12587
    },
    {
      "epoch": 2.3376044568245127,
      "grad_norm": 1.5431066280281802,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8748523592948914,
      "num_tokens": 438012083.0,
      "step": 12588
    },
    {
      "epoch": 2.337790157845868,
      "grad_norm": 1.8598035363373189,
      "learning_rate": 1e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8689454197883606,
      "num_tokens": 438040723.0,
      "step": 12589
    },
    {
      "epoch": 2.3379758588672237,
      "grad_norm": 1.54926675442265,
      "learning_rate": 1e-06,
      "loss": 0.2902,
      "mean_token_accuracy": 0.8979060649871826,
      "num_tokens": 438073438.0,
      "step": 12590
    },
    {
      "epoch": 2.3381615598885794,
      "grad_norm": 1.606795038338468,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8734663724899292,
      "num_tokens": 438109455.0,
      "step": 12591
    },
    {
      "epoch": 2.338347260909935,
      "grad_norm": 1.4520495278577767,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8853275179862976,
      "num_tokens": 438147854.0,
      "step": 12592
    },
    {
      "epoch": 2.3385329619312905,
      "grad_norm": 1.457276097720976,
      "learning_rate": 1e-06,
      "loss": 0.2954,
      "mean_token_accuracy": 0.895704984664917,
      "num_tokens": 438182776.0,
      "step": 12593
    },
    {
      "epoch": 2.338718662952646,
      "grad_norm": 1.4850112991038729,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8753225803375244,
      "num_tokens": 438223248.0,
      "step": 12594
    },
    {
      "epoch": 2.338904363974002,
      "grad_norm": 1.5943996885690728,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8868238925933838,
      "num_tokens": 438256492.0,
      "step": 12595
    },
    {
      "epoch": 2.3390900649953577,
      "grad_norm": 1.5858004820783769,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8901724815368652,
      "num_tokens": 438291285.0,
      "step": 12596
    },
    {
      "epoch": 2.339275766016713,
      "grad_norm": 1.5578022587697373,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8813902735710144,
      "num_tokens": 438330468.0,
      "step": 12597
    },
    {
      "epoch": 2.3394614670380687,
      "grad_norm": 1.5564962828610018,
      "learning_rate": 1e-06,
      "loss": 0.3045,
      "mean_token_accuracy": 0.8937835693359375,
      "num_tokens": 438366734.0,
      "step": 12598
    },
    {
      "epoch": 2.3396471680594244,
      "grad_norm": 1.6483461266983803,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8958669900894165,
      "num_tokens": 438397145.0,
      "step": 12599
    },
    {
      "epoch": 2.33983286908078,
      "grad_norm": 1.5053292032242813,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8902117013931274,
      "num_tokens": 438436012.0,
      "step": 12600
    },
    {
      "epoch": 2.3400185701021354,
      "grad_norm": 1.614949914521044,
      "learning_rate": 1e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8697616457939148,
      "num_tokens": 438468294.0,
      "step": 12601
    },
    {
      "epoch": 2.340204271123491,
      "grad_norm": 1.6097685455557749,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8810359239578247,
      "num_tokens": 438502521.0,
      "step": 12602
    },
    {
      "epoch": 2.340389972144847,
      "grad_norm": 1.7862766532942242,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.8931604623794556,
      "num_tokens": 438526547.0,
      "step": 12603
    },
    {
      "epoch": 2.340575673166202,
      "grad_norm": 1.5683577317803916,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8829560875892639,
      "num_tokens": 438561014.0,
      "step": 12604
    },
    {
      "epoch": 2.340761374187558,
      "grad_norm": 1.5726314871074794,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8748348355293274,
      "num_tokens": 438598092.0,
      "step": 12605
    },
    {
      "epoch": 2.3409470752089137,
      "grad_norm": 1.5223534162206955,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8844885230064392,
      "num_tokens": 438633355.0,
      "step": 12606
    },
    {
      "epoch": 2.3411327762302694,
      "grad_norm": 1.614244999518431,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8843473196029663,
      "num_tokens": 438665597.0,
      "step": 12607
    },
    {
      "epoch": 2.341318477251625,
      "grad_norm": 1.5409349016519391,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8826740980148315,
      "num_tokens": 438704882.0,
      "step": 12608
    },
    {
      "epoch": 2.3415041782729804,
      "grad_norm": 1.5971573324662143,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8732510209083557,
      "num_tokens": 438740401.0,
      "step": 12609
    },
    {
      "epoch": 2.341689879294336,
      "grad_norm": 1.5118541582063276,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8917969465255737,
      "num_tokens": 438775375.0,
      "step": 12610
    },
    {
      "epoch": 2.341875580315692,
      "grad_norm": 1.491750779480472,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8830088376998901,
      "num_tokens": 438811123.0,
      "step": 12611
    },
    {
      "epoch": 2.342061281337047,
      "grad_norm": 1.472059516823774,
      "learning_rate": 1e-06,
      "loss": 0.2775,
      "mean_token_accuracy": 0.9032261371612549,
      "num_tokens": 438846759.0,
      "step": 12612
    },
    {
      "epoch": 2.342246982358403,
      "grad_norm": 1.5806612495533081,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.881386399269104,
      "num_tokens": 438885227.0,
      "step": 12613
    },
    {
      "epoch": 2.3424326833797586,
      "grad_norm": 1.910725070945846,
      "learning_rate": 1e-06,
      "loss": 0.2963,
      "mean_token_accuracy": 0.8942254781723022,
      "num_tokens": 438920705.0,
      "step": 12614
    },
    {
      "epoch": 2.3426183844011144,
      "grad_norm": 1.5316836552403605,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.894169807434082,
      "num_tokens": 438953301.0,
      "step": 12615
    },
    {
      "epoch": 2.3428040854224697,
      "grad_norm": 1.4461519680988757,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.8975368142127991,
      "num_tokens": 438990882.0,
      "step": 12616
    },
    {
      "epoch": 2.3429897864438254,
      "grad_norm": 1.4930285080715568,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8923463225364685,
      "num_tokens": 439028440.0,
      "step": 12617
    },
    {
      "epoch": 2.343175487465181,
      "grad_norm": 1.5756329664754254,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8904476165771484,
      "num_tokens": 439061569.0,
      "step": 12618
    },
    {
      "epoch": 2.343361188486537,
      "grad_norm": 1.581307720534236,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8810634016990662,
      "num_tokens": 439094740.0,
      "step": 12619
    },
    {
      "epoch": 2.343546889507892,
      "grad_norm": 1.621621501848817,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8886150121688843,
      "num_tokens": 439128139.0,
      "step": 12620
    },
    {
      "epoch": 2.343732590529248,
      "grad_norm": 1.626767730088471,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8915634155273438,
      "num_tokens": 439162056.0,
      "step": 12621
    },
    {
      "epoch": 2.3439182915506036,
      "grad_norm": 1.4220233411895733,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8868982791900635,
      "num_tokens": 439201437.0,
      "step": 12622
    },
    {
      "epoch": 2.3441039925719593,
      "grad_norm": 1.491359323889145,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8925650715827942,
      "num_tokens": 439240581.0,
      "step": 12623
    },
    {
      "epoch": 2.3442896935933146,
      "grad_norm": 1.6988710397488305,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8834226131439209,
      "num_tokens": 439273153.0,
      "step": 12624
    },
    {
      "epoch": 2.3444753946146704,
      "grad_norm": 1.544835252767626,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8756486177444458,
      "num_tokens": 439311825.0,
      "step": 12625
    },
    {
      "epoch": 2.344661095636026,
      "grad_norm": 1.6741152923277116,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.882404625415802,
      "num_tokens": 439343571.0,
      "step": 12626
    },
    {
      "epoch": 2.3448467966573814,
      "grad_norm": 1.7067626797548656,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8887182474136353,
      "num_tokens": 439377778.0,
      "step": 12627
    },
    {
      "epoch": 2.345032497678737,
      "grad_norm": 1.5483543070482026,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.8982725739479065,
      "num_tokens": 439407529.0,
      "step": 12628
    },
    {
      "epoch": 2.345218198700093,
      "grad_norm": 1.5865003625679324,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8921046257019043,
      "num_tokens": 439439637.0,
      "step": 12629
    },
    {
      "epoch": 2.3454038997214486,
      "grad_norm": 1.6697589959358987,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8788906335830688,
      "num_tokens": 439472994.0,
      "step": 12630
    },
    {
      "epoch": 2.3455896007428043,
      "grad_norm": 1.6172321803898986,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.885282039642334,
      "num_tokens": 439507574.0,
      "step": 12631
    },
    {
      "epoch": 2.3457753017641596,
      "grad_norm": 1.6113005903361763,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8913954496383667,
      "num_tokens": 439538003.0,
      "step": 12632
    },
    {
      "epoch": 2.3459610027855153,
      "grad_norm": 1.5972718707494478,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8827592134475708,
      "num_tokens": 439571866.0,
      "step": 12633
    },
    {
      "epoch": 2.346146703806871,
      "grad_norm": 1.3636144794547844,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8853164911270142,
      "num_tokens": 439615034.0,
      "step": 12634
    },
    {
      "epoch": 2.3463324048282264,
      "grad_norm": 1.7137503998024248,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8743880391120911,
      "num_tokens": 439647608.0,
      "step": 12635
    },
    {
      "epoch": 2.346518105849582,
      "grad_norm": 1.7054107611255853,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8834789991378784,
      "num_tokens": 439677332.0,
      "step": 12636
    },
    {
      "epoch": 2.346703806870938,
      "grad_norm": 1.6215504136514711,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8838996887207031,
      "num_tokens": 439708031.0,
      "step": 12637
    },
    {
      "epoch": 2.3468895078922936,
      "grad_norm": 1.6679335156720696,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8905410766601562,
      "num_tokens": 439737574.0,
      "step": 12638
    },
    {
      "epoch": 2.347075208913649,
      "grad_norm": 1.5359611240829727,
      "learning_rate": 1e-06,
      "loss": 0.2823,
      "mean_token_accuracy": 0.8992702960968018,
      "num_tokens": 439771289.0,
      "step": 12639
    },
    {
      "epoch": 2.3472609099350046,
      "grad_norm": 1.5665856996883276,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8898072242736816,
      "num_tokens": 439807555.0,
      "step": 12640
    },
    {
      "epoch": 2.3474466109563603,
      "grad_norm": 1.7022365273498103,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.878191351890564,
      "num_tokens": 439840243.0,
      "step": 12641
    },
    {
      "epoch": 2.347632311977716,
      "grad_norm": 1.6356310949690067,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8833516836166382,
      "num_tokens": 439874268.0,
      "step": 12642
    },
    {
      "epoch": 2.3478180129990713,
      "grad_norm": 1.5531095581568377,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8900748491287231,
      "num_tokens": 439906318.0,
      "step": 12643
    },
    {
      "epoch": 2.348003714020427,
      "grad_norm": 1.7689477156150204,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8825125098228455,
      "num_tokens": 439942727.0,
      "step": 12644
    },
    {
      "epoch": 2.348189415041783,
      "grad_norm": 1.6437208705717625,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8791577816009521,
      "num_tokens": 439975418.0,
      "step": 12645
    },
    {
      "epoch": 2.3483751160631385,
      "grad_norm": 1.422648239039715,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8890405893325806,
      "num_tokens": 440016802.0,
      "step": 12646
    },
    {
      "epoch": 2.348560817084494,
      "grad_norm": 1.6823881810085453,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8734219074249268,
      "num_tokens": 440046407.0,
      "step": 12647
    },
    {
      "epoch": 2.3487465181058496,
      "grad_norm": 1.580397770503369,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8821606636047363,
      "num_tokens": 440080792.0,
      "step": 12648
    },
    {
      "epoch": 2.3489322191272053,
      "grad_norm": 1.670553745641084,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8927367925643921,
      "num_tokens": 440107537.0,
      "step": 12649
    },
    {
      "epoch": 2.3491179201485606,
      "grad_norm": 1.9120020554771457,
      "learning_rate": 1e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8708370923995972,
      "num_tokens": 440134641.0,
      "step": 12650
    },
    {
      "epoch": 2.3493036211699163,
      "grad_norm": 1.6890063829863062,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8894073963165283,
      "num_tokens": 440162542.0,
      "step": 12651
    },
    {
      "epoch": 2.349489322191272,
      "grad_norm": 1.4334452751541942,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8931427597999573,
      "num_tokens": 440199688.0,
      "step": 12652
    },
    {
      "epoch": 2.349675023212628,
      "grad_norm": 1.465979079153368,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8966662883758545,
      "num_tokens": 440235675.0,
      "step": 12653
    },
    {
      "epoch": 2.3498607242339835,
      "grad_norm": 1.7248515803128837,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.8872538805007935,
      "num_tokens": 440271581.0,
      "step": 12654
    },
    {
      "epoch": 2.350046425255339,
      "grad_norm": 1.5995142211357603,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8798934817314148,
      "num_tokens": 440305811.0,
      "step": 12655
    },
    {
      "epoch": 2.3502321262766945,
      "grad_norm": 1.4876748639887865,
      "learning_rate": 1e-06,
      "loss": 0.2696,
      "mean_token_accuracy": 0.900659441947937,
      "num_tokens": 440338064.0,
      "step": 12656
    },
    {
      "epoch": 2.3504178272980503,
      "grad_norm": 1.552631448503422,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8946394324302673,
      "num_tokens": 440374669.0,
      "step": 12657
    },
    {
      "epoch": 2.3506035283194056,
      "grad_norm": 1.387209273974769,
      "learning_rate": 1e-06,
      "loss": 0.2724,
      "mean_token_accuracy": 0.9025610685348511,
      "num_tokens": 440412574.0,
      "step": 12658
    },
    {
      "epoch": 2.3507892293407613,
      "grad_norm": 1.619724804565014,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.882917046546936,
      "num_tokens": 440446242.0,
      "step": 12659
    },
    {
      "epoch": 2.350974930362117,
      "grad_norm": 1.5241635301552197,
      "learning_rate": 1e-06,
      "loss": 0.2616,
      "mean_token_accuracy": 0.9054927825927734,
      "num_tokens": 440478837.0,
      "step": 12660
    },
    {
      "epoch": 2.3511606313834728,
      "grad_norm": 1.4586389180085686,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8913350701332092,
      "num_tokens": 440518880.0,
      "step": 12661
    },
    {
      "epoch": 2.351346332404828,
      "grad_norm": 1.4266335288056917,
      "learning_rate": 1e-06,
      "loss": 0.2788,
      "mean_token_accuracy": 0.8968080282211304,
      "num_tokens": 440556543.0,
      "step": 12662
    },
    {
      "epoch": 2.3515320334261838,
      "grad_norm": 1.5713515540335927,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8915428519248962,
      "num_tokens": 440591070.0,
      "step": 12663
    },
    {
      "epoch": 2.3517177344475395,
      "grad_norm": 1.508799514539108,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8855607509613037,
      "num_tokens": 440627030.0,
      "step": 12664
    },
    {
      "epoch": 2.3519034354688952,
      "grad_norm": 1.6997689497704216,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8892261385917664,
      "num_tokens": 440661862.0,
      "step": 12665
    },
    {
      "epoch": 2.3520891364902505,
      "grad_norm": 1.6172105775542374,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8834460973739624,
      "num_tokens": 440697600.0,
      "step": 12666
    },
    {
      "epoch": 2.3522748375116063,
      "grad_norm": 1.633156209726828,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8903571963310242,
      "num_tokens": 440729185.0,
      "step": 12667
    },
    {
      "epoch": 2.352460538532962,
      "grad_norm": 1.5860080243804349,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8878774642944336,
      "num_tokens": 440764594.0,
      "step": 12668
    },
    {
      "epoch": 2.3526462395543177,
      "grad_norm": 1.6646139899611505,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8738225698471069,
      "num_tokens": 440798505.0,
      "step": 12669
    },
    {
      "epoch": 2.352831940575673,
      "grad_norm": 1.6000086885339784,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8813362121582031,
      "num_tokens": 440832428.0,
      "step": 12670
    },
    {
      "epoch": 2.3530176415970288,
      "grad_norm": 1.46522469197126,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8910666108131409,
      "num_tokens": 440872068.0,
      "step": 12671
    },
    {
      "epoch": 2.3532033426183845,
      "grad_norm": 1.5924987201048866,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8912250399589539,
      "num_tokens": 440901049.0,
      "step": 12672
    },
    {
      "epoch": 2.3533890436397398,
      "grad_norm": 1.568520303477912,
      "learning_rate": 1e-06,
      "loss": 0.2981,
      "mean_token_accuracy": 0.8935328722000122,
      "num_tokens": 440934315.0,
      "step": 12673
    },
    {
      "epoch": 2.3535747446610955,
      "grad_norm": 1.6227238872117502,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8755346536636353,
      "num_tokens": 440968822.0,
      "step": 12674
    },
    {
      "epoch": 2.3537604456824512,
      "grad_norm": 1.672494404172226,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8874406814575195,
      "num_tokens": 440999606.0,
      "step": 12675
    },
    {
      "epoch": 2.353946146703807,
      "grad_norm": 1.6247907621062412,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8797342777252197,
      "num_tokens": 441032287.0,
      "step": 12676
    },
    {
      "epoch": 2.3541318477251627,
      "grad_norm": 1.6284225056612909,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8775584697723389,
      "num_tokens": 441066923.0,
      "step": 12677
    },
    {
      "epoch": 2.354317548746518,
      "grad_norm": 1.5027587498599388,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8776678442955017,
      "num_tokens": 441101058.0,
      "step": 12678
    },
    {
      "epoch": 2.3545032497678737,
      "grad_norm": 1.5645109257770062,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8875289559364319,
      "num_tokens": 441134452.0,
      "step": 12679
    },
    {
      "epoch": 2.3546889507892295,
      "grad_norm": 1.7132620709421782,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.883897602558136,
      "num_tokens": 441165573.0,
      "step": 12680
    },
    {
      "epoch": 2.3548746518105848,
      "grad_norm": 1.597896599269497,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8803331255912781,
      "num_tokens": 441196989.0,
      "step": 12681
    },
    {
      "epoch": 2.3550603528319405,
      "grad_norm": 1.5750637783071846,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8945073485374451,
      "num_tokens": 441229911.0,
      "step": 12682
    },
    {
      "epoch": 2.355246053853296,
      "grad_norm": 1.605909786687276,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8875384330749512,
      "num_tokens": 441266594.0,
      "step": 12683
    },
    {
      "epoch": 2.355431754874652,
      "grad_norm": 1.5517504414436407,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8978779315948486,
      "num_tokens": 441297995.0,
      "step": 12684
    },
    {
      "epoch": 2.3556174558960072,
      "grad_norm": 1.5542393826285779,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8843356966972351,
      "num_tokens": 441332960.0,
      "step": 12685
    },
    {
      "epoch": 2.355803156917363,
      "grad_norm": 1.4150406417622314,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8718270063400269,
      "num_tokens": 441380012.0,
      "step": 12686
    },
    {
      "epoch": 2.3559888579387187,
      "grad_norm": 1.6608644121874734,
      "learning_rate": 1e-06,
      "loss": 0.2889,
      "mean_token_accuracy": 0.8967359662055969,
      "num_tokens": 441409108.0,
      "step": 12687
    },
    {
      "epoch": 2.3561745589600744,
      "grad_norm": 1.6033594394776682,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8896279335021973,
      "num_tokens": 441444139.0,
      "step": 12688
    },
    {
      "epoch": 2.3563602599814297,
      "grad_norm": 1.5975553220924907,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.8959600925445557,
      "num_tokens": 441476040.0,
      "step": 12689
    },
    {
      "epoch": 2.3565459610027855,
      "grad_norm": 1.5249682274478589,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.8902384042739868,
      "num_tokens": 441512761.0,
      "step": 12690
    },
    {
      "epoch": 2.356731662024141,
      "grad_norm": 1.6898166448034384,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8865224719047546,
      "num_tokens": 441543593.0,
      "step": 12691
    },
    {
      "epoch": 2.356917363045497,
      "grad_norm": 1.4551268885421984,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8803861141204834,
      "num_tokens": 441584984.0,
      "step": 12692
    },
    {
      "epoch": 2.357103064066852,
      "grad_norm": 1.5795942116087336,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8908563852310181,
      "num_tokens": 441619070.0,
      "step": 12693
    },
    {
      "epoch": 2.357288765088208,
      "grad_norm": 1.8846724023066082,
      "learning_rate": 1e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8615075945854187,
      "num_tokens": 441650601.0,
      "step": 12694
    },
    {
      "epoch": 2.3574744661095637,
      "grad_norm": 1.5169856243122666,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8939497470855713,
      "num_tokens": 441685814.0,
      "step": 12695
    },
    {
      "epoch": 2.357660167130919,
      "grad_norm": 1.621020322452233,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8796024918556213,
      "num_tokens": 441721632.0,
      "step": 12696
    },
    {
      "epoch": 2.3578458681522747,
      "grad_norm": 1.4796252822993374,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8890097141265869,
      "num_tokens": 441758493.0,
      "step": 12697
    },
    {
      "epoch": 2.3580315691736304,
      "grad_norm": 1.6784317458865146,
      "learning_rate": 1e-06,
      "loss": 0.3335,
      "mean_token_accuracy": 0.881089448928833,
      "num_tokens": 441792719.0,
      "step": 12698
    },
    {
      "epoch": 2.358217270194986,
      "grad_norm": 1.5844323259689248,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8924849033355713,
      "num_tokens": 441823331.0,
      "step": 12699
    },
    {
      "epoch": 2.358402971216342,
      "grad_norm": 1.5849497676736441,
      "learning_rate": 1e-06,
      "loss": 0.2817,
      "mean_token_accuracy": 0.8995089530944824,
      "num_tokens": 441855786.0,
      "step": 12700
    },
    {
      "epoch": 2.358588672237697,
      "grad_norm": 1.8402792948864546,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8784452080726624,
      "num_tokens": 441886554.0,
      "step": 12701
    },
    {
      "epoch": 2.358774373259053,
      "grad_norm": 1.5343273576025527,
      "learning_rate": 1e-06,
      "loss": 0.2842,
      "mean_token_accuracy": 0.8979252576828003,
      "num_tokens": 441917968.0,
      "step": 12702
    },
    {
      "epoch": 2.3589600742804087,
      "grad_norm": 1.5153129152961244,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.896190881729126,
      "num_tokens": 441952705.0,
      "step": 12703
    },
    {
      "epoch": 2.359145775301764,
      "grad_norm": 1.4802077081664964,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8825794458389282,
      "num_tokens": 441990784.0,
      "step": 12704
    },
    {
      "epoch": 2.3593314763231197,
      "grad_norm": 1.4890250200036705,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8819907903671265,
      "num_tokens": 442034292.0,
      "step": 12705
    },
    {
      "epoch": 2.3595171773444754,
      "grad_norm": 1.917788033707857,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8661674857139587,
      "num_tokens": 442062004.0,
      "step": 12706
    },
    {
      "epoch": 2.359702878365831,
      "grad_norm": 1.761422288225096,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8889871835708618,
      "num_tokens": 442086906.0,
      "step": 12707
    },
    {
      "epoch": 2.3598885793871864,
      "grad_norm": 1.4647595072240358,
      "learning_rate": 1e-06,
      "loss": 0.2803,
      "mean_token_accuracy": 0.900451123714447,
      "num_tokens": 442125818.0,
      "step": 12708
    },
    {
      "epoch": 2.360074280408542,
      "grad_norm": 1.7361593903741153,
      "learning_rate": 1e-06,
      "loss": 0.3768,
      "mean_token_accuracy": 0.8703217506408691,
      "num_tokens": 442158056.0,
      "step": 12709
    },
    {
      "epoch": 2.360259981429898,
      "grad_norm": 1.508915929227795,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8867626786231995,
      "num_tokens": 442195971.0,
      "step": 12710
    },
    {
      "epoch": 2.3604456824512536,
      "grad_norm": 1.581699199548891,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8790463209152222,
      "num_tokens": 442231958.0,
      "step": 12711
    },
    {
      "epoch": 2.360631383472609,
      "grad_norm": 1.5179520652792886,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8900974988937378,
      "num_tokens": 442267157.0,
      "step": 12712
    },
    {
      "epoch": 2.3608170844939647,
      "grad_norm": 1.5159346563891274,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8859140872955322,
      "num_tokens": 442303393.0,
      "step": 12713
    },
    {
      "epoch": 2.3610027855153204,
      "grad_norm": 1.6641167557617216,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8767279982566833,
      "num_tokens": 442335157.0,
      "step": 12714
    },
    {
      "epoch": 2.361188486536676,
      "grad_norm": 1.613631066005608,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8856703042984009,
      "num_tokens": 442368672.0,
      "step": 12715
    },
    {
      "epoch": 2.3613741875580314,
      "grad_norm": 1.4298442910828084,
      "learning_rate": 1e-06,
      "loss": 0.2992,
      "mean_token_accuracy": 0.892550528049469,
      "num_tokens": 442409223.0,
      "step": 12716
    },
    {
      "epoch": 2.361559888579387,
      "grad_norm": 1.6718406576413092,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8807213306427002,
      "num_tokens": 442441250.0,
      "step": 12717
    },
    {
      "epoch": 2.361745589600743,
      "grad_norm": 1.7522513463524871,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8756037354469299,
      "num_tokens": 442470999.0,
      "step": 12718
    },
    {
      "epoch": 2.3619312906220986,
      "grad_norm": 1.5789920634292203,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.877881646156311,
      "num_tokens": 442508301.0,
      "step": 12719
    },
    {
      "epoch": 2.362116991643454,
      "grad_norm": 1.5313391750836511,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8792203664779663,
      "num_tokens": 442545571.0,
      "step": 12720
    },
    {
      "epoch": 2.3623026926648096,
      "grad_norm": 1.5912136050715973,
      "learning_rate": 1e-06,
      "loss": 0.276,
      "mean_token_accuracy": 0.9026891589164734,
      "num_tokens": 442577407.0,
      "step": 12721
    },
    {
      "epoch": 2.3624883936861654,
      "grad_norm": 1.4868966368674894,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8938612937927246,
      "num_tokens": 442610300.0,
      "step": 12722
    },
    {
      "epoch": 2.362674094707521,
      "grad_norm": 1.5506229650943881,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8759201765060425,
      "num_tokens": 442648268.0,
      "step": 12723
    },
    {
      "epoch": 2.3628597957288764,
      "grad_norm": 1.562834173610575,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8780777454376221,
      "num_tokens": 442684730.0,
      "step": 12724
    },
    {
      "epoch": 2.363045496750232,
      "grad_norm": 1.5693877305399746,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8797968626022339,
      "num_tokens": 442720163.0,
      "step": 12725
    },
    {
      "epoch": 2.363231197771588,
      "grad_norm": 1.5001380362606507,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8907814621925354,
      "num_tokens": 442756398.0,
      "step": 12726
    },
    {
      "epoch": 2.363416898792943,
      "grad_norm": 1.6009697544739891,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8909479379653931,
      "num_tokens": 442789984.0,
      "step": 12727
    },
    {
      "epoch": 2.363602599814299,
      "grad_norm": 1.5687826487131356,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8855866193771362,
      "num_tokens": 442829374.0,
      "step": 12728
    },
    {
      "epoch": 2.3637883008356546,
      "grad_norm": 1.6369014654594487,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8929821848869324,
      "num_tokens": 442862083.0,
      "step": 12729
    },
    {
      "epoch": 2.3639740018570103,
      "grad_norm": 1.9573418914038856,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8690963983535767,
      "num_tokens": 442888634.0,
      "step": 12730
    },
    {
      "epoch": 2.364159702878366,
      "grad_norm": 1.663355513568294,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8862525224685669,
      "num_tokens": 442919334.0,
      "step": 12731
    },
    {
      "epoch": 2.3643454038997214,
      "grad_norm": 1.613136645982686,
      "learning_rate": 1e-06,
      "loss": 0.2828,
      "mean_token_accuracy": 0.9000562429428101,
      "num_tokens": 442948381.0,
      "step": 12732
    },
    {
      "epoch": 2.364531104921077,
      "grad_norm": 1.4862835030415584,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8891615867614746,
      "num_tokens": 442983554.0,
      "step": 12733
    },
    {
      "epoch": 2.364716805942433,
      "grad_norm": 1.5140873143761049,
      "learning_rate": 1e-06,
      "loss": 0.3088,
      "mean_token_accuracy": 0.8898580074310303,
      "num_tokens": 443019509.0,
      "step": 12734
    },
    {
      "epoch": 2.364902506963788,
      "grad_norm": 1.5786468070546535,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8970710039138794,
      "num_tokens": 443048985.0,
      "step": 12735
    },
    {
      "epoch": 2.365088207985144,
      "grad_norm": 1.460482167749856,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8868967294692993,
      "num_tokens": 443087956.0,
      "step": 12736
    },
    {
      "epoch": 2.3652739090064996,
      "grad_norm": 1.476423861068307,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8905502557754517,
      "num_tokens": 443127786.0,
      "step": 12737
    },
    {
      "epoch": 2.3654596100278553,
      "grad_norm": 1.70443606309223,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8789415955543518,
      "num_tokens": 443158418.0,
      "step": 12738
    },
    {
      "epoch": 2.3656453110492106,
      "grad_norm": 1.4872510351809138,
      "learning_rate": 1e-06,
      "loss": 0.2852,
      "mean_token_accuracy": 0.897257924079895,
      "num_tokens": 443193098.0,
      "step": 12739
    },
    {
      "epoch": 2.3658310120705663,
      "grad_norm": 1.633910554291478,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8904881477355957,
      "num_tokens": 443227386.0,
      "step": 12740
    },
    {
      "epoch": 2.366016713091922,
      "grad_norm": 1.4315168975699262,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8716894388198853,
      "num_tokens": 443271392.0,
      "step": 12741
    },
    {
      "epoch": 2.366202414113278,
      "grad_norm": 1.3861957117451815,
      "learning_rate": 1e-06,
      "loss": 0.2926,
      "mean_token_accuracy": 0.8961687088012695,
      "num_tokens": 443310991.0,
      "step": 12742
    },
    {
      "epoch": 2.366388115134633,
      "grad_norm": 1.6257063313129223,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8905113935470581,
      "num_tokens": 443339752.0,
      "step": 12743
    },
    {
      "epoch": 2.366573816155989,
      "grad_norm": 1.5888268614390664,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8913335204124451,
      "num_tokens": 443370302.0,
      "step": 12744
    },
    {
      "epoch": 2.3667595171773446,
      "grad_norm": 1.6199887529077202,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8712813258171082,
      "num_tokens": 443406287.0,
      "step": 12745
    },
    {
      "epoch": 2.3669452181987003,
      "grad_norm": 1.537276770161992,
      "learning_rate": 1e-06,
      "loss": 0.2735,
      "mean_token_accuracy": 0.9007197618484497,
      "num_tokens": 443434523.0,
      "step": 12746
    },
    {
      "epoch": 2.3671309192200556,
      "grad_norm": 1.6764572572301097,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8954585790634155,
      "num_tokens": 443465207.0,
      "step": 12747
    },
    {
      "epoch": 2.3673166202414113,
      "grad_norm": 1.63048857099738,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8835023641586304,
      "num_tokens": 443497941.0,
      "step": 12748
    },
    {
      "epoch": 2.367502321262767,
      "grad_norm": 1.7684451843597777,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8800705671310425,
      "num_tokens": 443529163.0,
      "step": 12749
    },
    {
      "epoch": 2.3676880222841223,
      "grad_norm": 1.6896205807467135,
      "learning_rate": 1e-06,
      "loss": 0.3718,
      "mean_token_accuracy": 0.8760762214660645,
      "num_tokens": 443564074.0,
      "step": 12750
    },
    {
      "epoch": 2.367873723305478,
      "grad_norm": 1.6788346460788712,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8909305930137634,
      "num_tokens": 443594038.0,
      "step": 12751
    },
    {
      "epoch": 2.368059424326834,
      "grad_norm": 1.6764563562724264,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8789010047912598,
      "num_tokens": 443629294.0,
      "step": 12752
    },
    {
      "epoch": 2.3682451253481895,
      "grad_norm": 1.6567120208507906,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8741647005081177,
      "num_tokens": 443664799.0,
      "step": 12753
    },
    {
      "epoch": 2.3684308263695453,
      "grad_norm": 1.55979434822055,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8794386982917786,
      "num_tokens": 443704475.0,
      "step": 12754
    },
    {
      "epoch": 2.3686165273909006,
      "grad_norm": 1.8059524296987284,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8740917444229126,
      "num_tokens": 443733278.0,
      "step": 12755
    },
    {
      "epoch": 2.3688022284122563,
      "grad_norm": 1.5674727634656007,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8719476461410522,
      "num_tokens": 443771733.0,
      "step": 12756
    },
    {
      "epoch": 2.368987929433612,
      "grad_norm": 1.65250312196153,
      "learning_rate": 1e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8699228167533875,
      "num_tokens": 443809775.0,
      "step": 12757
    },
    {
      "epoch": 2.3691736304549673,
      "grad_norm": 1.6409612062666832,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8879373073577881,
      "num_tokens": 443842100.0,
      "step": 12758
    },
    {
      "epoch": 2.369359331476323,
      "grad_norm": 1.676080767963517,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8935149908065796,
      "num_tokens": 443872348.0,
      "step": 12759
    },
    {
      "epoch": 2.369545032497679,
      "grad_norm": 1.5897721050646334,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8820786476135254,
      "num_tokens": 443907876.0,
      "step": 12760
    },
    {
      "epoch": 2.3697307335190345,
      "grad_norm": 1.4673796734419307,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8902393579483032,
      "num_tokens": 443944374.0,
      "step": 12761
    },
    {
      "epoch": 2.36991643454039,
      "grad_norm": 1.4923594019705206,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8882753849029541,
      "num_tokens": 443985549.0,
      "step": 12762
    },
    {
      "epoch": 2.3701021355617455,
      "grad_norm": 1.516633739371141,
      "learning_rate": 1e-06,
      "loss": 0.2753,
      "mean_token_accuracy": 0.900351881980896,
      "num_tokens": 444017901.0,
      "step": 12763
    },
    {
      "epoch": 2.3702878365831013,
      "grad_norm": 1.5721253295608222,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8873560428619385,
      "num_tokens": 444050648.0,
      "step": 12764
    },
    {
      "epoch": 2.370473537604457,
      "grad_norm": 1.6569356068985541,
      "learning_rate": 1e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8710380792617798,
      "num_tokens": 444086197.0,
      "step": 12765
    },
    {
      "epoch": 2.3706592386258123,
      "grad_norm": 1.5438644750014516,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8895189166069031,
      "num_tokens": 444122024.0,
      "step": 12766
    },
    {
      "epoch": 2.370844939647168,
      "grad_norm": 1.6403337537991403,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8808982372283936,
      "num_tokens": 444154608.0,
      "step": 12767
    },
    {
      "epoch": 2.3710306406685238,
      "grad_norm": 1.4799218715920732,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8950339555740356,
      "num_tokens": 444194005.0,
      "step": 12768
    },
    {
      "epoch": 2.3712163416898795,
      "grad_norm": 1.694824687398902,
      "learning_rate": 1e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8769952654838562,
      "num_tokens": 444225531.0,
      "step": 12769
    },
    {
      "epoch": 2.371402042711235,
      "grad_norm": 1.5737635833408838,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.9031175374984741,
      "num_tokens": 444258208.0,
      "step": 12770
    },
    {
      "epoch": 2.3715877437325905,
      "grad_norm": 1.5569686470504263,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8937163352966309,
      "num_tokens": 444292580.0,
      "step": 12771
    },
    {
      "epoch": 2.3717734447539462,
      "grad_norm": 1.542591975024992,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8759374022483826,
      "num_tokens": 444332529.0,
      "step": 12772
    },
    {
      "epoch": 2.3719591457753015,
      "grad_norm": 1.4858944541714803,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.8935903310775757,
      "num_tokens": 444369920.0,
      "step": 12773
    },
    {
      "epoch": 2.3721448467966573,
      "grad_norm": 1.6320657986723388,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8865342736244202,
      "num_tokens": 444403458.0,
      "step": 12774
    },
    {
      "epoch": 2.372330547818013,
      "grad_norm": 1.5899274376368162,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8893219232559204,
      "num_tokens": 444438672.0,
      "step": 12775
    },
    {
      "epoch": 2.3725162488393687,
      "grad_norm": 1.529980110494929,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8785393238067627,
      "num_tokens": 444473697.0,
      "step": 12776
    },
    {
      "epoch": 2.3727019498607245,
      "grad_norm": 1.4517555756691471,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8861281871795654,
      "num_tokens": 444513303.0,
      "step": 12777
    },
    {
      "epoch": 2.3728876508820798,
      "grad_norm": 1.9445666103657808,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8828110098838806,
      "num_tokens": 444535493.0,
      "step": 12778
    },
    {
      "epoch": 2.3730733519034355,
      "grad_norm": 1.5089427002769182,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8793691992759705,
      "num_tokens": 444577153.0,
      "step": 12779
    },
    {
      "epoch": 2.3732590529247912,
      "grad_norm": 1.4802757410145877,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8940685987472534,
      "num_tokens": 444611209.0,
      "step": 12780
    },
    {
      "epoch": 2.3734447539461465,
      "grad_norm": 1.5823453259453029,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8867304921150208,
      "num_tokens": 444645563.0,
      "step": 12781
    },
    {
      "epoch": 2.3736304549675022,
      "grad_norm": 1.7334059932948953,
      "learning_rate": 1e-06,
      "loss": 0.3762,
      "mean_token_accuracy": 0.8698983192443848,
      "num_tokens": 444674154.0,
      "step": 12782
    },
    {
      "epoch": 2.373816155988858,
      "grad_norm": 1.5413487838595985,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8908270597457886,
      "num_tokens": 444707999.0,
      "step": 12783
    },
    {
      "epoch": 2.3740018570102137,
      "grad_norm": 1.5104768411743483,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8861702084541321,
      "num_tokens": 444743186.0,
      "step": 12784
    },
    {
      "epoch": 2.374187558031569,
      "grad_norm": 1.720733249749788,
      "learning_rate": 1e-06,
      "loss": 0.2926,
      "mean_token_accuracy": 0.8963613510131836,
      "num_tokens": 444769381.0,
      "step": 12785
    },
    {
      "epoch": 2.3743732590529247,
      "grad_norm": 1.6126593597610996,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8763148784637451,
      "num_tokens": 444805059.0,
      "step": 12786
    },
    {
      "epoch": 2.3745589600742805,
      "grad_norm": 1.644048547500624,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8890156149864197,
      "num_tokens": 444836549.0,
      "step": 12787
    },
    {
      "epoch": 2.374744661095636,
      "grad_norm": 1.472763676007015,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.891211986541748,
      "num_tokens": 444873570.0,
      "step": 12788
    },
    {
      "epoch": 2.3749303621169915,
      "grad_norm": 1.4504617490756029,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8849425315856934,
      "num_tokens": 444913840.0,
      "step": 12789
    },
    {
      "epoch": 2.3751160631383472,
      "grad_norm": 1.5721532327735455,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8804399371147156,
      "num_tokens": 444948721.0,
      "step": 12790
    },
    {
      "epoch": 2.375301764159703,
      "grad_norm": 1.6001873280865662,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8890445828437805,
      "num_tokens": 444983710.0,
      "step": 12791
    },
    {
      "epoch": 2.3754874651810587,
      "grad_norm": 1.6224941829166042,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8920530080795288,
      "num_tokens": 445015517.0,
      "step": 12792
    },
    {
      "epoch": 2.375673166202414,
      "grad_norm": 1.560757802841523,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.895790696144104,
      "num_tokens": 445049372.0,
      "step": 12793
    },
    {
      "epoch": 2.3758588672237697,
      "grad_norm": 1.8852565687702845,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.885231077671051,
      "num_tokens": 445076864.0,
      "step": 12794
    },
    {
      "epoch": 2.3760445682451254,
      "grad_norm": 1.4458442725378606,
      "learning_rate": 1e-06,
      "loss": 0.2762,
      "mean_token_accuracy": 0.9017438292503357,
      "num_tokens": 445114421.0,
      "step": 12795
    },
    {
      "epoch": 2.3762302692664807,
      "grad_norm": 1.4837916037690684,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8912423849105835,
      "num_tokens": 445150248.0,
      "step": 12796
    },
    {
      "epoch": 2.3764159702878365,
      "grad_norm": 1.679087715969336,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8784872889518738,
      "num_tokens": 445181560.0,
      "step": 12797
    },
    {
      "epoch": 2.376601671309192,
      "grad_norm": 1.5217688846377002,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8817764520645142,
      "num_tokens": 445222947.0,
      "step": 12798
    },
    {
      "epoch": 2.376787372330548,
      "grad_norm": 1.517203313062023,
      "learning_rate": 1e-06,
      "loss": 0.2945,
      "mean_token_accuracy": 0.8963807821273804,
      "num_tokens": 445255689.0,
      "step": 12799
    },
    {
      "epoch": 2.3769730733519037,
      "grad_norm": 1.668606399141064,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8743106126785278,
      "num_tokens": 445286214.0,
      "step": 12800
    },
    {
      "epoch": 2.377158774373259,
      "grad_norm": 1.623165531480467,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8789432048797607,
      "num_tokens": 445318561.0,
      "step": 12801
    },
    {
      "epoch": 2.3773444753946147,
      "grad_norm": 1.6568286404519887,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8824665546417236,
      "num_tokens": 445350304.0,
      "step": 12802
    },
    {
      "epoch": 2.3775301764159704,
      "grad_norm": 1.6130182512772375,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8922396302223206,
      "num_tokens": 445385291.0,
      "step": 12803
    },
    {
      "epoch": 2.3777158774373257,
      "grad_norm": 1.4267288284176347,
      "learning_rate": 1e-06,
      "loss": 0.2799,
      "mean_token_accuracy": 0.898939847946167,
      "num_tokens": 445424319.0,
      "step": 12804
    },
    {
      "epoch": 2.3779015784586814,
      "grad_norm": 1.6108139492956286,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8809704780578613,
      "num_tokens": 445459075.0,
      "step": 12805
    },
    {
      "epoch": 2.378087279480037,
      "grad_norm": 1.5211707369712906,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.8969526290893555,
      "num_tokens": 445493603.0,
      "step": 12806
    },
    {
      "epoch": 2.378272980501393,
      "grad_norm": 1.5404445267832871,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8899096250534058,
      "num_tokens": 445525495.0,
      "step": 12807
    },
    {
      "epoch": 2.378458681522748,
      "grad_norm": 1.5330804179824633,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8913341760635376,
      "num_tokens": 445561908.0,
      "step": 12808
    },
    {
      "epoch": 2.378644382544104,
      "grad_norm": 1.4812081081211692,
      "learning_rate": 1e-06,
      "loss": 0.2769,
      "mean_token_accuracy": 0.9005111455917358,
      "num_tokens": 445597545.0,
      "step": 12809
    },
    {
      "epoch": 2.3788300835654597,
      "grad_norm": 1.6055216613648287,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8867141604423523,
      "num_tokens": 445630199.0,
      "step": 12810
    },
    {
      "epoch": 2.3790157845868154,
      "grad_norm": 1.649060242897638,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8757323026657104,
      "num_tokens": 445667816.0,
      "step": 12811
    },
    {
      "epoch": 2.3792014856081707,
      "grad_norm": 1.5388446158120996,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8840643167495728,
      "num_tokens": 445705337.0,
      "step": 12812
    },
    {
      "epoch": 2.3793871866295264,
      "grad_norm": 1.5396637593944318,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.882571816444397,
      "num_tokens": 445741008.0,
      "step": 12813
    },
    {
      "epoch": 2.379572887650882,
      "grad_norm": 1.6586298696622965,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8790611028671265,
      "num_tokens": 445774115.0,
      "step": 12814
    },
    {
      "epoch": 2.379758588672238,
      "grad_norm": 1.6660472092454515,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.9005365967750549,
      "num_tokens": 445805130.0,
      "step": 12815
    },
    {
      "epoch": 2.379944289693593,
      "grad_norm": 1.5173520840137618,
      "learning_rate": 1e-06,
      "loss": 0.2805,
      "mean_token_accuracy": 0.8994391560554504,
      "num_tokens": 445837829.0,
      "step": 12816
    },
    {
      "epoch": 2.380129990714949,
      "grad_norm": 1.6224867982593187,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8921623229980469,
      "num_tokens": 445870279.0,
      "step": 12817
    },
    {
      "epoch": 2.3803156917363046,
      "grad_norm": 1.5480140573933632,
      "learning_rate": 1e-06,
      "loss": 0.2967,
      "mean_token_accuracy": 0.8943111300468445,
      "num_tokens": 445904984.0,
      "step": 12818
    },
    {
      "epoch": 2.38050139275766,
      "grad_norm": 1.6499658894161184,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8866246938705444,
      "num_tokens": 445935900.0,
      "step": 12819
    },
    {
      "epoch": 2.3806870937790157,
      "grad_norm": 1.5884473338631622,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8830731511116028,
      "num_tokens": 445969996.0,
      "step": 12820
    },
    {
      "epoch": 2.3808727948003714,
      "grad_norm": 1.6887207286528279,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8820046186447144,
      "num_tokens": 446009192.0,
      "step": 12821
    },
    {
      "epoch": 2.381058495821727,
      "grad_norm": 1.5769669797778052,
      "learning_rate": 1e-06,
      "loss": 0.3056,
      "mean_token_accuracy": 0.8934473991394043,
      "num_tokens": 446041275.0,
      "step": 12822
    },
    {
      "epoch": 2.381244196843083,
      "grad_norm": 1.6091705671833494,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8869194984436035,
      "num_tokens": 446073687.0,
      "step": 12823
    },
    {
      "epoch": 2.381429897864438,
      "grad_norm": 1.4634226203297052,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8876917362213135,
      "num_tokens": 446115441.0,
      "step": 12824
    },
    {
      "epoch": 2.381615598885794,
      "grad_norm": 1.6635039697589722,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.88664311170578,
      "num_tokens": 446146882.0,
      "step": 12825
    },
    {
      "epoch": 2.3818012999071496,
      "grad_norm": 1.5713823642521407,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8925391435623169,
      "num_tokens": 446183266.0,
      "step": 12826
    },
    {
      "epoch": 2.381987000928505,
      "grad_norm": 1.8157025775230202,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8723112344741821,
      "num_tokens": 446213002.0,
      "step": 12827
    },
    {
      "epoch": 2.3821727019498606,
      "grad_norm": 1.4900751787379645,
      "learning_rate": 1e-06,
      "loss": 0.295,
      "mean_token_accuracy": 0.8939505815505981,
      "num_tokens": 446248583.0,
      "step": 12828
    },
    {
      "epoch": 2.3823584029712164,
      "grad_norm": 1.6214927816177325,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.897840678691864,
      "num_tokens": 446278244.0,
      "step": 12829
    },
    {
      "epoch": 2.382544103992572,
      "grad_norm": 1.3936726042292258,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8883086442947388,
      "num_tokens": 446319562.0,
      "step": 12830
    },
    {
      "epoch": 2.3827298050139274,
      "grad_norm": 1.5714722067814368,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.877874493598938,
      "num_tokens": 446355020.0,
      "step": 12831
    },
    {
      "epoch": 2.382915506035283,
      "grad_norm": 1.5981607522566998,
      "learning_rate": 1e-06,
      "loss": 0.2906,
      "mean_token_accuracy": 0.895201563835144,
      "num_tokens": 446385896.0,
      "step": 12832
    },
    {
      "epoch": 2.383101207056639,
      "grad_norm": 1.6312920218932705,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8869093656539917,
      "num_tokens": 446418455.0,
      "step": 12833
    },
    {
      "epoch": 2.3832869080779946,
      "grad_norm": 1.6584068676016086,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8940203189849854,
      "num_tokens": 446446828.0,
      "step": 12834
    },
    {
      "epoch": 2.38347260909935,
      "grad_norm": 1.5094053231535818,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8812257051467896,
      "num_tokens": 446484358.0,
      "step": 12835
    },
    {
      "epoch": 2.3836583101207056,
      "grad_norm": 1.5730118796087025,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.887758195400238,
      "num_tokens": 446518998.0,
      "step": 12836
    },
    {
      "epoch": 2.3838440111420613,
      "grad_norm": 1.7990749900577143,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8827617168426514,
      "num_tokens": 446549208.0,
      "step": 12837
    },
    {
      "epoch": 2.384029712163417,
      "grad_norm": 1.6253484136186442,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8824065923690796,
      "num_tokens": 446585328.0,
      "step": 12838
    },
    {
      "epoch": 2.3842154131847724,
      "grad_norm": 1.3980540468904772,
      "learning_rate": 1e-06,
      "loss": 0.2988,
      "mean_token_accuracy": 0.8944236040115356,
      "num_tokens": 446625721.0,
      "step": 12839
    },
    {
      "epoch": 2.384401114206128,
      "grad_norm": 2.018529843777321,
      "learning_rate": 1e-06,
      "loss": 0.3609,
      "mean_token_accuracy": 0.8759945631027222,
      "num_tokens": 446649291.0,
      "step": 12840
    },
    {
      "epoch": 2.384586815227484,
      "grad_norm": 1.5725595832464072,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8882042169570923,
      "num_tokens": 446684330.0,
      "step": 12841
    },
    {
      "epoch": 2.384772516248839,
      "grad_norm": 1.627250216797403,
      "learning_rate": 1e-06,
      "loss": 0.3411,
      "mean_token_accuracy": 0.8835436105728149,
      "num_tokens": 446717733.0,
      "step": 12842
    },
    {
      "epoch": 2.384958217270195,
      "grad_norm": 1.6899010983484781,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8766157031059265,
      "num_tokens": 446753324.0,
      "step": 12843
    },
    {
      "epoch": 2.3851439182915506,
      "grad_norm": 1.5926100134956822,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8933884501457214,
      "num_tokens": 446787718.0,
      "step": 12844
    },
    {
      "epoch": 2.3853296193129063,
      "grad_norm": 1.5223116697916774,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8937634229660034,
      "num_tokens": 446823593.0,
      "step": 12845
    },
    {
      "epoch": 2.385515320334262,
      "grad_norm": 1.6780446537581115,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8917834162712097,
      "num_tokens": 446853004.0,
      "step": 12846
    },
    {
      "epoch": 2.3857010213556173,
      "grad_norm": 1.5334664462271053,
      "learning_rate": 1e-06,
      "loss": 0.304,
      "mean_token_accuracy": 0.8945045471191406,
      "num_tokens": 446887426.0,
      "step": 12847
    },
    {
      "epoch": 2.385886722376973,
      "grad_norm": 1.4510403396056397,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8815677165985107,
      "num_tokens": 446926218.0,
      "step": 12848
    },
    {
      "epoch": 2.386072423398329,
      "grad_norm": 1.5501393838309179,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.882277250289917,
      "num_tokens": 446959054.0,
      "step": 12849
    },
    {
      "epoch": 2.386258124419684,
      "grad_norm": 1.6522474217304792,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8993257284164429,
      "num_tokens": 446986020.0,
      "step": 12850
    },
    {
      "epoch": 2.38644382544104,
      "grad_norm": 1.4928907859226848,
      "learning_rate": 1e-06,
      "loss": 0.278,
      "mean_token_accuracy": 0.9026932716369629,
      "num_tokens": 447023014.0,
      "step": 12851
    },
    {
      "epoch": 2.3866295264623956,
      "grad_norm": 1.6059379964894986,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8878321647644043,
      "num_tokens": 447057531.0,
      "step": 12852
    },
    {
      "epoch": 2.3868152274837513,
      "grad_norm": 1.5418789193525793,
      "learning_rate": 1e-06,
      "loss": 0.3088,
      "mean_token_accuracy": 0.8908216953277588,
      "num_tokens": 447096080.0,
      "step": 12853
    },
    {
      "epoch": 2.3870009285051066,
      "grad_norm": 1.6443535581483695,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8747543096542358,
      "num_tokens": 447129084.0,
      "step": 12854
    },
    {
      "epoch": 2.3871866295264623,
      "grad_norm": 1.6366559456429604,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8886522650718689,
      "num_tokens": 447161556.0,
      "step": 12855
    },
    {
      "epoch": 2.387372330547818,
      "grad_norm": 1.5660802538166296,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8858621716499329,
      "num_tokens": 447194630.0,
      "step": 12856
    },
    {
      "epoch": 2.387558031569174,
      "grad_norm": 1.4654253742622674,
      "learning_rate": 1e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.8944998979568481,
      "num_tokens": 447232206.0,
      "step": 12857
    },
    {
      "epoch": 2.387743732590529,
      "grad_norm": 1.5718275817536713,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8693469762802124,
      "num_tokens": 447269124.0,
      "step": 12858
    },
    {
      "epoch": 2.387929433611885,
      "grad_norm": 1.4886985435395452,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8868983387947083,
      "num_tokens": 447310004.0,
      "step": 12859
    },
    {
      "epoch": 2.3881151346332405,
      "grad_norm": 1.4686407110116626,
      "learning_rate": 1e-06,
      "loss": 0.2794,
      "mean_token_accuracy": 0.9008625149726868,
      "num_tokens": 447345984.0,
      "step": 12860
    },
    {
      "epoch": 2.3883008356545963,
      "grad_norm": 1.5259807443308229,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8891223669052124,
      "num_tokens": 447383226.0,
      "step": 12861
    },
    {
      "epoch": 2.3884865366759516,
      "grad_norm": 1.5797205979811904,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8904232978820801,
      "num_tokens": 447419373.0,
      "step": 12862
    },
    {
      "epoch": 2.3886722376973073,
      "grad_norm": 1.5596842896190442,
      "learning_rate": 1e-06,
      "loss": 0.3045,
      "mean_token_accuracy": 0.8901869058609009,
      "num_tokens": 447452375.0,
      "step": 12863
    },
    {
      "epoch": 2.388857938718663,
      "grad_norm": 1.5218340547724465,
      "learning_rate": 1e-06,
      "loss": 0.296,
      "mean_token_accuracy": 0.894535481929779,
      "num_tokens": 447487780.0,
      "step": 12864
    },
    {
      "epoch": 2.3890436397400183,
      "grad_norm": 1.5595414304793358,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8808282613754272,
      "num_tokens": 447524353.0,
      "step": 12865
    },
    {
      "epoch": 2.389229340761374,
      "grad_norm": 1.4418129828032509,
      "learning_rate": 1e-06,
      "loss": 0.2887,
      "mean_token_accuracy": 0.8951658606529236,
      "num_tokens": 447563343.0,
      "step": 12866
    },
    {
      "epoch": 2.38941504178273,
      "grad_norm": 1.4974693295422523,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8907747268676758,
      "num_tokens": 447598197.0,
      "step": 12867
    },
    {
      "epoch": 2.3896007428040855,
      "grad_norm": 1.5045390404795052,
      "learning_rate": 1e-06,
      "loss": 0.268,
      "mean_token_accuracy": 0.8998086452484131,
      "num_tokens": 447631415.0,
      "step": 12868
    },
    {
      "epoch": 2.3897864438254413,
      "grad_norm": 1.5607983678726014,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.886660099029541,
      "num_tokens": 447666588.0,
      "step": 12869
    },
    {
      "epoch": 2.3899721448467965,
      "grad_norm": 1.6834966165288354,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8837385177612305,
      "num_tokens": 447698766.0,
      "step": 12870
    },
    {
      "epoch": 2.3901578458681523,
      "grad_norm": 1.4681901036483282,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.889541506767273,
      "num_tokens": 447739159.0,
      "step": 12871
    },
    {
      "epoch": 2.390343546889508,
      "grad_norm": 1.8072298147228718,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8857426643371582,
      "num_tokens": 447779604.0,
      "step": 12872
    },
    {
      "epoch": 2.3905292479108633,
      "grad_norm": 1.5235529150902019,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8892785906791687,
      "num_tokens": 447817298.0,
      "step": 12873
    },
    {
      "epoch": 2.390714948932219,
      "grad_norm": 1.6584657833051917,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.888740599155426,
      "num_tokens": 447849203.0,
      "step": 12874
    },
    {
      "epoch": 2.3909006499535748,
      "grad_norm": 1.5941176494644005,
      "learning_rate": 1e-06,
      "loss": 0.2861,
      "mean_token_accuracy": 0.8982985019683838,
      "num_tokens": 447880008.0,
      "step": 12875
    },
    {
      "epoch": 2.3910863509749305,
      "grad_norm": 1.6071547636613268,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8902232646942139,
      "num_tokens": 447917178.0,
      "step": 12876
    },
    {
      "epoch": 2.391272051996286,
      "grad_norm": 1.6369836984387838,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8782403469085693,
      "num_tokens": 447954278.0,
      "step": 12877
    },
    {
      "epoch": 2.3914577530176415,
      "grad_norm": 1.446724488114634,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8793327808380127,
      "num_tokens": 447998235.0,
      "step": 12878
    },
    {
      "epoch": 2.3916434540389973,
      "grad_norm": 1.592565299473614,
      "learning_rate": 1e-06,
      "loss": 0.2907,
      "mean_token_accuracy": 0.8986170291900635,
      "num_tokens": 448030507.0,
      "step": 12879
    },
    {
      "epoch": 2.391829155060353,
      "grad_norm": 1.4842000886959046,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8868763446807861,
      "num_tokens": 448068336.0,
      "step": 12880
    },
    {
      "epoch": 2.3920148560817083,
      "grad_norm": 1.9004082502618738,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8773558139801025,
      "num_tokens": 448092264.0,
      "step": 12881
    },
    {
      "epoch": 2.392200557103064,
      "grad_norm": 1.6116906460847111,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8870360255241394,
      "num_tokens": 448127904.0,
      "step": 12882
    },
    {
      "epoch": 2.3923862581244197,
      "grad_norm": 1.4932329455279798,
      "learning_rate": 1e-06,
      "loss": 0.2785,
      "mean_token_accuracy": 0.9007927775382996,
      "num_tokens": 448161608.0,
      "step": 12883
    },
    {
      "epoch": 2.3925719591457755,
      "grad_norm": 1.5582539589343622,
      "learning_rate": 1e-06,
      "loss": 0.2875,
      "mean_token_accuracy": 0.8997652530670166,
      "num_tokens": 448196030.0,
      "step": 12884
    },
    {
      "epoch": 2.3927576601671308,
      "grad_norm": 1.4881841897317665,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.885410487651825,
      "num_tokens": 448234993.0,
      "step": 12885
    },
    {
      "epoch": 2.3929433611884865,
      "grad_norm": 1.687221826709704,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.88213050365448,
      "num_tokens": 448265727.0,
      "step": 12886
    },
    {
      "epoch": 2.3931290622098422,
      "grad_norm": 1.4706560256911099,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.881043016910553,
      "num_tokens": 448306387.0,
      "step": 12887
    },
    {
      "epoch": 2.393314763231198,
      "grad_norm": 1.4709985618339392,
      "learning_rate": 1e-06,
      "loss": 0.3023,
      "mean_token_accuracy": 0.8914781212806702,
      "num_tokens": 448345650.0,
      "step": 12888
    },
    {
      "epoch": 2.3935004642525533,
      "grad_norm": 1.619352121103444,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8955700397491455,
      "num_tokens": 448378084.0,
      "step": 12889
    },
    {
      "epoch": 2.393686165273909,
      "grad_norm": 1.5343670089697639,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8800286054611206,
      "num_tokens": 448411696.0,
      "step": 12890
    },
    {
      "epoch": 2.3938718662952647,
      "grad_norm": 1.5835624741589507,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8785755038261414,
      "num_tokens": 448447806.0,
      "step": 12891
    },
    {
      "epoch": 2.3940575673166204,
      "grad_norm": 1.4109191555359208,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8906620740890503,
      "num_tokens": 448488615.0,
      "step": 12892
    },
    {
      "epoch": 2.3942432683379757,
      "grad_norm": 1.6941195230850563,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.880978524684906,
      "num_tokens": 448522257.0,
      "step": 12893
    },
    {
      "epoch": 2.3944289693593315,
      "grad_norm": 1.7278126285265873,
      "learning_rate": 1e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.8719065189361572,
      "num_tokens": 448553786.0,
      "step": 12894
    },
    {
      "epoch": 2.394614670380687,
      "grad_norm": 1.5461543289085538,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8952903747558594,
      "num_tokens": 448589615.0,
      "step": 12895
    },
    {
      "epoch": 2.3948003714020425,
      "grad_norm": 1.405993507796507,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8811342716217041,
      "num_tokens": 448629940.0,
      "step": 12896
    },
    {
      "epoch": 2.3949860724233982,
      "grad_norm": 1.7923585485692193,
      "learning_rate": 1e-06,
      "loss": 0.2972,
      "mean_token_accuracy": 0.8927741646766663,
      "num_tokens": 448652999.0,
      "step": 12897
    },
    {
      "epoch": 2.395171773444754,
      "grad_norm": 1.5028897261831848,
      "learning_rate": 1e-06,
      "loss": 0.2846,
      "mean_token_accuracy": 0.8945578336715698,
      "num_tokens": 448687062.0,
      "step": 12898
    },
    {
      "epoch": 2.3953574744661097,
      "grad_norm": 1.477562571826684,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8788837194442749,
      "num_tokens": 448726116.0,
      "step": 12899
    },
    {
      "epoch": 2.3955431754874654,
      "grad_norm": 1.577069558177803,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8822328448295593,
      "num_tokens": 448758710.0,
      "step": 12900
    },
    {
      "epoch": 2.3957288765088207,
      "grad_norm": 1.6995382418794196,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.8827763795852661,
      "num_tokens": 448792780.0,
      "step": 12901
    },
    {
      "epoch": 2.3959145775301764,
      "grad_norm": 1.5753348766121613,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8810039758682251,
      "num_tokens": 448828037.0,
      "step": 12902
    },
    {
      "epoch": 2.396100278551532,
      "grad_norm": 1.5987974270855088,
      "learning_rate": 1e-06,
      "loss": 0.2926,
      "mean_token_accuracy": 0.896392822265625,
      "num_tokens": 448860537.0,
      "step": 12903
    },
    {
      "epoch": 2.3962859795728875,
      "grad_norm": 1.6257306805620646,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8751180171966553,
      "num_tokens": 448894667.0,
      "step": 12904
    },
    {
      "epoch": 2.396471680594243,
      "grad_norm": 1.5154656581022827,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8923898339271545,
      "num_tokens": 448934507.0,
      "step": 12905
    },
    {
      "epoch": 2.396657381615599,
      "grad_norm": 1.6404260124643302,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.893964946269989,
      "num_tokens": 448964194.0,
      "step": 12906
    },
    {
      "epoch": 2.3968430826369547,
      "grad_norm": 1.6022714889900938,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8910154104232788,
      "num_tokens": 448995563.0,
      "step": 12907
    },
    {
      "epoch": 2.39702878365831,
      "grad_norm": 1.7512015244322408,
      "learning_rate": 1e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8595869541168213,
      "num_tokens": 449026847.0,
      "step": 12908
    },
    {
      "epoch": 2.3972144846796657,
      "grad_norm": 1.478426855612231,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8860470056533813,
      "num_tokens": 449066457.0,
      "step": 12909
    },
    {
      "epoch": 2.3974001857010214,
      "grad_norm": 1.579199977532418,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8890000581741333,
      "num_tokens": 449103734.0,
      "step": 12910
    },
    {
      "epoch": 2.397585886722377,
      "grad_norm": 1.566914584463044,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.8984882831573486,
      "num_tokens": 449133447.0,
      "step": 12911
    },
    {
      "epoch": 2.3977715877437324,
      "grad_norm": 1.6934766432449992,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.881271481513977,
      "num_tokens": 449165881.0,
      "step": 12912
    },
    {
      "epoch": 2.397957288765088,
      "grad_norm": 1.5635242764683652,
      "learning_rate": 1e-06,
      "loss": 0.3102,
      "mean_token_accuracy": 0.8871853947639465,
      "num_tokens": 449196311.0,
      "step": 12913
    },
    {
      "epoch": 2.398142989786444,
      "grad_norm": 1.622192082631372,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.88055419921875,
      "num_tokens": 449232738.0,
      "step": 12914
    },
    {
      "epoch": 2.3983286908077996,
      "grad_norm": 1.5307184313632956,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8911353349685669,
      "num_tokens": 449267532.0,
      "step": 12915
    },
    {
      "epoch": 2.398514391829155,
      "grad_norm": 1.664895903668111,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.879668116569519,
      "num_tokens": 449301931.0,
      "step": 12916
    },
    {
      "epoch": 2.3987000928505107,
      "grad_norm": 1.5701508152484598,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8955624103546143,
      "num_tokens": 449336304.0,
      "step": 12917
    },
    {
      "epoch": 2.3988857938718664,
      "grad_norm": 1.6361934197799053,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8827857971191406,
      "num_tokens": 449371131.0,
      "step": 12918
    },
    {
      "epoch": 2.3990714948932217,
      "grad_norm": 1.654213664482178,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.873267412185669,
      "num_tokens": 449406467.0,
      "step": 12919
    },
    {
      "epoch": 2.3992571959145774,
      "grad_norm": 1.5488448501300625,
      "learning_rate": 1e-06,
      "loss": 0.3088,
      "mean_token_accuracy": 0.8876147270202637,
      "num_tokens": 449444371.0,
      "step": 12920
    },
    {
      "epoch": 2.399442896935933,
      "grad_norm": 1.625157698792215,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8774328827857971,
      "num_tokens": 449477740.0,
      "step": 12921
    },
    {
      "epoch": 2.399628597957289,
      "grad_norm": 1.4616395457818014,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8904115557670593,
      "num_tokens": 449520602.0,
      "step": 12922
    },
    {
      "epoch": 2.3998142989786446,
      "grad_norm": 1.4248707895675206,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8938212394714355,
      "num_tokens": 449559132.0,
      "step": 12923
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.4770876386914107,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8897283673286438,
      "num_tokens": 449595403.0,
      "step": 12924
    },
    {
      "epoch": 2.4001857010213556,
      "grad_norm": 1.506556606571512,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8902055025100708,
      "num_tokens": 449629303.0,
      "step": 12925
    },
    {
      "epoch": 2.4003714020427114,
      "grad_norm": 1.6307877912996782,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8871647119522095,
      "num_tokens": 449663696.0,
      "step": 12926
    },
    {
      "epoch": 2.4005571030640667,
      "grad_norm": 1.6186382306126095,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8935171961784363,
      "num_tokens": 449697823.0,
      "step": 12927
    },
    {
      "epoch": 2.4007428040854224,
      "grad_norm": 1.504067826345839,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8910531997680664,
      "num_tokens": 449733403.0,
      "step": 12928
    },
    {
      "epoch": 2.400928505106778,
      "grad_norm": 1.6731017798339811,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.888586163520813,
      "num_tokens": 449763567.0,
      "step": 12929
    },
    {
      "epoch": 2.401114206128134,
      "grad_norm": 1.5162616774177966,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8744183778762817,
      "num_tokens": 449804250.0,
      "step": 12930
    },
    {
      "epoch": 2.401299907149489,
      "grad_norm": 1.5021449741460016,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.9018571376800537,
      "num_tokens": 449840129.0,
      "step": 12931
    },
    {
      "epoch": 2.401485608170845,
      "grad_norm": 1.5206090190851085,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8811466693878174,
      "num_tokens": 449879082.0,
      "step": 12932
    },
    {
      "epoch": 2.4016713091922006,
      "grad_norm": 1.4980956004957513,
      "learning_rate": 1e-06,
      "loss": 0.2981,
      "mean_token_accuracy": 0.8927364349365234,
      "num_tokens": 449915063.0,
      "step": 12933
    },
    {
      "epoch": 2.4018570102135564,
      "grad_norm": 1.5128889618005559,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.890067994594574,
      "num_tokens": 449949632.0,
      "step": 12934
    },
    {
      "epoch": 2.4020427112349116,
      "grad_norm": 1.6043552770407374,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8877614736557007,
      "num_tokens": 449980631.0,
      "step": 12935
    },
    {
      "epoch": 2.4022284122562674,
      "grad_norm": 1.7075582095145876,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8958350419998169,
      "num_tokens": 450007745.0,
      "step": 12936
    },
    {
      "epoch": 2.402414113277623,
      "grad_norm": 1.4559088886037899,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8843252658843994,
      "num_tokens": 450048352.0,
      "step": 12937
    },
    {
      "epoch": 2.402599814298979,
      "grad_norm": 1.518167157710461,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.883899450302124,
      "num_tokens": 450083752.0,
      "step": 12938
    },
    {
      "epoch": 2.402785515320334,
      "grad_norm": 1.5452546034318464,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8842094540596008,
      "num_tokens": 450118890.0,
      "step": 12939
    },
    {
      "epoch": 2.40297121634169,
      "grad_norm": 1.3750355384510458,
      "learning_rate": 1e-06,
      "loss": 0.2755,
      "mean_token_accuracy": 0.9020495414733887,
      "num_tokens": 450158192.0,
      "step": 12940
    },
    {
      "epoch": 2.4031569173630456,
      "grad_norm": 1.525214940411672,
      "learning_rate": 1e-06,
      "loss": 0.2933,
      "mean_token_accuracy": 0.8949233293533325,
      "num_tokens": 450192138.0,
      "step": 12941
    },
    {
      "epoch": 2.403342618384401,
      "grad_norm": 1.5401887443830626,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8875012397766113,
      "num_tokens": 450227884.0,
      "step": 12942
    },
    {
      "epoch": 2.4035283194057566,
      "grad_norm": 1.620690641778399,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8910486698150635,
      "num_tokens": 450261076.0,
      "step": 12943
    },
    {
      "epoch": 2.4037140204271124,
      "grad_norm": 1.5066092282528885,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8927215337753296,
      "num_tokens": 450296285.0,
      "step": 12944
    },
    {
      "epoch": 2.403899721448468,
      "grad_norm": 1.595990471244773,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8890249133110046,
      "num_tokens": 450328833.0,
      "step": 12945
    },
    {
      "epoch": 2.404085422469824,
      "grad_norm": 1.7169360148367836,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.893909752368927,
      "num_tokens": 450360368.0,
      "step": 12946
    },
    {
      "epoch": 2.404271123491179,
      "grad_norm": 1.5022596329430242,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8852116465568542,
      "num_tokens": 450395854.0,
      "step": 12947
    },
    {
      "epoch": 2.404456824512535,
      "grad_norm": 1.6382743016740737,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8808835744857788,
      "num_tokens": 450430550.0,
      "step": 12948
    },
    {
      "epoch": 2.4046425255338906,
      "grad_norm": 1.5702461661584748,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.8912879228591919,
      "num_tokens": 450462723.0,
      "step": 12949
    },
    {
      "epoch": 2.404828226555246,
      "grad_norm": 1.603778979379632,
      "learning_rate": 1e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8834104537963867,
      "num_tokens": 450497782.0,
      "step": 12950
    },
    {
      "epoch": 2.4050139275766016,
      "grad_norm": 1.6797680836135742,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.8962392807006836,
      "num_tokens": 450528913.0,
      "step": 12951
    },
    {
      "epoch": 2.4051996285979573,
      "grad_norm": 1.5795491546562992,
      "learning_rate": 1e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8712798357009888,
      "num_tokens": 450564235.0,
      "step": 12952
    },
    {
      "epoch": 2.405385329619313,
      "grad_norm": 1.4215683334000586,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8909642696380615,
      "num_tokens": 450605168.0,
      "step": 12953
    },
    {
      "epoch": 2.4055710306406684,
      "grad_norm": 1.474014942350448,
      "learning_rate": 1e-06,
      "loss": 0.3343,
      "mean_token_accuracy": 0.8831679224967957,
      "num_tokens": 450644334.0,
      "step": 12954
    },
    {
      "epoch": 2.405756731662024,
      "grad_norm": 1.6076788697872173,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8817903995513916,
      "num_tokens": 450679654.0,
      "step": 12955
    },
    {
      "epoch": 2.40594243268338,
      "grad_norm": 1.6037566652707842,
      "learning_rate": 1e-06,
      "loss": 0.2956,
      "mean_token_accuracy": 0.8980674743652344,
      "num_tokens": 450711434.0,
      "step": 12956
    },
    {
      "epoch": 2.4061281337047355,
      "grad_norm": 1.830694358417106,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8831188082695007,
      "num_tokens": 450739047.0,
      "step": 12957
    },
    {
      "epoch": 2.406313834726091,
      "grad_norm": 1.5210469637636108,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.874862790107727,
      "num_tokens": 450776467.0,
      "step": 12958
    },
    {
      "epoch": 2.4064995357474466,
      "grad_norm": 1.5023034010792815,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8917912244796753,
      "num_tokens": 450812587.0,
      "step": 12959
    },
    {
      "epoch": 2.4066852367688023,
      "grad_norm": 1.6034255372137385,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8870436549186707,
      "num_tokens": 450847287.0,
      "step": 12960
    },
    {
      "epoch": 2.406870937790158,
      "grad_norm": 1.5369480035070278,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8807908296585083,
      "num_tokens": 450882567.0,
      "step": 12961
    },
    {
      "epoch": 2.4070566388115133,
      "grad_norm": 1.7317308210255118,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8760026693344116,
      "num_tokens": 450910492.0,
      "step": 12962
    },
    {
      "epoch": 2.407242339832869,
      "grad_norm": 1.4658372423480182,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.8988603353500366,
      "num_tokens": 450944951.0,
      "step": 12963
    },
    {
      "epoch": 2.407428040854225,
      "grad_norm": 1.4952963093301928,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.8985906839370728,
      "num_tokens": 450979686.0,
      "step": 12964
    },
    {
      "epoch": 2.40761374187558,
      "grad_norm": 1.5318752497168633,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8836442232131958,
      "num_tokens": 451017467.0,
      "step": 12965
    },
    {
      "epoch": 2.407799442896936,
      "grad_norm": 1.5473016055630793,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.870280385017395,
      "num_tokens": 451055241.0,
      "step": 12966
    },
    {
      "epoch": 2.4079851439182915,
      "grad_norm": 1.614648547164854,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8834623098373413,
      "num_tokens": 451088781.0,
      "step": 12967
    },
    {
      "epoch": 2.4081708449396473,
      "grad_norm": 1.672861327384737,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.874714195728302,
      "num_tokens": 451123980.0,
      "step": 12968
    },
    {
      "epoch": 2.408356545961003,
      "grad_norm": 1.71012280855246,
      "learning_rate": 1e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8718847632408142,
      "num_tokens": 451155657.0,
      "step": 12969
    },
    {
      "epoch": 2.4085422469823583,
      "grad_norm": 1.620350581441802,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8972093462944031,
      "num_tokens": 451189553.0,
      "step": 12970
    },
    {
      "epoch": 2.408727948003714,
      "grad_norm": 1.6003504633882117,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8942077159881592,
      "num_tokens": 451221541.0,
      "step": 12971
    },
    {
      "epoch": 2.4089136490250698,
      "grad_norm": 1.6715994124856839,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8943381309509277,
      "num_tokens": 451252561.0,
      "step": 12972
    },
    {
      "epoch": 2.409099350046425,
      "grad_norm": 1.410686253011308,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.887508749961853,
      "num_tokens": 451299232.0,
      "step": 12973
    },
    {
      "epoch": 2.409285051067781,
      "grad_norm": 1.4769663580864978,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8916233777999878,
      "num_tokens": 451336493.0,
      "step": 12974
    },
    {
      "epoch": 2.4094707520891365,
      "grad_norm": 1.6383601757158097,
      "learning_rate": 1e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.865703821182251,
      "num_tokens": 451373496.0,
      "step": 12975
    },
    {
      "epoch": 2.4096564531104923,
      "grad_norm": 1.7318480982923214,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8784542083740234,
      "num_tokens": 451406157.0,
      "step": 12976
    },
    {
      "epoch": 2.4098421541318475,
      "grad_norm": 1.6635610163546994,
      "learning_rate": 1e-06,
      "loss": 0.2799,
      "mean_token_accuracy": 0.9007920622825623,
      "num_tokens": 451437063.0,
      "step": 12977
    },
    {
      "epoch": 2.4100278551532033,
      "grad_norm": 1.7640380631668113,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8773127794265747,
      "num_tokens": 451468038.0,
      "step": 12978
    },
    {
      "epoch": 2.410213556174559,
      "grad_norm": 1.6965519025712186,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8842224478721619,
      "num_tokens": 451498363.0,
      "step": 12979
    },
    {
      "epoch": 2.4103992571959147,
      "grad_norm": 1.6897517141130476,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8712916374206543,
      "num_tokens": 451532898.0,
      "step": 12980
    },
    {
      "epoch": 2.41058495821727,
      "grad_norm": 1.509823842135553,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8839238286018372,
      "num_tokens": 451569392.0,
      "step": 12981
    },
    {
      "epoch": 2.4107706592386258,
      "grad_norm": 1.5686771891979638,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8879559636116028,
      "num_tokens": 451604943.0,
      "step": 12982
    },
    {
      "epoch": 2.4109563602599815,
      "grad_norm": 1.5155528982734872,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8900464773178101,
      "num_tokens": 451641365.0,
      "step": 12983
    },
    {
      "epoch": 2.4111420612813372,
      "grad_norm": 1.6451867415704529,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8836727142333984,
      "num_tokens": 451673090.0,
      "step": 12984
    },
    {
      "epoch": 2.4113277623026925,
      "grad_norm": 1.6803322298517054,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8873953819274902,
      "num_tokens": 451699708.0,
      "step": 12985
    },
    {
      "epoch": 2.4115134633240483,
      "grad_norm": 1.5598647489574542,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8873744010925293,
      "num_tokens": 451733822.0,
      "step": 12986
    },
    {
      "epoch": 2.411699164345404,
      "grad_norm": 1.6889126328592947,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8725833296775818,
      "num_tokens": 451769395.0,
      "step": 12987
    },
    {
      "epoch": 2.4118848653667593,
      "grad_norm": 1.6252952218266423,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8927663564682007,
      "num_tokens": 451803899.0,
      "step": 12988
    },
    {
      "epoch": 2.412070566388115,
      "grad_norm": 1.694741519066372,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8901705145835876,
      "num_tokens": 451838122.0,
      "step": 12989
    },
    {
      "epoch": 2.4122562674094707,
      "grad_norm": 1.5504425879894301,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8713799715042114,
      "num_tokens": 451877740.0,
      "step": 12990
    },
    {
      "epoch": 2.4124419684308265,
      "grad_norm": 1.5083408358919417,
      "learning_rate": 1e-06,
      "loss": 0.2849,
      "mean_token_accuracy": 0.900962233543396,
      "num_tokens": 451911750.0,
      "step": 12991
    },
    {
      "epoch": 2.412627669452182,
      "grad_norm": 1.6034286424560482,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8836430907249451,
      "num_tokens": 451946387.0,
      "step": 12992
    },
    {
      "epoch": 2.4128133704735375,
      "grad_norm": 1.6082001877420857,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.895618736743927,
      "num_tokens": 451980671.0,
      "step": 12993
    },
    {
      "epoch": 2.4129990714948932,
      "grad_norm": 1.6058847437907704,
      "learning_rate": 1e-06,
      "loss": 0.2927,
      "mean_token_accuracy": 0.8947322368621826,
      "num_tokens": 452013594.0,
      "step": 12994
    },
    {
      "epoch": 2.413184772516249,
      "grad_norm": 1.6503197242654248,
      "learning_rate": 1e-06,
      "loss": 0.3574,
      "mean_token_accuracy": 0.878631055355072,
      "num_tokens": 452050348.0,
      "step": 12995
    },
    {
      "epoch": 2.4133704735376043,
      "grad_norm": 1.6348352388280123,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8710964918136597,
      "num_tokens": 452086149.0,
      "step": 12996
    },
    {
      "epoch": 2.41355617455896,
      "grad_norm": 1.5284037463347289,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8940707445144653,
      "num_tokens": 452124521.0,
      "step": 12997
    },
    {
      "epoch": 2.4137418755803157,
      "grad_norm": 1.7314323499854265,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8801226615905762,
      "num_tokens": 452157072.0,
      "step": 12998
    },
    {
      "epoch": 2.4139275766016715,
      "grad_norm": 1.5102227101628327,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8812645077705383,
      "num_tokens": 452195623.0,
      "step": 12999
    },
    {
      "epoch": 2.4141132776230267,
      "grad_norm": 1.6043925234252985,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8944551944732666,
      "num_tokens": 452229159.0,
      "step": 13000
    },
    {
      "epoch": 2.4142989786443825,
      "grad_norm": 1.4805598326083027,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8848024010658264,
      "num_tokens": 452267861.0,
      "step": 13001
    },
    {
      "epoch": 2.414484679665738,
      "grad_norm": 1.6835490506985193,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8886072635650635,
      "num_tokens": 452297176.0,
      "step": 13002
    },
    {
      "epoch": 2.414670380687094,
      "grad_norm": 1.4838420018401923,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8906484842300415,
      "num_tokens": 452333184.0,
      "step": 13003
    },
    {
      "epoch": 2.4148560817084492,
      "grad_norm": 1.497816857831778,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8863073587417603,
      "num_tokens": 452368631.0,
      "step": 13004
    },
    {
      "epoch": 2.415041782729805,
      "grad_norm": 1.5087983330345505,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8931928277015686,
      "num_tokens": 452402704.0,
      "step": 13005
    },
    {
      "epoch": 2.4152274837511607,
      "grad_norm": 1.5317703040347537,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.8990800380706787,
      "num_tokens": 452435224.0,
      "step": 13006
    },
    {
      "epoch": 2.4154131847725164,
      "grad_norm": 1.5654561527607378,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8896667957305908,
      "num_tokens": 452469090.0,
      "step": 13007
    },
    {
      "epoch": 2.4155988857938717,
      "grad_norm": 1.6423507858192354,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8822721838951111,
      "num_tokens": 452502038.0,
      "step": 13008
    },
    {
      "epoch": 2.4157845868152275,
      "grad_norm": 1.4932312431162265,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8888789415359497,
      "num_tokens": 452540384.0,
      "step": 13009
    },
    {
      "epoch": 2.415970287836583,
      "grad_norm": 1.9260244722588615,
      "learning_rate": 1e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8757740259170532,
      "num_tokens": 452567667.0,
      "step": 13010
    },
    {
      "epoch": 2.4161559888579385,
      "grad_norm": 1.57923520975527,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8853828310966492,
      "num_tokens": 452600104.0,
      "step": 13011
    },
    {
      "epoch": 2.416341689879294,
      "grad_norm": 1.544396879666808,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8901385068893433,
      "num_tokens": 452637015.0,
      "step": 13012
    },
    {
      "epoch": 2.41652739090065,
      "grad_norm": 1.4381308261659667,
      "learning_rate": 1e-06,
      "loss": 0.2757,
      "mean_token_accuracy": 0.9011648893356323,
      "num_tokens": 452674325.0,
      "step": 13013
    },
    {
      "epoch": 2.4167130919220057,
      "grad_norm": 1.5770184093114477,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8799006938934326,
      "num_tokens": 452711853.0,
      "step": 13014
    },
    {
      "epoch": 2.4168987929433614,
      "grad_norm": 1.5936974895900748,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8719466924667358,
      "num_tokens": 452747674.0,
      "step": 13015
    },
    {
      "epoch": 2.4170844939647167,
      "grad_norm": 1.5941242113678833,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.8952744603157043,
      "num_tokens": 452780840.0,
      "step": 13016
    },
    {
      "epoch": 2.4172701949860724,
      "grad_norm": 1.5754806788556524,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8764501810073853,
      "num_tokens": 452818645.0,
      "step": 13017
    },
    {
      "epoch": 2.417455896007428,
      "grad_norm": 1.5067050667653867,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8892381191253662,
      "num_tokens": 452854692.0,
      "step": 13018
    },
    {
      "epoch": 2.4176415970287835,
      "grad_norm": 1.5391704741312384,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8846929669380188,
      "num_tokens": 452890495.0,
      "step": 13019
    },
    {
      "epoch": 2.417827298050139,
      "grad_norm": 1.6943222102241207,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8784471750259399,
      "num_tokens": 452921722.0,
      "step": 13020
    },
    {
      "epoch": 2.418012999071495,
      "grad_norm": 1.518305163971296,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8883812427520752,
      "num_tokens": 452960120.0,
      "step": 13021
    },
    {
      "epoch": 2.4181987000928507,
      "grad_norm": 1.552615091207874,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8820924162864685,
      "num_tokens": 453001843.0,
      "step": 13022
    },
    {
      "epoch": 2.418384401114206,
      "grad_norm": 1.698709582773104,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8837767839431763,
      "num_tokens": 453031979.0,
      "step": 13023
    },
    {
      "epoch": 2.4185701021355617,
      "grad_norm": 1.4869044371677738,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8862079381942749,
      "num_tokens": 453069113.0,
      "step": 13024
    },
    {
      "epoch": 2.4187558031569174,
      "grad_norm": 1.6518716449582216,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8851162791252136,
      "num_tokens": 453102856.0,
      "step": 13025
    },
    {
      "epoch": 2.418941504178273,
      "grad_norm": 1.6304428484538491,
      "learning_rate": 1e-06,
      "loss": 0.2962,
      "mean_token_accuracy": 0.8950489163398743,
      "num_tokens": 453133120.0,
      "step": 13026
    },
    {
      "epoch": 2.4191272051996284,
      "grad_norm": 1.5445257764327154,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8831897974014282,
      "num_tokens": 453170583.0,
      "step": 13027
    },
    {
      "epoch": 2.419312906220984,
      "grad_norm": 1.559967550929129,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8821899890899658,
      "num_tokens": 453204678.0,
      "step": 13028
    },
    {
      "epoch": 2.41949860724234,
      "grad_norm": 1.5418437956788225,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8836221098899841,
      "num_tokens": 453242274.0,
      "step": 13029
    },
    {
      "epoch": 2.4196843082636956,
      "grad_norm": 1.6048136451363688,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8869032859802246,
      "num_tokens": 453277894.0,
      "step": 13030
    },
    {
      "epoch": 2.419870009285051,
      "grad_norm": 1.6946673381188906,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8724238872528076,
      "num_tokens": 453311699.0,
      "step": 13031
    },
    {
      "epoch": 2.4200557103064066,
      "grad_norm": 1.733086161388187,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.9014847278594971,
      "num_tokens": 453338855.0,
      "step": 13032
    },
    {
      "epoch": 2.4202414113277624,
      "grad_norm": 1.5530184654528179,
      "learning_rate": 1e-06,
      "loss": 0.299,
      "mean_token_accuracy": 0.8952300548553467,
      "num_tokens": 453372131.0,
      "step": 13033
    },
    {
      "epoch": 2.4204271123491177,
      "grad_norm": 1.6639798363467992,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8862420320510864,
      "num_tokens": 453402089.0,
      "step": 13034
    },
    {
      "epoch": 2.4206128133704734,
      "grad_norm": 1.48303691554871,
      "learning_rate": 1e-06,
      "loss": 0.2973,
      "mean_token_accuracy": 0.8955307602882385,
      "num_tokens": 453438611.0,
      "step": 13035
    },
    {
      "epoch": 2.420798514391829,
      "grad_norm": 1.6352460721053068,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8931995630264282,
      "num_tokens": 453468088.0,
      "step": 13036
    },
    {
      "epoch": 2.420984215413185,
      "grad_norm": 1.6440852763308442,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8942829370498657,
      "num_tokens": 453499457.0,
      "step": 13037
    },
    {
      "epoch": 2.4211699164345406,
      "grad_norm": 1.5419837689268188,
      "learning_rate": 1e-06,
      "loss": 0.3022,
      "mean_token_accuracy": 0.8926951885223389,
      "num_tokens": 453533144.0,
      "step": 13038
    },
    {
      "epoch": 2.421355617455896,
      "grad_norm": 1.8447783825361599,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8787950277328491,
      "num_tokens": 453561918.0,
      "step": 13039
    },
    {
      "epoch": 2.4215413184772516,
      "grad_norm": 1.5852857151472186,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8935452699661255,
      "num_tokens": 453592834.0,
      "step": 13040
    },
    {
      "epoch": 2.4217270194986074,
      "grad_norm": 1.7169999389909587,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8817936778068542,
      "num_tokens": 453624004.0,
      "step": 13041
    },
    {
      "epoch": 2.4219127205199626,
      "grad_norm": 1.5560180291218222,
      "learning_rate": 1e-06,
      "loss": 0.286,
      "mean_token_accuracy": 0.8990631103515625,
      "num_tokens": 453659594.0,
      "step": 13042
    },
    {
      "epoch": 2.4220984215413184,
      "grad_norm": 1.637420918602251,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8771909475326538,
      "num_tokens": 453692286.0,
      "step": 13043
    },
    {
      "epoch": 2.422284122562674,
      "grad_norm": 1.4843462688370719,
      "learning_rate": 1e-06,
      "loss": 0.3255,
      "mean_token_accuracy": 0.8863028883934021,
      "num_tokens": 453733176.0,
      "step": 13044
    },
    {
      "epoch": 2.42246982358403,
      "grad_norm": 1.5638445293931962,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.880089521408081,
      "num_tokens": 453766587.0,
      "step": 13045
    },
    {
      "epoch": 2.422655524605385,
      "grad_norm": 1.455100654027162,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8905695676803589,
      "num_tokens": 453803268.0,
      "step": 13046
    },
    {
      "epoch": 2.422841225626741,
      "grad_norm": 1.499638622138783,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8854007720947266,
      "num_tokens": 453841516.0,
      "step": 13047
    },
    {
      "epoch": 2.4230269266480966,
      "grad_norm": 1.6112514387512504,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8768718242645264,
      "num_tokens": 453874110.0,
      "step": 13048
    },
    {
      "epoch": 2.4232126276694523,
      "grad_norm": 1.5762819518064408,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8743931651115417,
      "num_tokens": 453908600.0,
      "step": 13049
    },
    {
      "epoch": 2.4233983286908076,
      "grad_norm": 1.501630380269821,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8848745226860046,
      "num_tokens": 453947704.0,
      "step": 13050
    },
    {
      "epoch": 2.4235840297121634,
      "grad_norm": 1.5540109754888343,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8932181596755981,
      "num_tokens": 453981220.0,
      "step": 13051
    },
    {
      "epoch": 2.423769730733519,
      "grad_norm": 1.7211570274466277,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8872107267379761,
      "num_tokens": 454011498.0,
      "step": 13052
    },
    {
      "epoch": 2.423955431754875,
      "grad_norm": 1.5394600392147417,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8897081613540649,
      "num_tokens": 454046841.0,
      "step": 13053
    },
    {
      "epoch": 2.42414113277623,
      "grad_norm": 1.6471175856638278,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8910967111587524,
      "num_tokens": 454077438.0,
      "step": 13054
    },
    {
      "epoch": 2.424326833797586,
      "grad_norm": 1.6232963895125372,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.8824727535247803,
      "num_tokens": 454116611.0,
      "step": 13055
    },
    {
      "epoch": 2.4245125348189416,
      "grad_norm": 1.4932027158915437,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8802596926689148,
      "num_tokens": 454159469.0,
      "step": 13056
    },
    {
      "epoch": 2.4246982358402973,
      "grad_norm": 1.6311396196802608,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8836753368377686,
      "num_tokens": 454190720.0,
      "step": 13057
    },
    {
      "epoch": 2.4248839368616526,
      "grad_norm": 1.621676880727144,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.8729636073112488,
      "num_tokens": 454228125.0,
      "step": 13058
    },
    {
      "epoch": 2.4250696378830083,
      "grad_norm": 1.5725770828855115,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8905232548713684,
      "num_tokens": 454262376.0,
      "step": 13059
    },
    {
      "epoch": 2.425255338904364,
      "grad_norm": 1.4300810197221903,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8946000337600708,
      "num_tokens": 454300944.0,
      "step": 13060
    },
    {
      "epoch": 2.42544103992572,
      "grad_norm": 1.477254659057689,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8892369270324707,
      "num_tokens": 454340771.0,
      "step": 13061
    },
    {
      "epoch": 2.425626740947075,
      "grad_norm": 1.5151710250306003,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8772480487823486,
      "num_tokens": 454379394.0,
      "step": 13062
    },
    {
      "epoch": 2.425812441968431,
      "grad_norm": 1.6165198573945012,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8872936964035034,
      "num_tokens": 454410684.0,
      "step": 13063
    },
    {
      "epoch": 2.4259981429897866,
      "grad_norm": 1.529041385777445,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8904205560684204,
      "num_tokens": 454445261.0,
      "step": 13064
    },
    {
      "epoch": 2.426183844011142,
      "grad_norm": 1.6226017631031013,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.8696261644363403,
      "num_tokens": 454480245.0,
      "step": 13065
    },
    {
      "epoch": 2.4263695450324976,
      "grad_norm": 2.5024481812642816,
      "learning_rate": 1e-06,
      "loss": 0.2864,
      "mean_token_accuracy": 0.8962670564651489,
      "num_tokens": 454512126.0,
      "step": 13066
    },
    {
      "epoch": 2.4265552460538533,
      "grad_norm": 1.5946168566782501,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8884608745574951,
      "num_tokens": 454544335.0,
      "step": 13067
    },
    {
      "epoch": 2.426740947075209,
      "grad_norm": 1.5480017065800353,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.890609860420227,
      "num_tokens": 454580912.0,
      "step": 13068
    },
    {
      "epoch": 2.4269266480965648,
      "grad_norm": 1.6748348575320484,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8835970759391785,
      "num_tokens": 454611813.0,
      "step": 13069
    },
    {
      "epoch": 2.42711234911792,
      "grad_norm": 1.566111827374133,
      "learning_rate": 1e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8733941316604614,
      "num_tokens": 454648380.0,
      "step": 13070
    },
    {
      "epoch": 2.427298050139276,
      "grad_norm": 1.6225140986367572,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8895924091339111,
      "num_tokens": 454679679.0,
      "step": 13071
    },
    {
      "epoch": 2.4274837511606315,
      "grad_norm": 1.6667465956062517,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8888951539993286,
      "num_tokens": 454710658.0,
      "step": 13072
    },
    {
      "epoch": 2.427669452181987,
      "grad_norm": 1.4775954851003892,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.895218014717102,
      "num_tokens": 454747495.0,
      "step": 13073
    },
    {
      "epoch": 2.4278551532033426,
      "grad_norm": 1.4833401187874595,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8865340948104858,
      "num_tokens": 454785225.0,
      "step": 13074
    },
    {
      "epoch": 2.4280408542246983,
      "grad_norm": 1.561100706918912,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.8958595395088196,
      "num_tokens": 454825758.0,
      "step": 13075
    },
    {
      "epoch": 2.428226555246054,
      "grad_norm": 1.8352779931101901,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8767902255058289,
      "num_tokens": 454853880.0,
      "step": 13076
    },
    {
      "epoch": 2.4284122562674093,
      "grad_norm": 1.6116765565100688,
      "learning_rate": 1e-06,
      "loss": 0.3655,
      "mean_token_accuracy": 0.8723170161247253,
      "num_tokens": 454887663.0,
      "step": 13077
    },
    {
      "epoch": 2.428597957288765,
      "grad_norm": 1.5873575590770064,
      "learning_rate": 1e-06,
      "loss": 0.3816,
      "mean_token_accuracy": 0.8676109910011292,
      "num_tokens": 454925489.0,
      "step": 13078
    },
    {
      "epoch": 2.4287836583101208,
      "grad_norm": 1.492678907982835,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8802695274353027,
      "num_tokens": 454965120.0,
      "step": 13079
    },
    {
      "epoch": 2.4289693593314765,
      "grad_norm": 1.5642345014888028,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8816609382629395,
      "num_tokens": 454999329.0,
      "step": 13080
    },
    {
      "epoch": 2.429155060352832,
      "grad_norm": 1.5092507751065634,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.8947156071662903,
      "num_tokens": 455032809.0,
      "step": 13081
    },
    {
      "epoch": 2.4293407613741875,
      "grad_norm": 1.5872327173383862,
      "learning_rate": 1e-06,
      "loss": 0.3638,
      "mean_token_accuracy": 0.8764315843582153,
      "num_tokens": 455070749.0,
      "step": 13082
    },
    {
      "epoch": 2.4295264623955433,
      "grad_norm": 1.5130645860781402,
      "learning_rate": 1e-06,
      "loss": 0.2865,
      "mean_token_accuracy": 0.8970893025398254,
      "num_tokens": 455103972.0,
      "step": 13083
    },
    {
      "epoch": 2.429712163416899,
      "grad_norm": 1.5763383155293555,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8828192949295044,
      "num_tokens": 455141966.0,
      "step": 13084
    },
    {
      "epoch": 2.4298978644382543,
      "grad_norm": 1.434254421241899,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8711937069892883,
      "num_tokens": 455189208.0,
      "step": 13085
    },
    {
      "epoch": 2.43008356545961,
      "grad_norm": 1.5803999680728933,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.887030303478241,
      "num_tokens": 455224243.0,
      "step": 13086
    },
    {
      "epoch": 2.4302692664809658,
      "grad_norm": 1.642531536979537,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.888574481010437,
      "num_tokens": 455259179.0,
      "step": 13087
    },
    {
      "epoch": 2.430454967502321,
      "grad_norm": 1.5595971650710831,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8838748931884766,
      "num_tokens": 455292422.0,
      "step": 13088
    },
    {
      "epoch": 2.4306406685236768,
      "grad_norm": 1.4818542233806613,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8974753022193909,
      "num_tokens": 455329667.0,
      "step": 13089
    },
    {
      "epoch": 2.4308263695450325,
      "grad_norm": 1.685863381182347,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8772474527359009,
      "num_tokens": 455364401.0,
      "step": 13090
    },
    {
      "epoch": 2.4310120705663882,
      "grad_norm": 1.5384974492254513,
      "learning_rate": 1e-06,
      "loss": 0.2977,
      "mean_token_accuracy": 0.8944442868232727,
      "num_tokens": 455400543.0,
      "step": 13091
    },
    {
      "epoch": 2.431197771587744,
      "grad_norm": 1.7787068792248517,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8770252466201782,
      "num_tokens": 455433316.0,
      "step": 13092
    },
    {
      "epoch": 2.4313834726090993,
      "grad_norm": 1.9250737189122649,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.884121298789978,
      "num_tokens": 455471695.0,
      "step": 13093
    },
    {
      "epoch": 2.431569173630455,
      "grad_norm": 1.6779068639530237,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8888299465179443,
      "num_tokens": 455501470.0,
      "step": 13094
    },
    {
      "epoch": 2.4317548746518107,
      "grad_norm": 1.5283773957931905,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8953373432159424,
      "num_tokens": 455536294.0,
      "step": 13095
    },
    {
      "epoch": 2.431940575673166,
      "grad_norm": 1.536042158728663,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8876791000366211,
      "num_tokens": 455572521.0,
      "step": 13096
    },
    {
      "epoch": 2.4321262766945217,
      "grad_norm": 1.5491595131131313,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8909173607826233,
      "num_tokens": 455607648.0,
      "step": 13097
    },
    {
      "epoch": 2.4323119777158775,
      "grad_norm": 1.5546035264730194,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8783167600631714,
      "num_tokens": 455650012.0,
      "step": 13098
    },
    {
      "epoch": 2.432497678737233,
      "grad_norm": 1.6531214952932516,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8990219831466675,
      "num_tokens": 455679224.0,
      "step": 13099
    },
    {
      "epoch": 2.4326833797585885,
      "grad_norm": 1.8367355676029322,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8824079036712646,
      "num_tokens": 455709274.0,
      "step": 13100
    },
    {
      "epoch": 2.4328690807799442,
      "grad_norm": 1.5969569065635731,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.902825117111206,
      "num_tokens": 455738309.0,
      "step": 13101
    },
    {
      "epoch": 2.4330547818013,
      "grad_norm": 1.438746373978506,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8979641199111938,
      "num_tokens": 455776792.0,
      "step": 13102
    },
    {
      "epoch": 2.4332404828226557,
      "grad_norm": 1.5352482040707314,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.8882692456245422,
      "num_tokens": 455813358.0,
      "step": 13103
    },
    {
      "epoch": 2.433426183844011,
      "grad_norm": 1.730655765149306,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8798556327819824,
      "num_tokens": 455845787.0,
      "step": 13104
    },
    {
      "epoch": 2.4336118848653667,
      "grad_norm": 1.5728398499369054,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8920658826828003,
      "num_tokens": 455878434.0,
      "step": 13105
    },
    {
      "epoch": 2.4337975858867225,
      "grad_norm": 1.5810457269015858,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8849019408226013,
      "num_tokens": 455918193.0,
      "step": 13106
    },
    {
      "epoch": 2.433983286908078,
      "grad_norm": 1.5474082517160996,
      "learning_rate": 1e-06,
      "loss": 0.3035,
      "mean_token_accuracy": 0.895035445690155,
      "num_tokens": 455952334.0,
      "step": 13107
    },
    {
      "epoch": 2.4341689879294335,
      "grad_norm": 1.5794905580169298,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.8971513509750366,
      "num_tokens": 455983442.0,
      "step": 13108
    },
    {
      "epoch": 2.434354688950789,
      "grad_norm": 1.6410269648969753,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8884274959564209,
      "num_tokens": 456017328.0,
      "step": 13109
    },
    {
      "epoch": 2.434540389972145,
      "grad_norm": 1.654695043100453,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8835153579711914,
      "num_tokens": 456049108.0,
      "step": 13110
    },
    {
      "epoch": 2.4347260909935002,
      "grad_norm": 1.5759190017920859,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8856168985366821,
      "num_tokens": 456085678.0,
      "step": 13111
    },
    {
      "epoch": 2.434911792014856,
      "grad_norm": 1.593560895848285,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8748712539672852,
      "num_tokens": 456121895.0,
      "step": 13112
    },
    {
      "epoch": 2.4350974930362117,
      "grad_norm": 1.6637229336910375,
      "learning_rate": 1e-06,
      "loss": 0.3193,
      "mean_token_accuracy": 0.8884661197662354,
      "num_tokens": 456155351.0,
      "step": 13113
    },
    {
      "epoch": 2.4352831940575674,
      "grad_norm": 1.6044894732152177,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8814744353294373,
      "num_tokens": 456192264.0,
      "step": 13114
    },
    {
      "epoch": 2.435468895078923,
      "grad_norm": 1.573592322039521,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8856978416442871,
      "num_tokens": 456225031.0,
      "step": 13115
    },
    {
      "epoch": 2.4356545961002785,
      "grad_norm": 1.6155821215686372,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.883018970489502,
      "num_tokens": 456259391.0,
      "step": 13116
    },
    {
      "epoch": 2.435840297121634,
      "grad_norm": 1.659145711563912,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8851714134216309,
      "num_tokens": 456290874.0,
      "step": 13117
    },
    {
      "epoch": 2.43602599814299,
      "grad_norm": 1.6473814663147923,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8803386688232422,
      "num_tokens": 456324923.0,
      "step": 13118
    },
    {
      "epoch": 2.436211699164345,
      "grad_norm": 1.61729826693462,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.885944128036499,
      "num_tokens": 456359192.0,
      "step": 13119
    },
    {
      "epoch": 2.436397400185701,
      "grad_norm": 1.5600311871824826,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8827966451644897,
      "num_tokens": 456394150.0,
      "step": 13120
    },
    {
      "epoch": 2.4365831012070567,
      "grad_norm": 1.5736846194889516,
      "learning_rate": 1e-06,
      "loss": 0.3084,
      "mean_token_accuracy": 0.8917669653892517,
      "num_tokens": 456428388.0,
      "step": 13121
    },
    {
      "epoch": 2.4367688022284124,
      "grad_norm": 1.5423406084750615,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8787355422973633,
      "num_tokens": 456467011.0,
      "step": 13122
    },
    {
      "epoch": 2.4369545032497677,
      "grad_norm": 1.4059391960297172,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8862254619598389,
      "num_tokens": 456511150.0,
      "step": 13123
    },
    {
      "epoch": 2.4371402042711234,
      "grad_norm": 1.4703941501868332,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.882236123085022,
      "num_tokens": 456552740.0,
      "step": 13124
    },
    {
      "epoch": 2.437325905292479,
      "grad_norm": 1.695339074135182,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8930479288101196,
      "num_tokens": 456581462.0,
      "step": 13125
    },
    {
      "epoch": 2.437511606313835,
      "grad_norm": 1.5792292858635448,
      "learning_rate": 1e-06,
      "loss": 0.3042,
      "mean_token_accuracy": 0.8882567286491394,
      "num_tokens": 456616011.0,
      "step": 13126
    },
    {
      "epoch": 2.43769730733519,
      "grad_norm": 1.7643353801390818,
      "learning_rate": 1e-06,
      "loss": 0.2855,
      "mean_token_accuracy": 0.8963989019393921,
      "num_tokens": 456644942.0,
      "step": 13127
    },
    {
      "epoch": 2.437883008356546,
      "grad_norm": 1.6326762003186135,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.893231987953186,
      "num_tokens": 456674185.0,
      "step": 13128
    },
    {
      "epoch": 2.4380687093779017,
      "grad_norm": 1.6330424336333058,
      "learning_rate": 1e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8723258972167969,
      "num_tokens": 456709502.0,
      "step": 13129
    },
    {
      "epoch": 2.4382544103992574,
      "grad_norm": 1.4948775050488154,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.892279326915741,
      "num_tokens": 456747189.0,
      "step": 13130
    },
    {
      "epoch": 2.4384401114206127,
      "grad_norm": 1.8449770846242572,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8706753849983215,
      "num_tokens": 456778178.0,
      "step": 13131
    },
    {
      "epoch": 2.4386258124419684,
      "grad_norm": 1.6499819525010648,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.881493866443634,
      "num_tokens": 456816761.0,
      "step": 13132
    },
    {
      "epoch": 2.438811513463324,
      "grad_norm": 1.744367154900273,
      "learning_rate": 1e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8694770932197571,
      "num_tokens": 456850897.0,
      "step": 13133
    },
    {
      "epoch": 2.4389972144846794,
      "grad_norm": 1.6073634233362868,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.882987380027771,
      "num_tokens": 456884266.0,
      "step": 13134
    },
    {
      "epoch": 2.439182915506035,
      "grad_norm": 1.563830365955585,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8867605924606323,
      "num_tokens": 456923721.0,
      "step": 13135
    },
    {
      "epoch": 2.439368616527391,
      "grad_norm": 1.6354316970793865,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8920385837554932,
      "num_tokens": 456953865.0,
      "step": 13136
    },
    {
      "epoch": 2.4395543175487466,
      "grad_norm": 1.4876988934693063,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8915749788284302,
      "num_tokens": 456993215.0,
      "step": 13137
    },
    {
      "epoch": 2.4397400185701024,
      "grad_norm": 1.5271393236173172,
      "learning_rate": 1e-06,
      "loss": 0.2956,
      "mean_token_accuracy": 0.8932201862335205,
      "num_tokens": 457029310.0,
      "step": 13138
    },
    {
      "epoch": 2.4399257195914577,
      "grad_norm": 1.5727770009915627,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8805921673774719,
      "num_tokens": 457063594.0,
      "step": 13139
    },
    {
      "epoch": 2.4401114206128134,
      "grad_norm": 1.7283632917498135,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8802473545074463,
      "num_tokens": 457093403.0,
      "step": 13140
    },
    {
      "epoch": 2.440297121634169,
      "grad_norm": 1.3756601224007152,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8844447135925293,
      "num_tokens": 457139148.0,
      "step": 13141
    },
    {
      "epoch": 2.4404828226555244,
      "grad_norm": 1.6419393810387082,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8945592641830444,
      "num_tokens": 457169557.0,
      "step": 13142
    },
    {
      "epoch": 2.44066852367688,
      "grad_norm": 1.5925934434119415,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8945661783218384,
      "num_tokens": 457203469.0,
      "step": 13143
    },
    {
      "epoch": 2.440854224698236,
      "grad_norm": 1.7805601877311943,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8816683888435364,
      "num_tokens": 457231995.0,
      "step": 13144
    },
    {
      "epoch": 2.4410399257195916,
      "grad_norm": 1.4973548139932193,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8888040781021118,
      "num_tokens": 457270185.0,
      "step": 13145
    },
    {
      "epoch": 2.441225626740947,
      "grad_norm": 1.618761742361098,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8741747140884399,
      "num_tokens": 457307782.0,
      "step": 13146
    },
    {
      "epoch": 2.4414113277623026,
      "grad_norm": 1.5781025133943538,
      "learning_rate": 1e-06,
      "loss": 0.267,
      "mean_token_accuracy": 0.904773473739624,
      "num_tokens": 457344913.0,
      "step": 13147
    },
    {
      "epoch": 2.4415970287836584,
      "grad_norm": 1.6100207529507184,
      "learning_rate": 1e-06,
      "loss": 0.2831,
      "mean_token_accuracy": 0.9011015892028809,
      "num_tokens": 457376822.0,
      "step": 13148
    },
    {
      "epoch": 2.441782729805014,
      "grad_norm": 1.5492792074943529,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8854660391807556,
      "num_tokens": 457413033.0,
      "step": 13149
    },
    {
      "epoch": 2.4419684308263694,
      "grad_norm": 1.6904496511547118,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8898322582244873,
      "num_tokens": 457444700.0,
      "step": 13150
    },
    {
      "epoch": 2.442154131847725,
      "grad_norm": 1.574453633220242,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8825180530548096,
      "num_tokens": 457476662.0,
      "step": 13151
    },
    {
      "epoch": 2.442339832869081,
      "grad_norm": 1.7239654670364752,
      "learning_rate": 1e-06,
      "loss": 0.3432,
      "mean_token_accuracy": 0.8825051188468933,
      "num_tokens": 457506436.0,
      "step": 13152
    },
    {
      "epoch": 2.4425255338904366,
      "grad_norm": 1.43428839344755,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8852429986000061,
      "num_tokens": 457545751.0,
      "step": 13153
    },
    {
      "epoch": 2.442711234911792,
      "grad_norm": 1.5317683063374221,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8916443586349487,
      "num_tokens": 457580399.0,
      "step": 13154
    },
    {
      "epoch": 2.4428969359331476,
      "grad_norm": 1.6262955739342457,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8914724588394165,
      "num_tokens": 457610714.0,
      "step": 13155
    },
    {
      "epoch": 2.4430826369545033,
      "grad_norm": 1.5382688913025788,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.8767743110656738,
      "num_tokens": 457648796.0,
      "step": 13156
    },
    {
      "epoch": 2.4432683379758586,
      "grad_norm": 1.5201039746534017,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8826085329055786,
      "num_tokens": 457690060.0,
      "step": 13157
    },
    {
      "epoch": 2.4434540389972144,
      "grad_norm": 1.6346030109954943,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8866521120071411,
      "num_tokens": 457727855.0,
      "step": 13158
    },
    {
      "epoch": 2.44363974001857,
      "grad_norm": 1.5861826948786475,
      "learning_rate": 1e-06,
      "loss": 0.296,
      "mean_token_accuracy": 0.8926137089729309,
      "num_tokens": 457759257.0,
      "step": 13159
    },
    {
      "epoch": 2.443825441039926,
      "grad_norm": 1.5712216512523098,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8814119696617126,
      "num_tokens": 457795352.0,
      "step": 13160
    },
    {
      "epoch": 2.4440111420612816,
      "grad_norm": 1.625797693101364,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8773897290229797,
      "num_tokens": 457831955.0,
      "step": 13161
    },
    {
      "epoch": 2.444196843082637,
      "grad_norm": 1.8512277786645226,
      "learning_rate": 1e-06,
      "loss": 0.3335,
      "mean_token_accuracy": 0.8821984529495239,
      "num_tokens": 457860542.0,
      "step": 13162
    },
    {
      "epoch": 2.4443825441039926,
      "grad_norm": 1.5932758200860342,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8868787288665771,
      "num_tokens": 457897679.0,
      "step": 13163
    },
    {
      "epoch": 2.4445682451253483,
      "grad_norm": 1.774581605047062,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8813437819480896,
      "num_tokens": 457925548.0,
      "step": 13164
    },
    {
      "epoch": 2.4447539461467036,
      "grad_norm": 1.64073613204311,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8816424608230591,
      "num_tokens": 457957599.0,
      "step": 13165
    },
    {
      "epoch": 2.4449396471680593,
      "grad_norm": 1.6303542189939766,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8765196204185486,
      "num_tokens": 457991212.0,
      "step": 13166
    },
    {
      "epoch": 2.445125348189415,
      "grad_norm": 1.592020113216935,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8798189163208008,
      "num_tokens": 458030596.0,
      "step": 13167
    },
    {
      "epoch": 2.445311049210771,
      "grad_norm": 1.663003971632491,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8908063769340515,
      "num_tokens": 458062463.0,
      "step": 13168
    },
    {
      "epoch": 2.445496750232126,
      "grad_norm": 1.4765382289460853,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8961158394813538,
      "num_tokens": 458097702.0,
      "step": 13169
    },
    {
      "epoch": 2.445682451253482,
      "grad_norm": 1.5394785990239575,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8975377678871155,
      "num_tokens": 458135392.0,
      "step": 13170
    },
    {
      "epoch": 2.4458681522748376,
      "grad_norm": 1.5971910699884093,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8862457275390625,
      "num_tokens": 458172833.0,
      "step": 13171
    },
    {
      "epoch": 2.4460538532961933,
      "grad_norm": 1.5955646907442522,
      "learning_rate": 1e-06,
      "loss": 0.2934,
      "mean_token_accuracy": 0.8962195515632629,
      "num_tokens": 458205156.0,
      "step": 13172
    },
    {
      "epoch": 2.4462395543175486,
      "grad_norm": 1.4804170500124185,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.8969948291778564,
      "num_tokens": 458240762.0,
      "step": 13173
    },
    {
      "epoch": 2.4464252553389043,
      "grad_norm": 1.4934983646013253,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8811131715774536,
      "num_tokens": 458278056.0,
      "step": 13174
    },
    {
      "epoch": 2.44661095636026,
      "grad_norm": 1.5780446632906213,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.8987689018249512,
      "num_tokens": 458307728.0,
      "step": 13175
    },
    {
      "epoch": 2.446796657381616,
      "grad_norm": 1.4950916950465793,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8906493782997131,
      "num_tokens": 458343808.0,
      "step": 13176
    },
    {
      "epoch": 2.446982358402971,
      "grad_norm": 1.5977250715397855,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8959271907806396,
      "num_tokens": 458376642.0,
      "step": 13177
    },
    {
      "epoch": 2.447168059424327,
      "grad_norm": 1.6357838683008334,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8725361824035645,
      "num_tokens": 458410133.0,
      "step": 13178
    },
    {
      "epoch": 2.4473537604456825,
      "grad_norm": 1.5658731051794788,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8881546258926392,
      "num_tokens": 458446719.0,
      "step": 13179
    },
    {
      "epoch": 2.447539461467038,
      "grad_norm": 1.496286982665815,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8860386610031128,
      "num_tokens": 458484703.0,
      "step": 13180
    },
    {
      "epoch": 2.4477251624883936,
      "grad_norm": 1.521867185903376,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8917884230613708,
      "num_tokens": 458519553.0,
      "step": 13181
    },
    {
      "epoch": 2.4479108635097493,
      "grad_norm": 1.5805164343730664,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8837916851043701,
      "num_tokens": 458550511.0,
      "step": 13182
    },
    {
      "epoch": 2.448096564531105,
      "grad_norm": 1.4773739356968854,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8946154713630676,
      "num_tokens": 458586801.0,
      "step": 13183
    },
    {
      "epoch": 2.4482822655524608,
      "grad_norm": 1.4274685445978588,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8858047723770142,
      "num_tokens": 458628343.0,
      "step": 13184
    },
    {
      "epoch": 2.448467966573816,
      "grad_norm": 1.6469951629261976,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.889268696308136,
      "num_tokens": 458661540.0,
      "step": 13185
    },
    {
      "epoch": 2.448653667595172,
      "grad_norm": 1.6714579486027477,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8795151710510254,
      "num_tokens": 458693818.0,
      "step": 13186
    },
    {
      "epoch": 2.4488393686165275,
      "grad_norm": 1.6280689152367862,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8867760896682739,
      "num_tokens": 458725436.0,
      "step": 13187
    },
    {
      "epoch": 2.449025069637883,
      "grad_norm": 1.5660778566785079,
      "learning_rate": 1e-06,
      "loss": 0.3529,
      "mean_token_accuracy": 0.8749452829360962,
      "num_tokens": 458761909.0,
      "step": 13188
    },
    {
      "epoch": 2.4492107706592385,
      "grad_norm": 1.6217692535873998,
      "learning_rate": 1e-06,
      "loss": 0.2905,
      "mean_token_accuracy": 0.8959929943084717,
      "num_tokens": 458798276.0,
      "step": 13189
    },
    {
      "epoch": 2.4493964716805943,
      "grad_norm": 1.7350381717123085,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8815497756004333,
      "num_tokens": 458831054.0,
      "step": 13190
    },
    {
      "epoch": 2.44958217270195,
      "grad_norm": 1.5533518486132158,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8846259117126465,
      "num_tokens": 458874495.0,
      "step": 13191
    },
    {
      "epoch": 2.4497678737233053,
      "grad_norm": 1.6524570464179582,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8782896995544434,
      "num_tokens": 458909155.0,
      "step": 13192
    },
    {
      "epoch": 2.449953574744661,
      "grad_norm": 1.6377669929586156,
      "learning_rate": 1e-06,
      "loss": 0.3255,
      "mean_token_accuracy": 0.884110152721405,
      "num_tokens": 458939690.0,
      "step": 13193
    },
    {
      "epoch": 2.4501392757660168,
      "grad_norm": 1.5373309880201893,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8980796337127686,
      "num_tokens": 458972088.0,
      "step": 13194
    },
    {
      "epoch": 2.4503249767873725,
      "grad_norm": 1.4387362584943209,
      "learning_rate": 1e-06,
      "loss": 0.2916,
      "mean_token_accuracy": 0.8965519070625305,
      "num_tokens": 459008617.0,
      "step": 13195
    },
    {
      "epoch": 2.4505106778087278,
      "grad_norm": 1.6113855782752617,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8796551823616028,
      "num_tokens": 459046849.0,
      "step": 13196
    },
    {
      "epoch": 2.4506963788300835,
      "grad_norm": 1.6492899440584856,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8841931819915771,
      "num_tokens": 459080016.0,
      "step": 13197
    },
    {
      "epoch": 2.4508820798514392,
      "grad_norm": 1.52799560230817,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8856778144836426,
      "num_tokens": 459115945.0,
      "step": 13198
    },
    {
      "epoch": 2.451067780872795,
      "grad_norm": 1.5558479934198577,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8862423896789551,
      "num_tokens": 459148966.0,
      "step": 13199
    },
    {
      "epoch": 2.4512534818941503,
      "grad_norm": 1.6297701544553165,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8755847215652466,
      "num_tokens": 459183555.0,
      "step": 13200
    },
    {
      "epoch": 2.451439182915506,
      "grad_norm": 1.4523045505522598,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.8926460146903992,
      "num_tokens": 459222150.0,
      "step": 13201
    },
    {
      "epoch": 2.4516248839368617,
      "grad_norm": 1.5191362687588696,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8731575608253479,
      "num_tokens": 459257947.0,
      "step": 13202
    },
    {
      "epoch": 2.451810584958217,
      "grad_norm": 1.638815783369156,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8916462063789368,
      "num_tokens": 459287099.0,
      "step": 13203
    },
    {
      "epoch": 2.4519962859795728,
      "grad_norm": 1.6324736556141082,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8772597908973694,
      "num_tokens": 459321247.0,
      "step": 13204
    },
    {
      "epoch": 2.4521819870009285,
      "grad_norm": 1.4348918463362008,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8947241902351379,
      "num_tokens": 459358471.0,
      "step": 13205
    },
    {
      "epoch": 2.452367688022284,
      "grad_norm": 1.5134573986346245,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8747619390487671,
      "num_tokens": 459396269.0,
      "step": 13206
    },
    {
      "epoch": 2.45255338904364,
      "grad_norm": 1.5091276363014559,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8726820945739746,
      "num_tokens": 459436763.0,
      "step": 13207
    },
    {
      "epoch": 2.4527390900649952,
      "grad_norm": 1.7401475394324157,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8863663077354431,
      "num_tokens": 459464294.0,
      "step": 13208
    },
    {
      "epoch": 2.452924791086351,
      "grad_norm": 1.6552596479461812,
      "learning_rate": 1e-06,
      "loss": 0.2891,
      "mean_token_accuracy": 0.8985249996185303,
      "num_tokens": 459492641.0,
      "step": 13209
    },
    {
      "epoch": 2.4531104921077067,
      "grad_norm": 1.6024661217642593,
      "learning_rate": 1e-06,
      "loss": 0.35,
      "mean_token_accuracy": 0.8768690228462219,
      "num_tokens": 459525391.0,
      "step": 13210
    },
    {
      "epoch": 2.453296193129062,
      "grad_norm": 1.5435921672619215,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8929651975631714,
      "num_tokens": 459560853.0,
      "step": 13211
    },
    {
      "epoch": 2.4534818941504177,
      "grad_norm": 1.5743369290234477,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.876032292842865,
      "num_tokens": 459593788.0,
      "step": 13212
    },
    {
      "epoch": 2.4536675951717735,
      "grad_norm": 1.4833202152663594,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8846020698547363,
      "num_tokens": 459635150.0,
      "step": 13213
    },
    {
      "epoch": 2.453853296193129,
      "grad_norm": 1.6732720878077094,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8832820653915405,
      "num_tokens": 459667090.0,
      "step": 13214
    },
    {
      "epoch": 2.4540389972144845,
      "grad_norm": 1.6761799206607695,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8843473196029663,
      "num_tokens": 459698567.0,
      "step": 13215
    },
    {
      "epoch": 2.45422469823584,
      "grad_norm": 1.709390881040947,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.870978593826294,
      "num_tokens": 459733785.0,
      "step": 13216
    },
    {
      "epoch": 2.454410399257196,
      "grad_norm": 1.725827368098528,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8916149139404297,
      "num_tokens": 459762503.0,
      "step": 13217
    },
    {
      "epoch": 2.4545961002785517,
      "grad_norm": 1.511034266832311,
      "learning_rate": 1e-06,
      "loss": 0.2633,
      "mean_token_accuracy": 0.9060099124908447,
      "num_tokens": 459795676.0,
      "step": 13218
    },
    {
      "epoch": 2.454781801299907,
      "grad_norm": 1.627618655241798,
      "learning_rate": 1e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8707791566848755,
      "num_tokens": 459831209.0,
      "step": 13219
    },
    {
      "epoch": 2.4549675023212627,
      "grad_norm": 1.5903586801688543,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8797699809074402,
      "num_tokens": 459868289.0,
      "step": 13220
    },
    {
      "epoch": 2.4551532033426184,
      "grad_norm": 1.7261079342320622,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8786251544952393,
      "num_tokens": 459901581.0,
      "step": 13221
    },
    {
      "epoch": 2.455338904363974,
      "grad_norm": 1.581227451520543,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8818089962005615,
      "num_tokens": 459938455.0,
      "step": 13222
    },
    {
      "epoch": 2.4555246053853295,
      "grad_norm": 1.6331159360990481,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8954333066940308,
      "num_tokens": 459971561.0,
      "step": 13223
    },
    {
      "epoch": 2.455710306406685,
      "grad_norm": 1.530365623607387,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8917884230613708,
      "num_tokens": 460006916.0,
      "step": 13224
    },
    {
      "epoch": 2.455896007428041,
      "grad_norm": 1.6128652651982334,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8760790824890137,
      "num_tokens": 460040663.0,
      "step": 13225
    },
    {
      "epoch": 2.4560817084493967,
      "grad_norm": 1.6338439236253537,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8854950666427612,
      "num_tokens": 460078825.0,
      "step": 13226
    },
    {
      "epoch": 2.456267409470752,
      "grad_norm": 1.4835980908846196,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8904905319213867,
      "num_tokens": 460116113.0,
      "step": 13227
    },
    {
      "epoch": 2.4564531104921077,
      "grad_norm": 1.451339508259783,
      "learning_rate": 1e-06,
      "loss": 0.2767,
      "mean_token_accuracy": 0.8994179964065552,
      "num_tokens": 460149730.0,
      "step": 13228
    },
    {
      "epoch": 2.4566388115134634,
      "grad_norm": 1.5366502078316593,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8997430801391602,
      "num_tokens": 460185637.0,
      "step": 13229
    },
    {
      "epoch": 2.456824512534819,
      "grad_norm": 1.5909412014079072,
      "learning_rate": 1e-06,
      "loss": 0.2908,
      "mean_token_accuracy": 0.9007742404937744,
      "num_tokens": 460217739.0,
      "step": 13230
    },
    {
      "epoch": 2.4570102135561744,
      "grad_norm": 1.5364885916881459,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.887901782989502,
      "num_tokens": 460252800.0,
      "step": 13231
    },
    {
      "epoch": 2.45719591457753,
      "grad_norm": 1.650358031442341,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8758018016815186,
      "num_tokens": 460292919.0,
      "step": 13232
    },
    {
      "epoch": 2.457381615598886,
      "grad_norm": 1.7268757923345386,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8825454711914062,
      "num_tokens": 460320533.0,
      "step": 13233
    },
    {
      "epoch": 2.457567316620241,
      "grad_norm": 1.4503968598956771,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.886753499507904,
      "num_tokens": 460359319.0,
      "step": 13234
    },
    {
      "epoch": 2.457753017641597,
      "grad_norm": 1.5435303817071462,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8835703134536743,
      "num_tokens": 460401049.0,
      "step": 13235
    },
    {
      "epoch": 2.4579387186629527,
      "grad_norm": 1.7538657270451212,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8819476366043091,
      "num_tokens": 460430671.0,
      "step": 13236
    },
    {
      "epoch": 2.4581244196843084,
      "grad_norm": 1.7589794807090788,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8815096616744995,
      "num_tokens": 460460589.0,
      "step": 13237
    },
    {
      "epoch": 2.458310120705664,
      "grad_norm": 1.6131155866868097,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8902295827865601,
      "num_tokens": 460494175.0,
      "step": 13238
    },
    {
      "epoch": 2.4584958217270194,
      "grad_norm": 1.7189987582786153,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8964799642562866,
      "num_tokens": 460522996.0,
      "step": 13239
    },
    {
      "epoch": 2.458681522748375,
      "grad_norm": 1.594812861183804,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8954904675483704,
      "num_tokens": 460556124.0,
      "step": 13240
    },
    {
      "epoch": 2.458867223769731,
      "grad_norm": 1.6178544496769915,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8797622919082642,
      "num_tokens": 460590907.0,
      "step": 13241
    },
    {
      "epoch": 2.459052924791086,
      "grad_norm": 1.7475473693264059,
      "learning_rate": 1e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.8728154301643372,
      "num_tokens": 460622986.0,
      "step": 13242
    },
    {
      "epoch": 2.459238625812442,
      "grad_norm": 1.7864245417438005,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8799723386764526,
      "num_tokens": 460653959.0,
      "step": 13243
    },
    {
      "epoch": 2.4594243268337976,
      "grad_norm": 1.585784233161895,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8779511451721191,
      "num_tokens": 460688800.0,
      "step": 13244
    },
    {
      "epoch": 2.4596100278551534,
      "grad_norm": 1.7176418553563955,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.888025164604187,
      "num_tokens": 460721339.0,
      "step": 13245
    },
    {
      "epoch": 2.4597957288765087,
      "grad_norm": 1.4746978805743411,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8884826898574829,
      "num_tokens": 460755504.0,
      "step": 13246
    },
    {
      "epoch": 2.4599814298978644,
      "grad_norm": 1.6779993892775484,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8832603693008423,
      "num_tokens": 460786761.0,
      "step": 13247
    },
    {
      "epoch": 2.46016713091922,
      "grad_norm": 1.6032490423353158,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8907309770584106,
      "num_tokens": 460820695.0,
      "step": 13248
    },
    {
      "epoch": 2.460352831940576,
      "grad_norm": 1.4607695424756686,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8829891085624695,
      "num_tokens": 460860764.0,
      "step": 13249
    },
    {
      "epoch": 2.460538532961931,
      "grad_norm": 1.474162232781675,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8886680603027344,
      "num_tokens": 460899920.0,
      "step": 13250
    },
    {
      "epoch": 2.460724233983287,
      "grad_norm": 1.681387415273985,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8832713961601257,
      "num_tokens": 460928024.0,
      "step": 13251
    },
    {
      "epoch": 2.4609099350046426,
      "grad_norm": 1.5319761108703043,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8891165256500244,
      "num_tokens": 460961740.0,
      "step": 13252
    },
    {
      "epoch": 2.4610956360259983,
      "grad_norm": 1.5654789476291824,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8884965181350708,
      "num_tokens": 460996690.0,
      "step": 13253
    },
    {
      "epoch": 2.4612813370473536,
      "grad_norm": 1.5748305603545076,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8872015476226807,
      "num_tokens": 461030577.0,
      "step": 13254
    },
    {
      "epoch": 2.4614670380687094,
      "grad_norm": 1.505455490567421,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8780828714370728,
      "num_tokens": 461069571.0,
      "step": 13255
    },
    {
      "epoch": 2.461652739090065,
      "grad_norm": 1.5370947451515184,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8861579895019531,
      "num_tokens": 461102870.0,
      "step": 13256
    },
    {
      "epoch": 2.4618384401114204,
      "grad_norm": 1.6820422246881277,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8679465055465698,
      "num_tokens": 461136947.0,
      "step": 13257
    },
    {
      "epoch": 2.462024141132776,
      "grad_norm": 1.7948579629601245,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8787164688110352,
      "num_tokens": 461168566.0,
      "step": 13258
    },
    {
      "epoch": 2.462209842154132,
      "grad_norm": 1.5320102909770532,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8892909288406372,
      "num_tokens": 461201806.0,
      "step": 13259
    },
    {
      "epoch": 2.4623955431754876,
      "grad_norm": 1.4126039010519207,
      "learning_rate": 1e-06,
      "loss": 0.2726,
      "mean_token_accuracy": 0.9007446765899658,
      "num_tokens": 461238868.0,
      "step": 13260
    },
    {
      "epoch": 2.4625812441968433,
      "grad_norm": 1.6813576087292201,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8787786364555359,
      "num_tokens": 461272604.0,
      "step": 13261
    },
    {
      "epoch": 2.4627669452181986,
      "grad_norm": 1.6589360771542252,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8778047561645508,
      "num_tokens": 461306063.0,
      "step": 13262
    },
    {
      "epoch": 2.4629526462395543,
      "grad_norm": 1.663993382092904,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8940284848213196,
      "num_tokens": 461335675.0,
      "step": 13263
    },
    {
      "epoch": 2.46313834726091,
      "grad_norm": 1.5817791918221147,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8902558088302612,
      "num_tokens": 461373247.0,
      "step": 13264
    },
    {
      "epoch": 2.4633240482822654,
      "grad_norm": 1.5918260772452855,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8819712400436401,
      "num_tokens": 461406340.0,
      "step": 13265
    },
    {
      "epoch": 2.463509749303621,
      "grad_norm": 1.4900322988722463,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8876174688339233,
      "num_tokens": 461445553.0,
      "step": 13266
    },
    {
      "epoch": 2.463695450324977,
      "grad_norm": 1.652662720903008,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8868963718414307,
      "num_tokens": 461477729.0,
      "step": 13267
    },
    {
      "epoch": 2.4638811513463326,
      "grad_norm": 1.641076139579121,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8782433271408081,
      "num_tokens": 461513755.0,
      "step": 13268
    },
    {
      "epoch": 2.464066852367688,
      "grad_norm": 1.6093533415187966,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8848237991333008,
      "num_tokens": 461549928.0,
      "step": 13269
    },
    {
      "epoch": 2.4642525533890436,
      "grad_norm": 1.4280248990684254,
      "learning_rate": 1e-06,
      "loss": 0.2623,
      "mean_token_accuracy": 0.9055605530738831,
      "num_tokens": 461585730.0,
      "step": 13270
    },
    {
      "epoch": 2.4644382544103993,
      "grad_norm": 1.5478292900025044,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8882376551628113,
      "num_tokens": 461622353.0,
      "step": 13271
    },
    {
      "epoch": 2.464623955431755,
      "grad_norm": 1.563347971381284,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8920586109161377,
      "num_tokens": 461657712.0,
      "step": 13272
    },
    {
      "epoch": 2.4648096564531103,
      "grad_norm": 1.7913425893842718,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8880884051322937,
      "num_tokens": 461687393.0,
      "step": 13273
    },
    {
      "epoch": 2.464995357474466,
      "grad_norm": 1.5602755001514312,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8771712183952332,
      "num_tokens": 461727726.0,
      "step": 13274
    },
    {
      "epoch": 2.465181058495822,
      "grad_norm": 1.6081330990961282,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8836808800697327,
      "num_tokens": 461761480.0,
      "step": 13275
    },
    {
      "epoch": 2.4653667595171775,
      "grad_norm": 1.5325883996131922,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8924540281295776,
      "num_tokens": 461800354.0,
      "step": 13276
    },
    {
      "epoch": 2.465552460538533,
      "grad_norm": 1.5646182644717854,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8910990357398987,
      "num_tokens": 461840365.0,
      "step": 13277
    },
    {
      "epoch": 2.4657381615598886,
      "grad_norm": 1.675800992571353,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8885753154754639,
      "num_tokens": 461871629.0,
      "step": 13278
    },
    {
      "epoch": 2.4659238625812443,
      "grad_norm": 1.52503937172788,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8931328058242798,
      "num_tokens": 461909104.0,
      "step": 13279
    },
    {
      "epoch": 2.4661095636025996,
      "grad_norm": 1.6499269309171913,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8787782192230225,
      "num_tokens": 461942563.0,
      "step": 13280
    },
    {
      "epoch": 2.4662952646239553,
      "grad_norm": 1.5780869913977105,
      "learning_rate": 1e-06,
      "loss": 0.3577,
      "mean_token_accuracy": 0.8765882849693298,
      "num_tokens": 461977415.0,
      "step": 13281
    },
    {
      "epoch": 2.466480965645311,
      "grad_norm": 1.5230815084620415,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8958004713058472,
      "num_tokens": 462011591.0,
      "step": 13282
    },
    {
      "epoch": 2.466666666666667,
      "grad_norm": 1.6291018659137926,
      "learning_rate": 1e-06,
      "loss": 0.2968,
      "mean_token_accuracy": 0.896020770072937,
      "num_tokens": 462045101.0,
      "step": 13283
    },
    {
      "epoch": 2.4668523676880225,
      "grad_norm": 1.6146395490623566,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8935381174087524,
      "num_tokens": 462078836.0,
      "step": 13284
    },
    {
      "epoch": 2.467038068709378,
      "grad_norm": 1.5653936972460496,
      "learning_rate": 1e-06,
      "loss": 0.2906,
      "mean_token_accuracy": 0.897205114364624,
      "num_tokens": 462110511.0,
      "step": 13285
    },
    {
      "epoch": 2.4672237697307335,
      "grad_norm": 1.5027772622299362,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8808408379554749,
      "num_tokens": 462149230.0,
      "step": 13286
    },
    {
      "epoch": 2.4674094707520893,
      "grad_norm": 1.8301096869264821,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8782879114151001,
      "num_tokens": 462178530.0,
      "step": 13287
    },
    {
      "epoch": 2.4675951717734446,
      "grad_norm": 1.559248344189788,
      "learning_rate": 1e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8724282383918762,
      "num_tokens": 462218151.0,
      "step": 13288
    },
    {
      "epoch": 2.4677808727948003,
      "grad_norm": 1.883916722382254,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8659690618515015,
      "num_tokens": 462244013.0,
      "step": 13289
    },
    {
      "epoch": 2.467966573816156,
      "grad_norm": 1.6879597104300326,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.865761399269104,
      "num_tokens": 462279722.0,
      "step": 13290
    },
    {
      "epoch": 2.4681522748375118,
      "grad_norm": 1.4861126130347504,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8769157528877258,
      "num_tokens": 462319758.0,
      "step": 13291
    },
    {
      "epoch": 2.468337975858867,
      "grad_norm": 1.5141375711096299,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8872660398483276,
      "num_tokens": 462357791.0,
      "step": 13292
    },
    {
      "epoch": 2.468523676880223,
      "grad_norm": 1.547180607232732,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8909811973571777,
      "num_tokens": 462393929.0,
      "step": 13293
    },
    {
      "epoch": 2.4687093779015785,
      "grad_norm": 1.7601803459770298,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8761394023895264,
      "num_tokens": 462425307.0,
      "step": 13294
    },
    {
      "epoch": 2.4688950789229342,
      "grad_norm": 1.6219331280055256,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8872891068458557,
      "num_tokens": 462460886.0,
      "step": 13295
    },
    {
      "epoch": 2.4690807799442895,
      "grad_norm": 1.6818528001553392,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8801225423812866,
      "num_tokens": 462500079.0,
      "step": 13296
    },
    {
      "epoch": 2.4692664809656453,
      "grad_norm": 1.5166266425573802,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8886420726776123,
      "num_tokens": 462537415.0,
      "step": 13297
    },
    {
      "epoch": 2.469452181987001,
      "grad_norm": 1.4081179310128211,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.8936570882797241,
      "num_tokens": 462575926.0,
      "step": 13298
    },
    {
      "epoch": 2.4696378830083567,
      "grad_norm": 1.3923821486585173,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8904117941856384,
      "num_tokens": 462620006.0,
      "step": 13299
    },
    {
      "epoch": 2.469823584029712,
      "grad_norm": 1.4482856966924884,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8907883763313293,
      "num_tokens": 462655740.0,
      "step": 13300
    },
    {
      "epoch": 2.4700092850510678,
      "grad_norm": 1.8279397366896346,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.8789271116256714,
      "num_tokens": 462686792.0,
      "step": 13301
    },
    {
      "epoch": 2.4701949860724235,
      "grad_norm": 1.5929093617992902,
      "learning_rate": 1e-06,
      "loss": 0.3018,
      "mean_token_accuracy": 0.8954517245292664,
      "num_tokens": 462719981.0,
      "step": 13302
    },
    {
      "epoch": 2.470380687093779,
      "grad_norm": 1.5588601020751316,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8936732411384583,
      "num_tokens": 462753154.0,
      "step": 13303
    },
    {
      "epoch": 2.4705663881151345,
      "grad_norm": 1.5805803148017423,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8861662745475769,
      "num_tokens": 462786845.0,
      "step": 13304
    },
    {
      "epoch": 2.4707520891364902,
      "grad_norm": 1.614280085629983,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8758066296577454,
      "num_tokens": 462820356.0,
      "step": 13305
    },
    {
      "epoch": 2.470937790157846,
      "grad_norm": 1.5692171500279222,
      "learning_rate": 1e-06,
      "loss": 0.3722,
      "mean_token_accuracy": 0.8705923557281494,
      "num_tokens": 462857011.0,
      "step": 13306
    },
    {
      "epoch": 2.4711234911792017,
      "grad_norm": 1.7791438494755079,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8874577283859253,
      "num_tokens": 462882844.0,
      "step": 13307
    },
    {
      "epoch": 2.471309192200557,
      "grad_norm": 1.7142197927719582,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8890888094902039,
      "num_tokens": 462913590.0,
      "step": 13308
    },
    {
      "epoch": 2.4714948932219127,
      "grad_norm": 1.4914091395507603,
      "learning_rate": 1e-06,
      "loss": 0.2911,
      "mean_token_accuracy": 0.8945459127426147,
      "num_tokens": 462948197.0,
      "step": 13309
    },
    {
      "epoch": 2.4716805942432685,
      "grad_norm": 1.5260618121828322,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8889288902282715,
      "num_tokens": 462984938.0,
      "step": 13310
    },
    {
      "epoch": 2.4718662952646238,
      "grad_norm": 1.575666803956316,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8916643261909485,
      "num_tokens": 463020555.0,
      "step": 13311
    },
    {
      "epoch": 2.4720519962859795,
      "grad_norm": 1.5757819658181664,
      "learning_rate": 1e-06,
      "loss": 0.2908,
      "mean_token_accuracy": 0.8940502405166626,
      "num_tokens": 463053303.0,
      "step": 13312
    },
    {
      "epoch": 2.4722376973073352,
      "grad_norm": 1.6752251028074718,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.8688690662384033,
      "num_tokens": 463083213.0,
      "step": 13313
    },
    {
      "epoch": 2.472423398328691,
      "grad_norm": 1.510415664676798,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8857665061950684,
      "num_tokens": 463122335.0,
      "step": 13314
    },
    {
      "epoch": 2.4726090993500462,
      "grad_norm": 1.5148181325121708,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8701852560043335,
      "num_tokens": 463161545.0,
      "step": 13315
    },
    {
      "epoch": 2.472794800371402,
      "grad_norm": 1.4953190375986705,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8984184265136719,
      "num_tokens": 463203444.0,
      "step": 13316
    },
    {
      "epoch": 2.4729805013927577,
      "grad_norm": 1.5422756903684003,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8941752314567566,
      "num_tokens": 463240605.0,
      "step": 13317
    },
    {
      "epoch": 2.4731662024141134,
      "grad_norm": 1.5220771437973153,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8868686556816101,
      "num_tokens": 463276952.0,
      "step": 13318
    },
    {
      "epoch": 2.4733519034354687,
      "grad_norm": 1.5432561157204854,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8847118020057678,
      "num_tokens": 463311952.0,
      "step": 13319
    },
    {
      "epoch": 2.4735376044568245,
      "grad_norm": 1.6349160002687806,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8782837986946106,
      "num_tokens": 463345533.0,
      "step": 13320
    },
    {
      "epoch": 2.47372330547818,
      "grad_norm": 1.4528886213060208,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8924426436424255,
      "num_tokens": 463385305.0,
      "step": 13321
    },
    {
      "epoch": 2.473909006499536,
      "grad_norm": 1.692476002565774,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8915265202522278,
      "num_tokens": 463415844.0,
      "step": 13322
    },
    {
      "epoch": 2.4740947075208912,
      "grad_norm": 1.8626263928228375,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8714561462402344,
      "num_tokens": 463447726.0,
      "step": 13323
    },
    {
      "epoch": 2.474280408542247,
      "grad_norm": 1.664708477361982,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.879261314868927,
      "num_tokens": 463482795.0,
      "step": 13324
    },
    {
      "epoch": 2.4744661095636027,
      "grad_norm": 1.4816847677875176,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.8956952691078186,
      "num_tokens": 463522294.0,
      "step": 13325
    },
    {
      "epoch": 2.474651810584958,
      "grad_norm": 1.6468081667203576,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8775230646133423,
      "num_tokens": 463554001.0,
      "step": 13326
    },
    {
      "epoch": 2.4748375116063137,
      "grad_norm": 1.6108503853603746,
      "learning_rate": 1e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8723230361938477,
      "num_tokens": 463590909.0,
      "step": 13327
    },
    {
      "epoch": 2.4750232126276694,
      "grad_norm": 1.6340278324724569,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8995845913887024,
      "num_tokens": 463621037.0,
      "step": 13328
    },
    {
      "epoch": 2.475208913649025,
      "grad_norm": 1.5751791330349503,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8854255676269531,
      "num_tokens": 463655211.0,
      "step": 13329
    },
    {
      "epoch": 2.475394614670381,
      "grad_norm": 1.5962044555834072,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.891404390335083,
      "num_tokens": 463687022.0,
      "step": 13330
    },
    {
      "epoch": 2.475580315691736,
      "grad_norm": 1.5948408823361113,
      "learning_rate": 1e-06,
      "loss": 0.2827,
      "mean_token_accuracy": 0.9004827737808228,
      "num_tokens": 463719945.0,
      "step": 13331
    },
    {
      "epoch": 2.475766016713092,
      "grad_norm": 1.8050020759288736,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8797832131385803,
      "num_tokens": 463752164.0,
      "step": 13332
    },
    {
      "epoch": 2.4759517177344477,
      "grad_norm": 1.6204861352280862,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8641555309295654,
      "num_tokens": 463792692.0,
      "step": 13333
    },
    {
      "epoch": 2.476137418755803,
      "grad_norm": 1.6228752149532024,
      "learning_rate": 1e-06,
      "loss": 0.2743,
      "mean_token_accuracy": 0.8999785780906677,
      "num_tokens": 463824481.0,
      "step": 13334
    },
    {
      "epoch": 2.4763231197771587,
      "grad_norm": 1.582883789189531,
      "learning_rate": 1e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8697451949119568,
      "num_tokens": 463863414.0,
      "step": 13335
    },
    {
      "epoch": 2.4765088207985144,
      "grad_norm": 1.7205759067959834,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8851078152656555,
      "num_tokens": 463894991.0,
      "step": 13336
    },
    {
      "epoch": 2.47669452181987,
      "grad_norm": 1.5843709157987285,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8889563083648682,
      "num_tokens": 463929068.0,
      "step": 13337
    },
    {
      "epoch": 2.4768802228412254,
      "grad_norm": 1.5674509431715085,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.88294917345047,
      "num_tokens": 463962850.0,
      "step": 13338
    },
    {
      "epoch": 2.477065923862581,
      "grad_norm": 1.753521486335862,
      "learning_rate": 1e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8764967918395996,
      "num_tokens": 463992041.0,
      "step": 13339
    },
    {
      "epoch": 2.477251624883937,
      "grad_norm": 1.6749785432356064,
      "learning_rate": 1e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8721641302108765,
      "num_tokens": 464029101.0,
      "step": 13340
    },
    {
      "epoch": 2.4774373259052926,
      "grad_norm": 1.7517641127791876,
      "learning_rate": 1e-06,
      "loss": 0.3658,
      "mean_token_accuracy": 0.875839352607727,
      "num_tokens": 464057533.0,
      "step": 13341
    },
    {
      "epoch": 2.477623026926648,
      "grad_norm": 1.5102049467704857,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8876445293426514,
      "num_tokens": 464094017.0,
      "step": 13342
    },
    {
      "epoch": 2.4778087279480037,
      "grad_norm": 1.6383280539882479,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8700792789459229,
      "num_tokens": 464128816.0,
      "step": 13343
    },
    {
      "epoch": 2.4779944289693594,
      "grad_norm": 1.4412726900397934,
      "learning_rate": 1e-06,
      "loss": 0.2911,
      "mean_token_accuracy": 0.8953055143356323,
      "num_tokens": 464170571.0,
      "step": 13344
    },
    {
      "epoch": 2.478180129990715,
      "grad_norm": 1.6503651472733134,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8792417049407959,
      "num_tokens": 464201657.0,
      "step": 13345
    },
    {
      "epoch": 2.4783658310120704,
      "grad_norm": 1.583046003873128,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8920709490776062,
      "num_tokens": 464234037.0,
      "step": 13346
    },
    {
      "epoch": 2.478551532033426,
      "grad_norm": 1.634393339264888,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8913647532463074,
      "num_tokens": 464265083.0,
      "step": 13347
    },
    {
      "epoch": 2.478737233054782,
      "grad_norm": 1.4174878617438769,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8911281228065491,
      "num_tokens": 464307661.0,
      "step": 13348
    },
    {
      "epoch": 2.478922934076137,
      "grad_norm": 1.5117919611573514,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.9001983404159546,
      "num_tokens": 464342550.0,
      "step": 13349
    },
    {
      "epoch": 2.479108635097493,
      "grad_norm": 1.5376471460360055,
      "learning_rate": 1e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8726282715797424,
      "num_tokens": 464378743.0,
      "step": 13350
    },
    {
      "epoch": 2.4792943361188486,
      "grad_norm": 1.481888011170981,
      "learning_rate": 1e-06,
      "loss": 0.2826,
      "mean_token_accuracy": 0.9033080339431763,
      "num_tokens": 464413950.0,
      "step": 13351
    },
    {
      "epoch": 2.4794800371402044,
      "grad_norm": 1.5633831777059786,
      "learning_rate": 1e-06,
      "loss": 0.2862,
      "mean_token_accuracy": 0.900905966758728,
      "num_tokens": 464446933.0,
      "step": 13352
    },
    {
      "epoch": 2.47966573816156,
      "grad_norm": 1.5665302208286138,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8887863159179688,
      "num_tokens": 464482013.0,
      "step": 13353
    },
    {
      "epoch": 2.4798514391829154,
      "grad_norm": 1.5903363500266665,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8915720582008362,
      "num_tokens": 464515945.0,
      "step": 13354
    },
    {
      "epoch": 2.480037140204271,
      "grad_norm": 1.6358635584892605,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8763083219528198,
      "num_tokens": 464549085.0,
      "step": 13355
    },
    {
      "epoch": 2.480222841225627,
      "grad_norm": 1.4525465734533964,
      "learning_rate": 1e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8843372464179993,
      "num_tokens": 464591131.0,
      "step": 13356
    },
    {
      "epoch": 2.480408542246982,
      "grad_norm": 1.611329047336896,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.881051778793335,
      "num_tokens": 464628279.0,
      "step": 13357
    },
    {
      "epoch": 2.480594243268338,
      "grad_norm": 1.5119694044953493,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8910834789276123,
      "num_tokens": 464662484.0,
      "step": 13358
    },
    {
      "epoch": 2.4807799442896936,
      "grad_norm": 1.62872518789118,
      "learning_rate": 1e-06,
      "loss": 0.2885,
      "mean_token_accuracy": 0.8941744565963745,
      "num_tokens": 464693015.0,
      "step": 13359
    },
    {
      "epoch": 2.4809656453110494,
      "grad_norm": 1.6207891092543447,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8846368789672852,
      "num_tokens": 464724217.0,
      "step": 13360
    },
    {
      "epoch": 2.4811513463324046,
      "grad_norm": 1.6591398119631857,
      "learning_rate": 1e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.8746442198753357,
      "num_tokens": 464759160.0,
      "step": 13361
    },
    {
      "epoch": 2.4813370473537604,
      "grad_norm": 1.5495551486179695,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8932995200157166,
      "num_tokens": 464793720.0,
      "step": 13362
    },
    {
      "epoch": 2.481522748375116,
      "grad_norm": 1.4706305640521296,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8902458548545837,
      "num_tokens": 464833371.0,
      "step": 13363
    },
    {
      "epoch": 2.481708449396472,
      "grad_norm": 1.5625997627789738,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8776844143867493,
      "num_tokens": 464870246.0,
      "step": 13364
    },
    {
      "epoch": 2.481894150417827,
      "grad_norm": 1.47807409647297,
      "learning_rate": 1e-06,
      "loss": 0.2564,
      "mean_token_accuracy": 0.9088269472122192,
      "num_tokens": 464904406.0,
      "step": 13365
    },
    {
      "epoch": 2.482079851439183,
      "grad_norm": 1.487679079894618,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.894546627998352,
      "num_tokens": 464941049.0,
      "step": 13366
    },
    {
      "epoch": 2.4822655524605386,
      "grad_norm": 1.5563303119092773,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8837045431137085,
      "num_tokens": 464975895.0,
      "step": 13367
    },
    {
      "epoch": 2.4824512534818943,
      "grad_norm": 1.606635095608848,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8884608149528503,
      "num_tokens": 465010139.0,
      "step": 13368
    },
    {
      "epoch": 2.4826369545032496,
      "grad_norm": 1.4811952365320733,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.885804295539856,
      "num_tokens": 465046513.0,
      "step": 13369
    },
    {
      "epoch": 2.4828226555246053,
      "grad_norm": 1.7729614485327876,
      "learning_rate": 1e-06,
      "loss": 0.347,
      "mean_token_accuracy": 0.8775522112846375,
      "num_tokens": 465074582.0,
      "step": 13370
    },
    {
      "epoch": 2.483008356545961,
      "grad_norm": 1.4218117365667975,
      "learning_rate": 1e-06,
      "loss": 0.281,
      "mean_token_accuracy": 0.9024395942687988,
      "num_tokens": 465114524.0,
      "step": 13371
    },
    {
      "epoch": 2.4831940575673164,
      "grad_norm": 1.4575680094357732,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8865289092063904,
      "num_tokens": 465153681.0,
      "step": 13372
    },
    {
      "epoch": 2.483379758588672,
      "grad_norm": 1.4565784816443983,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8928853273391724,
      "num_tokens": 465190857.0,
      "step": 13373
    },
    {
      "epoch": 2.483565459610028,
      "grad_norm": 1.5058384620954102,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8860733509063721,
      "num_tokens": 465227504.0,
      "step": 13374
    },
    {
      "epoch": 2.4837511606313836,
      "grad_norm": 1.8393756347387402,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8723888397216797,
      "num_tokens": 465258132.0,
      "step": 13375
    },
    {
      "epoch": 2.4839368616527393,
      "grad_norm": 1.4892187200829496,
      "learning_rate": 1e-06,
      "loss": 0.2882,
      "mean_token_accuracy": 0.8975932002067566,
      "num_tokens": 465293456.0,
      "step": 13376
    },
    {
      "epoch": 2.4841225626740946,
      "grad_norm": 1.4443884671743001,
      "learning_rate": 1e-06,
      "loss": 0.2895,
      "mean_token_accuracy": 0.8970947861671448,
      "num_tokens": 465331293.0,
      "step": 13377
    },
    {
      "epoch": 2.4843082636954503,
      "grad_norm": 1.7167672745617737,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8871333599090576,
      "num_tokens": 465363491.0,
      "step": 13378
    },
    {
      "epoch": 2.484493964716806,
      "grad_norm": 1.6555598705327643,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8761013746261597,
      "num_tokens": 465398233.0,
      "step": 13379
    },
    {
      "epoch": 2.4846796657381613,
      "grad_norm": 1.4339973399292016,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8902950286865234,
      "num_tokens": 465436935.0,
      "step": 13380
    },
    {
      "epoch": 2.484865366759517,
      "grad_norm": 1.6265126190382233,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8847147226333618,
      "num_tokens": 465469729.0,
      "step": 13381
    },
    {
      "epoch": 2.485051067780873,
      "grad_norm": 1.614546847747218,
      "learning_rate": 1e-06,
      "loss": 0.3131,
      "mean_token_accuracy": 0.8870704770088196,
      "num_tokens": 465499091.0,
      "step": 13382
    },
    {
      "epoch": 2.4852367688022285,
      "grad_norm": 1.5645402773764074,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8796467185020447,
      "num_tokens": 465534974.0,
      "step": 13383
    },
    {
      "epoch": 2.4854224698235843,
      "grad_norm": 1.620358368191457,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8835228681564331,
      "num_tokens": 465573390.0,
      "step": 13384
    },
    {
      "epoch": 2.4856081708449396,
      "grad_norm": 1.3962938427377654,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8825213313102722,
      "num_tokens": 465616544.0,
      "step": 13385
    },
    {
      "epoch": 2.4857938718662953,
      "grad_norm": 1.4961098313022867,
      "learning_rate": 1e-06,
      "loss": 0.2934,
      "mean_token_accuracy": 0.8947425484657288,
      "num_tokens": 465652832.0,
      "step": 13386
    },
    {
      "epoch": 2.485979572887651,
      "grad_norm": 1.6175544245088787,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8861225843429565,
      "num_tokens": 465687400.0,
      "step": 13387
    },
    {
      "epoch": 2.4861652739090063,
      "grad_norm": 1.5600370392473664,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8969250321388245,
      "num_tokens": 465718885.0,
      "step": 13388
    },
    {
      "epoch": 2.486350974930362,
      "grad_norm": 1.4797453942844068,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.886027455329895,
      "num_tokens": 465761902.0,
      "step": 13389
    },
    {
      "epoch": 2.486536675951718,
      "grad_norm": 1.416062374236905,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8874157667160034,
      "num_tokens": 465804705.0,
      "step": 13390
    },
    {
      "epoch": 2.4867223769730735,
      "grad_norm": 1.4286836337210782,
      "learning_rate": 1e-06,
      "loss": 0.3116,
      "mean_token_accuracy": 0.8915620446205139,
      "num_tokens": 465844032.0,
      "step": 13391
    },
    {
      "epoch": 2.486908077994429,
      "grad_norm": 1.4802778276747526,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8832578063011169,
      "num_tokens": 465883050.0,
      "step": 13392
    },
    {
      "epoch": 2.4870937790157845,
      "grad_norm": 1.5785330566884215,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8813862800598145,
      "num_tokens": 465921902.0,
      "step": 13393
    },
    {
      "epoch": 2.4872794800371403,
      "grad_norm": 1.3371767031887583,
      "learning_rate": 1e-06,
      "loss": 0.3048,
      "mean_token_accuracy": 0.8910194039344788,
      "num_tokens": 465969201.0,
      "step": 13394
    },
    {
      "epoch": 2.487465181058496,
      "grad_norm": 1.3742972126174091,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.8980680704116821,
      "num_tokens": 466011618.0,
      "step": 13395
    },
    {
      "epoch": 2.4876508820798513,
      "grad_norm": 1.4261500900588935,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8910689949989319,
      "num_tokens": 466051873.0,
      "step": 13396
    },
    {
      "epoch": 2.487836583101207,
      "grad_norm": 1.6133605076494117,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8888848423957825,
      "num_tokens": 466084748.0,
      "step": 13397
    },
    {
      "epoch": 2.4880222841225628,
      "grad_norm": 1.5539248437601714,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.8965755701065063,
      "num_tokens": 466117327.0,
      "step": 13398
    },
    {
      "epoch": 2.4882079851439185,
      "grad_norm": 1.530379960093226,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.882820188999176,
      "num_tokens": 466154316.0,
      "step": 13399
    },
    {
      "epoch": 2.488393686165274,
      "grad_norm": 1.5063929733900387,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8886675238609314,
      "num_tokens": 466194014.0,
      "step": 13400
    },
    {
      "epoch": 2.4885793871866295,
      "grad_norm": 1.415150079218855,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8889936208724976,
      "num_tokens": 466233636.0,
      "step": 13401
    },
    {
      "epoch": 2.4887650882079853,
      "grad_norm": 1.5729522933389999,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8846144676208496,
      "num_tokens": 466268624.0,
      "step": 13402
    },
    {
      "epoch": 2.4889507892293405,
      "grad_norm": 1.6070227741864291,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8783829212188721,
      "num_tokens": 466305269.0,
      "step": 13403
    },
    {
      "epoch": 2.4891364902506963,
      "grad_norm": 1.7575908726420275,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8782747983932495,
      "num_tokens": 466338526.0,
      "step": 13404
    },
    {
      "epoch": 2.489322191272052,
      "grad_norm": 1.560668336863561,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8813048601150513,
      "num_tokens": 466374088.0,
      "step": 13405
    },
    {
      "epoch": 2.4895078922934077,
      "grad_norm": 1.844064013823442,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8883230686187744,
      "num_tokens": 466401998.0,
      "step": 13406
    },
    {
      "epoch": 2.4896935933147635,
      "grad_norm": 1.5319691399405204,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8838775753974915,
      "num_tokens": 466438786.0,
      "step": 13407
    },
    {
      "epoch": 2.4898792943361188,
      "grad_norm": 1.4712124936074336,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8876844644546509,
      "num_tokens": 466477870.0,
      "step": 13408
    },
    {
      "epoch": 2.4900649953574745,
      "grad_norm": 1.4466867505467202,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8887689113616943,
      "num_tokens": 466517897.0,
      "step": 13409
    },
    {
      "epoch": 2.4902506963788302,
      "grad_norm": 1.5888401184852872,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8753354549407959,
      "num_tokens": 466553599.0,
      "step": 13410
    },
    {
      "epoch": 2.4904363974001855,
      "grad_norm": 1.6400315144204856,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.883463978767395,
      "num_tokens": 466585786.0,
      "step": 13411
    },
    {
      "epoch": 2.4906220984215413,
      "grad_norm": 1.64865013751349,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8824782967567444,
      "num_tokens": 466618954.0,
      "step": 13412
    },
    {
      "epoch": 2.490807799442897,
      "grad_norm": 1.604935408404211,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8863726258277893,
      "num_tokens": 466652179.0,
      "step": 13413
    },
    {
      "epoch": 2.4909935004642527,
      "grad_norm": 1.6192252435284342,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8926421403884888,
      "num_tokens": 466686461.0,
      "step": 13414
    },
    {
      "epoch": 2.491179201485608,
      "grad_norm": 1.6233373878751025,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8766262531280518,
      "num_tokens": 466724242.0,
      "step": 13415
    },
    {
      "epoch": 2.4913649025069637,
      "grad_norm": 1.8491948385389883,
      "learning_rate": 1e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8771121501922607,
      "num_tokens": 466753341.0,
      "step": 13416
    },
    {
      "epoch": 2.4915506035283195,
      "grad_norm": 1.5062375580269634,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8860828876495361,
      "num_tokens": 466789525.0,
      "step": 13417
    },
    {
      "epoch": 2.491736304549675,
      "grad_norm": 1.5626838984106701,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8901990652084351,
      "num_tokens": 466823457.0,
      "step": 13418
    },
    {
      "epoch": 2.4919220055710305,
      "grad_norm": 1.5495280829917677,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8868303298950195,
      "num_tokens": 466857060.0,
      "step": 13419
    },
    {
      "epoch": 2.4921077065923862,
      "grad_norm": 1.6291162280075222,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8835175037384033,
      "num_tokens": 466890756.0,
      "step": 13420
    },
    {
      "epoch": 2.492293407613742,
      "grad_norm": 1.7544588875316809,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.8864684104919434,
      "num_tokens": 466922814.0,
      "step": 13421
    },
    {
      "epoch": 2.4924791086350977,
      "grad_norm": 1.4856982046081348,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.9017041921615601,
      "num_tokens": 466959054.0,
      "step": 13422
    },
    {
      "epoch": 2.492664809656453,
      "grad_norm": 1.701746450183159,
      "learning_rate": 1e-06,
      "loss": 0.2907,
      "mean_token_accuracy": 0.8982590436935425,
      "num_tokens": 466986944.0,
      "step": 13423
    },
    {
      "epoch": 2.4928505106778087,
      "grad_norm": 1.4135332220271841,
      "learning_rate": 1e-06,
      "loss": 0.3001,
      "mean_token_accuracy": 0.8921035528182983,
      "num_tokens": 467027919.0,
      "step": 13424
    },
    {
      "epoch": 2.4930362116991645,
      "grad_norm": 1.5606018880932688,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8836553692817688,
      "num_tokens": 467061839.0,
      "step": 13425
    },
    {
      "epoch": 2.4932219127205197,
      "grad_norm": 1.473978816016304,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8766168355941772,
      "num_tokens": 467103314.0,
      "step": 13426
    },
    {
      "epoch": 2.4934076137418755,
      "grad_norm": 1.8426376399012547,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8917886018753052,
      "num_tokens": 467127877.0,
      "step": 13427
    },
    {
      "epoch": 2.493593314763231,
      "grad_norm": 1.403564670588194,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.898303210735321,
      "num_tokens": 467165471.0,
      "step": 13428
    },
    {
      "epoch": 2.493779015784587,
      "grad_norm": 1.6411310538763295,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8807204961776733,
      "num_tokens": 467200217.0,
      "step": 13429
    },
    {
      "epoch": 2.4939647168059427,
      "grad_norm": 1.6846103399584018,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8874348998069763,
      "num_tokens": 467232496.0,
      "step": 13430
    },
    {
      "epoch": 2.494150417827298,
      "grad_norm": 1.660510205830986,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8860455751419067,
      "num_tokens": 467264370.0,
      "step": 13431
    },
    {
      "epoch": 2.4943361188486537,
      "grad_norm": 1.6996527688883711,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8681673407554626,
      "num_tokens": 467296540.0,
      "step": 13432
    },
    {
      "epoch": 2.4945218198700094,
      "grad_norm": 1.6866038055634573,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8879587650299072,
      "num_tokens": 467327939.0,
      "step": 13433
    },
    {
      "epoch": 2.4947075208913647,
      "grad_norm": 1.7691383111398047,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8888206481933594,
      "num_tokens": 467358841.0,
      "step": 13434
    },
    {
      "epoch": 2.4948932219127204,
      "grad_norm": 1.5598998246456295,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.883080005645752,
      "num_tokens": 467396480.0,
      "step": 13435
    },
    {
      "epoch": 2.495078922934076,
      "grad_norm": 1.7418932301553354,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8953349590301514,
      "num_tokens": 467425167.0,
      "step": 13436
    },
    {
      "epoch": 2.495264623955432,
      "grad_norm": 1.5355271178013052,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8870514631271362,
      "num_tokens": 467459254.0,
      "step": 13437
    },
    {
      "epoch": 2.495450324976787,
      "grad_norm": 1.6442351392603276,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8828942775726318,
      "num_tokens": 467492154.0,
      "step": 13438
    },
    {
      "epoch": 2.495636025998143,
      "grad_norm": 1.467269113077493,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8970353603363037,
      "num_tokens": 467528744.0,
      "step": 13439
    },
    {
      "epoch": 2.4958217270194987,
      "grad_norm": 1.556149249474108,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8816242218017578,
      "num_tokens": 467567705.0,
      "step": 13440
    },
    {
      "epoch": 2.4960074280408544,
      "grad_norm": 1.5539543583138566,
      "learning_rate": 1e-06,
      "loss": 0.291,
      "mean_token_accuracy": 0.8963807821273804,
      "num_tokens": 467599637.0,
      "step": 13441
    },
    {
      "epoch": 2.4961931290622097,
      "grad_norm": 1.617961109488858,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.890644907951355,
      "num_tokens": 467633906.0,
      "step": 13442
    },
    {
      "epoch": 2.4963788300835654,
      "grad_norm": 1.5547057205762134,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8944247961044312,
      "num_tokens": 467667713.0,
      "step": 13443
    },
    {
      "epoch": 2.496564531104921,
      "grad_norm": 1.5065398578448308,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8877629041671753,
      "num_tokens": 467706605.0,
      "step": 13444
    },
    {
      "epoch": 2.496750232126277,
      "grad_norm": 1.5919778695185793,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8842877745628357,
      "num_tokens": 467741811.0,
      "step": 13445
    },
    {
      "epoch": 2.496935933147632,
      "grad_norm": 1.5810045543801574,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.87827068567276,
      "num_tokens": 467776849.0,
      "step": 13446
    },
    {
      "epoch": 2.497121634168988,
      "grad_norm": 1.696304474688696,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8815719485282898,
      "num_tokens": 467807756.0,
      "step": 13447
    },
    {
      "epoch": 2.4973073351903436,
      "grad_norm": 1.5323223738258152,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8940367102622986,
      "num_tokens": 467842865.0,
      "step": 13448
    },
    {
      "epoch": 2.497493036211699,
      "grad_norm": 1.411047973943916,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.892126202583313,
      "num_tokens": 467881577.0,
      "step": 13449
    },
    {
      "epoch": 2.4976787372330547,
      "grad_norm": 1.5521305528515161,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8817993402481079,
      "num_tokens": 467917376.0,
      "step": 13450
    },
    {
      "epoch": 2.4978644382544104,
      "grad_norm": 1.497154064918203,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8902888298034668,
      "num_tokens": 467955040.0,
      "step": 13451
    },
    {
      "epoch": 2.498050139275766,
      "grad_norm": 1.6534312861316802,
      "learning_rate": 1e-06,
      "loss": 0.3098,
      "mean_token_accuracy": 0.891336977481842,
      "num_tokens": 467989205.0,
      "step": 13452
    },
    {
      "epoch": 2.498235840297122,
      "grad_norm": 1.43842710978746,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.890299916267395,
      "num_tokens": 468028231.0,
      "step": 13453
    },
    {
      "epoch": 2.498421541318477,
      "grad_norm": 1.6179181514714127,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8893064260482788,
      "num_tokens": 468060539.0,
      "step": 13454
    },
    {
      "epoch": 2.498607242339833,
      "grad_norm": 1.6043441569216754,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8843244910240173,
      "num_tokens": 468097059.0,
      "step": 13455
    },
    {
      "epoch": 2.4987929433611886,
      "grad_norm": 1.6340491603950007,
      "learning_rate": 1e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8600807189941406,
      "num_tokens": 468135203.0,
      "step": 13456
    },
    {
      "epoch": 2.498978644382544,
      "grad_norm": 1.4821250346595087,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8895485401153564,
      "num_tokens": 468172288.0,
      "step": 13457
    },
    {
      "epoch": 2.4991643454038996,
      "grad_norm": 1.7413550170776866,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8826552033424377,
      "num_tokens": 468200531.0,
      "step": 13458
    },
    {
      "epoch": 2.4993500464252554,
      "grad_norm": 1.6439407051701662,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8817048072814941,
      "num_tokens": 468229195.0,
      "step": 13459
    },
    {
      "epoch": 2.499535747446611,
      "grad_norm": 1.4266216257012725,
      "learning_rate": 1e-06,
      "loss": 0.2749,
      "mean_token_accuracy": 0.9017981290817261,
      "num_tokens": 468263593.0,
      "step": 13460
    },
    {
      "epoch": 2.4997214484679664,
      "grad_norm": 1.4884664482117929,
      "learning_rate": 1e-06,
      "loss": 0.2993,
      "mean_token_accuracy": 0.8955645561218262,
      "num_tokens": 468301803.0,
      "step": 13461
    },
    {
      "epoch": 2.499907149489322,
      "grad_norm": 1.5389465919285883,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.8968233466148376,
      "num_tokens": 468334909.0,
      "step": 13462
    },
    {
      "epoch": 2.500092850510678,
      "grad_norm": 1.6051055155387128,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8813605308532715,
      "num_tokens": 468369658.0,
      "step": 13463
    },
    {
      "epoch": 2.500278551532033,
      "grad_norm": 1.5801800860798951,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8881474733352661,
      "num_tokens": 468402046.0,
      "step": 13464
    },
    {
      "epoch": 2.500464252553389,
      "grad_norm": 1.4739388806939566,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8925623893737793,
      "num_tokens": 468439156.0,
      "step": 13465
    },
    {
      "epoch": 2.5006499535747446,
      "grad_norm": 1.452432682761306,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8811879754066467,
      "num_tokens": 468479132.0,
      "step": 13466
    },
    {
      "epoch": 2.5008356545961004,
      "grad_norm": 1.626280131103193,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8838725686073303,
      "num_tokens": 468513390.0,
      "step": 13467
    },
    {
      "epoch": 2.501021355617456,
      "grad_norm": 1.723756515253602,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8850002884864807,
      "num_tokens": 468543827.0,
      "step": 13468
    },
    {
      "epoch": 2.5012070566388114,
      "grad_norm": 1.5127135544705672,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8940615653991699,
      "num_tokens": 468583668.0,
      "step": 13469
    },
    {
      "epoch": 2.501392757660167,
      "grad_norm": 1.6415141272289058,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8689206838607788,
      "num_tokens": 468621731.0,
      "step": 13470
    },
    {
      "epoch": 2.501578458681523,
      "grad_norm": 1.620110705239117,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8811126351356506,
      "num_tokens": 468656682.0,
      "step": 13471
    },
    {
      "epoch": 2.501764159702878,
      "grad_norm": 1.4924854964824428,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8887588977813721,
      "num_tokens": 468693000.0,
      "step": 13472
    },
    {
      "epoch": 2.501949860724234,
      "grad_norm": 1.5458715692612375,
      "learning_rate": 1e-06,
      "loss": 0.2779,
      "mean_token_accuracy": 0.9015525579452515,
      "num_tokens": 468725476.0,
      "step": 13473
    },
    {
      "epoch": 2.5021355617455896,
      "grad_norm": 1.730956995333056,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8970315456390381,
      "num_tokens": 468753605.0,
      "step": 13474
    },
    {
      "epoch": 2.5023212627669453,
      "grad_norm": 1.4888362560767483,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8875737190246582,
      "num_tokens": 468789890.0,
      "step": 13475
    },
    {
      "epoch": 2.502506963788301,
      "grad_norm": 1.5188531692101064,
      "learning_rate": 1e-06,
      "loss": 0.2664,
      "mean_token_accuracy": 0.9024400115013123,
      "num_tokens": 468823218.0,
      "step": 13476
    },
    {
      "epoch": 2.5026926648096564,
      "grad_norm": 1.5205637223895283,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8795795440673828,
      "num_tokens": 468863362.0,
      "step": 13477
    },
    {
      "epoch": 2.502878365831012,
      "grad_norm": 1.593600609280001,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8912721872329712,
      "num_tokens": 468895453.0,
      "step": 13478
    },
    {
      "epoch": 2.503064066852368,
      "grad_norm": 1.650510916418948,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8896132707595825,
      "num_tokens": 468923995.0,
      "step": 13479
    },
    {
      "epoch": 2.503249767873723,
      "grad_norm": 1.8131062710154953,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.878925621509552,
      "num_tokens": 468950019.0,
      "step": 13480
    },
    {
      "epoch": 2.503435468895079,
      "grad_norm": 1.6542078621560325,
      "learning_rate": 1e-06,
      "loss": 0.3454,
      "mean_token_accuracy": 0.878659188747406,
      "num_tokens": 468984972.0,
      "step": 13481
    },
    {
      "epoch": 2.5036211699164346,
      "grad_norm": 1.6706338611748464,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8919801712036133,
      "num_tokens": 469018089.0,
      "step": 13482
    },
    {
      "epoch": 2.5038068709377903,
      "grad_norm": 1.4490424186815627,
      "learning_rate": 1e-06,
      "loss": 0.3001,
      "mean_token_accuracy": 0.8931529521942139,
      "num_tokens": 469053566.0,
      "step": 13483
    },
    {
      "epoch": 2.503992571959146,
      "grad_norm": 1.6059536250019877,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.891037106513977,
      "num_tokens": 469084608.0,
      "step": 13484
    },
    {
      "epoch": 2.5041782729805013,
      "grad_norm": 1.4999090243665427,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8825214505195618,
      "num_tokens": 469124475.0,
      "step": 13485
    },
    {
      "epoch": 2.504363974001857,
      "grad_norm": 1.7076829825095252,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8837795853614807,
      "num_tokens": 469151481.0,
      "step": 13486
    },
    {
      "epoch": 2.5045496750232124,
      "grad_norm": 1.5525778541594522,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8837810158729553,
      "num_tokens": 469192315.0,
      "step": 13487
    },
    {
      "epoch": 2.504735376044568,
      "grad_norm": 1.5132582156410264,
      "learning_rate": 1e-06,
      "loss": 0.3344,
      "mean_token_accuracy": 0.8812135457992554,
      "num_tokens": 469232301.0,
      "step": 13488
    },
    {
      "epoch": 2.504921077065924,
      "grad_norm": 1.533995388886409,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8897214531898499,
      "num_tokens": 469268249.0,
      "step": 13489
    },
    {
      "epoch": 2.5051067780872796,
      "grad_norm": 1.5481319353149248,
      "learning_rate": 1e-06,
      "loss": 0.3022,
      "mean_token_accuracy": 0.8966001272201538,
      "num_tokens": 469301001.0,
      "step": 13490
    },
    {
      "epoch": 2.5052924791086353,
      "grad_norm": 1.705582253609482,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8847529292106628,
      "num_tokens": 469331077.0,
      "step": 13491
    },
    {
      "epoch": 2.5054781801299906,
      "grad_norm": 1.5551591766659203,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8845884799957275,
      "num_tokens": 469369650.0,
      "step": 13492
    },
    {
      "epoch": 2.5056638811513463,
      "grad_norm": 1.6069184424513154,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8742926716804504,
      "num_tokens": 469404964.0,
      "step": 13493
    },
    {
      "epoch": 2.505849582172702,
      "grad_norm": 1.5688632109697527,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.8892184495925903,
      "num_tokens": 469437327.0,
      "step": 13494
    },
    {
      "epoch": 2.5060352831940573,
      "grad_norm": 1.6121196895679677,
      "learning_rate": 1e-06,
      "loss": 0.2913,
      "mean_token_accuracy": 0.8988868594169617,
      "num_tokens": 469467829.0,
      "step": 13495
    },
    {
      "epoch": 2.506220984215413,
      "grad_norm": 1.5066538177575186,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8734104037284851,
      "num_tokens": 469503873.0,
      "step": 13496
    },
    {
      "epoch": 2.506406685236769,
      "grad_norm": 1.5195051074081782,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.878914475440979,
      "num_tokens": 469542019.0,
      "step": 13497
    },
    {
      "epoch": 2.5065923862581245,
      "grad_norm": 1.608717620774331,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8834596872329712,
      "num_tokens": 469573320.0,
      "step": 13498
    },
    {
      "epoch": 2.5067780872794803,
      "grad_norm": 1.5333800572295309,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8939305543899536,
      "num_tokens": 469606028.0,
      "step": 13499
    },
    {
      "epoch": 2.5069637883008355,
      "grad_norm": 1.5064066071233668,
      "learning_rate": 1e-06,
      "loss": 0.2828,
      "mean_token_accuracy": 0.8988240957260132,
      "num_tokens": 469640128.0,
      "step": 13500
    },
    {
      "epoch": 2.5071494893221913,
      "grad_norm": 1.500414820098238,
      "learning_rate": 1e-06,
      "loss": 0.2974,
      "mean_token_accuracy": 0.8954372406005859,
      "num_tokens": 469679978.0,
      "step": 13501
    },
    {
      "epoch": 2.507335190343547,
      "grad_norm": 1.6834192056938968,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8803743124008179,
      "num_tokens": 469711629.0,
      "step": 13502
    },
    {
      "epoch": 2.5075208913649023,
      "grad_norm": 1.6049615373290738,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8955682516098022,
      "num_tokens": 469745259.0,
      "step": 13503
    },
    {
      "epoch": 2.507706592386258,
      "grad_norm": 1.5679162901325907,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8837687969207764,
      "num_tokens": 469781853.0,
      "step": 13504
    },
    {
      "epoch": 2.5078922934076138,
      "grad_norm": 1.5782180179476002,
      "learning_rate": 1e-06,
      "loss": 0.3001,
      "mean_token_accuracy": 0.893582820892334,
      "num_tokens": 469815117.0,
      "step": 13505
    },
    {
      "epoch": 2.5080779944289695,
      "grad_norm": 1.4551978733579003,
      "learning_rate": 1e-06,
      "loss": 0.2677,
      "mean_token_accuracy": 0.9054104685783386,
      "num_tokens": 469853732.0,
      "step": 13506
    },
    {
      "epoch": 2.5082636954503252,
      "grad_norm": 1.5215007670403637,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8928459882736206,
      "num_tokens": 469891596.0,
      "step": 13507
    },
    {
      "epoch": 2.5084493964716805,
      "grad_norm": 1.525527498656118,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8870362043380737,
      "num_tokens": 469926241.0,
      "step": 13508
    },
    {
      "epoch": 2.5086350974930363,
      "grad_norm": 1.4926372930874043,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8851349353790283,
      "num_tokens": 469962264.0,
      "step": 13509
    },
    {
      "epoch": 2.508820798514392,
      "grad_norm": 1.50683403551756,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8823156356811523,
      "num_tokens": 469999858.0,
      "step": 13510
    },
    {
      "epoch": 2.5090064995357473,
      "grad_norm": 1.5774138118905172,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8970595598220825,
      "num_tokens": 470034853.0,
      "step": 13511
    },
    {
      "epoch": 2.509192200557103,
      "grad_norm": 1.4026951314797882,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.891112208366394,
      "num_tokens": 470075888.0,
      "step": 13512
    },
    {
      "epoch": 2.5093779015784587,
      "grad_norm": 1.405394201266296,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.882933497428894,
      "num_tokens": 470121052.0,
      "step": 13513
    },
    {
      "epoch": 2.5095636025998145,
      "grad_norm": 1.4644305379404698,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8906266689300537,
      "num_tokens": 470162611.0,
      "step": 13514
    },
    {
      "epoch": 2.5097493036211698,
      "grad_norm": 1.671195237396198,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8823950290679932,
      "num_tokens": 470192394.0,
      "step": 13515
    },
    {
      "epoch": 2.5099350046425255,
      "grad_norm": 1.748465681296367,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8789398074150085,
      "num_tokens": 470222697.0,
      "step": 13516
    },
    {
      "epoch": 2.5101207056638812,
      "grad_norm": 1.6838117143058597,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8702070713043213,
      "num_tokens": 470256149.0,
      "step": 13517
    },
    {
      "epoch": 2.5103064066852365,
      "grad_norm": 1.5523683200819383,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8914574980735779,
      "num_tokens": 470291160.0,
      "step": 13518
    },
    {
      "epoch": 2.5104921077065923,
      "grad_norm": 1.6496102342463745,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8955038189888,
      "num_tokens": 470321633.0,
      "step": 13519
    },
    {
      "epoch": 2.510677808727948,
      "grad_norm": 1.4921407638513458,
      "learning_rate": 1e-06,
      "loss": 0.2752,
      "mean_token_accuracy": 0.9002392292022705,
      "num_tokens": 470356050.0,
      "step": 13520
    },
    {
      "epoch": 2.5108635097493037,
      "grad_norm": 1.4896573356990226,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8917378187179565,
      "num_tokens": 470393455.0,
      "step": 13521
    },
    {
      "epoch": 2.5110492107706595,
      "grad_norm": 1.6278718317786267,
      "learning_rate": 1e-06,
      "loss": 0.2845,
      "mean_token_accuracy": 0.8965610265731812,
      "num_tokens": 470424913.0,
      "step": 13522
    },
    {
      "epoch": 2.5112349117920147,
      "grad_norm": 1.5091381848575371,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8851643800735474,
      "num_tokens": 470464154.0,
      "step": 13523
    },
    {
      "epoch": 2.5114206128133705,
      "grad_norm": 1.64623607282296,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8908843994140625,
      "num_tokens": 470493358.0,
      "step": 13524
    },
    {
      "epoch": 2.511606313834726,
      "grad_norm": 1.7086754403030862,
      "learning_rate": 1e-06,
      "loss": 0.3611,
      "mean_token_accuracy": 0.8768059611320496,
      "num_tokens": 470528747.0,
      "step": 13525
    },
    {
      "epoch": 2.5117920148560815,
      "grad_norm": 1.518316334657457,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8887057900428772,
      "num_tokens": 470568589.0,
      "step": 13526
    },
    {
      "epoch": 2.5119777158774372,
      "grad_norm": 1.4836662168619643,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8921215534210205,
      "num_tokens": 470606299.0,
      "step": 13527
    },
    {
      "epoch": 2.512163416898793,
      "grad_norm": 1.6671773157356515,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8708803653717041,
      "num_tokens": 470638720.0,
      "step": 13528
    },
    {
      "epoch": 2.5123491179201487,
      "grad_norm": 1.5432951508438715,
      "learning_rate": 1e-06,
      "loss": 0.296,
      "mean_token_accuracy": 0.8928234577178955,
      "num_tokens": 470676658.0,
      "step": 13529
    },
    {
      "epoch": 2.5125348189415044,
      "grad_norm": 1.5689881940948935,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8776683807373047,
      "num_tokens": 470718505.0,
      "step": 13530
    },
    {
      "epoch": 2.5127205199628597,
      "grad_norm": 1.5855132928516629,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8842796683311462,
      "num_tokens": 470752271.0,
      "step": 13531
    },
    {
      "epoch": 2.5129062209842155,
      "grad_norm": 1.6918940524156911,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.883564829826355,
      "num_tokens": 470785763.0,
      "step": 13532
    },
    {
      "epoch": 2.513091922005571,
      "grad_norm": 1.680302031682499,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8872289061546326,
      "num_tokens": 470815683.0,
      "step": 13533
    },
    {
      "epoch": 2.5132776230269265,
      "grad_norm": 1.497101255671581,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8892941474914551,
      "num_tokens": 470858037.0,
      "step": 13534
    },
    {
      "epoch": 2.513463324048282,
      "grad_norm": 1.5211983031654732,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8892664313316345,
      "num_tokens": 470892058.0,
      "step": 13535
    },
    {
      "epoch": 2.513649025069638,
      "grad_norm": 1.6151211970528045,
      "learning_rate": 1e-06,
      "loss": 0.3721,
      "mean_token_accuracy": 0.8718637824058533,
      "num_tokens": 470928715.0,
      "step": 13536
    },
    {
      "epoch": 2.5138347260909937,
      "grad_norm": 1.6289960982661207,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8922650814056396,
      "num_tokens": 470960115.0,
      "step": 13537
    },
    {
      "epoch": 2.514020427112349,
      "grad_norm": 1.8552360963953227,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8787837028503418,
      "num_tokens": 470987182.0,
      "step": 13538
    },
    {
      "epoch": 2.5142061281337047,
      "grad_norm": 1.5366821559380712,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8813890218734741,
      "num_tokens": 471025526.0,
      "step": 13539
    },
    {
      "epoch": 2.5143918291550604,
      "grad_norm": 1.5733935118908828,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8858305215835571,
      "num_tokens": 471062199.0,
      "step": 13540
    },
    {
      "epoch": 2.5145775301764157,
      "grad_norm": 1.664599650997917,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8821039795875549,
      "num_tokens": 471094811.0,
      "step": 13541
    },
    {
      "epoch": 2.5147632311977715,
      "grad_norm": 1.5286461468015318,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8822385668754578,
      "num_tokens": 471131815.0,
      "step": 13542
    },
    {
      "epoch": 2.514948932219127,
      "grad_norm": 1.6026069519785382,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8760625123977661,
      "num_tokens": 471168653.0,
      "step": 13543
    },
    {
      "epoch": 2.515134633240483,
      "grad_norm": 1.5922742967362766,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8894990682601929,
      "num_tokens": 471201384.0,
      "step": 13544
    },
    {
      "epoch": 2.5153203342618387,
      "grad_norm": 1.424692970459302,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8906956911087036,
      "num_tokens": 471241451.0,
      "step": 13545
    },
    {
      "epoch": 2.515506035283194,
      "grad_norm": 1.5574714213188336,
      "learning_rate": 1e-06,
      "loss": 0.2811,
      "mean_token_accuracy": 0.9002249240875244,
      "num_tokens": 471272421.0,
      "step": 13546
    },
    {
      "epoch": 2.5156917363045497,
      "grad_norm": 1.5214164307153333,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8846246004104614,
      "num_tokens": 471307815.0,
      "step": 13547
    },
    {
      "epoch": 2.5158774373259054,
      "grad_norm": 1.6020291359283547,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8734155893325806,
      "num_tokens": 471342118.0,
      "step": 13548
    },
    {
      "epoch": 2.5160631383472607,
      "grad_norm": 1.492267673835244,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8860293626785278,
      "num_tokens": 471379683.0,
      "step": 13549
    },
    {
      "epoch": 2.5162488393686164,
      "grad_norm": 1.7968096375044833,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8733783960342407,
      "num_tokens": 471408784.0,
      "step": 13550
    },
    {
      "epoch": 2.516434540389972,
      "grad_norm": 1.5406556353662768,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8844764232635498,
      "num_tokens": 471444343.0,
      "step": 13551
    },
    {
      "epoch": 2.516620241411328,
      "grad_norm": 1.702370592912435,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8761305809020996,
      "num_tokens": 471478421.0,
      "step": 13552
    },
    {
      "epoch": 2.5168059424326836,
      "grad_norm": 1.539625855385513,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8804436326026917,
      "num_tokens": 471514428.0,
      "step": 13553
    },
    {
      "epoch": 2.516991643454039,
      "grad_norm": 1.5855405144662886,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8888767957687378,
      "num_tokens": 471547952.0,
      "step": 13554
    },
    {
      "epoch": 2.5171773444753947,
      "grad_norm": 1.590182311226182,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8771228194236755,
      "num_tokens": 471581874.0,
      "step": 13555
    },
    {
      "epoch": 2.5173630454967504,
      "grad_norm": 1.6393262583730093,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8777581453323364,
      "num_tokens": 471618775.0,
      "step": 13556
    },
    {
      "epoch": 2.5175487465181057,
      "grad_norm": 1.7267619269475756,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8815672397613525,
      "num_tokens": 471649255.0,
      "step": 13557
    },
    {
      "epoch": 2.5177344475394614,
      "grad_norm": 1.7034514411468518,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8778212070465088,
      "num_tokens": 471685595.0,
      "step": 13558
    },
    {
      "epoch": 2.517920148560817,
      "grad_norm": 1.5642457077571241,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8819640874862671,
      "num_tokens": 471722603.0,
      "step": 13559
    },
    {
      "epoch": 2.518105849582173,
      "grad_norm": 1.5519498616497964,
      "learning_rate": 1e-06,
      "loss": 0.2936,
      "mean_token_accuracy": 0.8966416716575623,
      "num_tokens": 471754825.0,
      "step": 13560
    },
    {
      "epoch": 2.518291550603528,
      "grad_norm": 1.555836947894236,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.870037853717804,
      "num_tokens": 471791145.0,
      "step": 13561
    },
    {
      "epoch": 2.518477251624884,
      "grad_norm": 1.5640403758599255,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8888530731201172,
      "num_tokens": 471826913.0,
      "step": 13562
    },
    {
      "epoch": 2.5186629526462396,
      "grad_norm": 1.5039126947162231,
      "learning_rate": 1e-06,
      "loss": 0.2741,
      "mean_token_accuracy": 0.9024062156677246,
      "num_tokens": 471863697.0,
      "step": 13563
    },
    {
      "epoch": 2.518848653667595,
      "grad_norm": 1.6647535706069523,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8883589506149292,
      "num_tokens": 471894556.0,
      "step": 13564
    },
    {
      "epoch": 2.5190343546889506,
      "grad_norm": 1.5870445135299798,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8875923156738281,
      "num_tokens": 471927633.0,
      "step": 13565
    },
    {
      "epoch": 2.5192200557103064,
      "grad_norm": 1.830655764398596,
      "learning_rate": 1e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.869597315788269,
      "num_tokens": 471960163.0,
      "step": 13566
    },
    {
      "epoch": 2.519405756731662,
      "grad_norm": 1.5623656827411578,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8863811492919922,
      "num_tokens": 471995095.0,
      "step": 13567
    },
    {
      "epoch": 2.519591457753018,
      "grad_norm": 4.692633273068042,
      "learning_rate": 1e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8687881231307983,
      "num_tokens": 472026470.0,
      "step": 13568
    },
    {
      "epoch": 2.519777158774373,
      "grad_norm": 1.5058209262121462,
      "learning_rate": 1e-06,
      "loss": 0.2726,
      "mean_token_accuracy": 0.9014173746109009,
      "num_tokens": 472060116.0,
      "step": 13569
    },
    {
      "epoch": 2.519962859795729,
      "grad_norm": 1.7499347733328139,
      "learning_rate": 1e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8718686699867249,
      "num_tokens": 472091624.0,
      "step": 13570
    },
    {
      "epoch": 2.5201485608170846,
      "grad_norm": 1.6403440611330853,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8775631189346313,
      "num_tokens": 472127427.0,
      "step": 13571
    },
    {
      "epoch": 2.52033426183844,
      "grad_norm": 1.610901534694951,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8873739838600159,
      "num_tokens": 472158759.0,
      "step": 13572
    },
    {
      "epoch": 2.5205199628597956,
      "grad_norm": 1.4236070309027866,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8873124718666077,
      "num_tokens": 472203185.0,
      "step": 13573
    },
    {
      "epoch": 2.5207056638811514,
      "grad_norm": 1.6244099345585883,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8843802213668823,
      "num_tokens": 472237089.0,
      "step": 13574
    },
    {
      "epoch": 2.520891364902507,
      "grad_norm": 1.6269147321551045,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8864322900772095,
      "num_tokens": 472267909.0,
      "step": 13575
    },
    {
      "epoch": 2.521077065923863,
      "grad_norm": 1.7224555008149502,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8849847316741943,
      "num_tokens": 472299734.0,
      "step": 13576
    },
    {
      "epoch": 2.521262766945218,
      "grad_norm": 1.6600508896081616,
      "learning_rate": 1e-06,
      "loss": 0.3179,
      "mean_token_accuracy": 0.8886846303939819,
      "num_tokens": 472332406.0,
      "step": 13577
    },
    {
      "epoch": 2.521448467966574,
      "grad_norm": 1.475921745196991,
      "learning_rate": 1e-06,
      "loss": 0.2826,
      "mean_token_accuracy": 0.8984873294830322,
      "num_tokens": 472366702.0,
      "step": 13578
    },
    {
      "epoch": 2.5216341689879296,
      "grad_norm": 1.6512580256661857,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8786088228225708,
      "num_tokens": 472397954.0,
      "step": 13579
    },
    {
      "epoch": 2.521819870009285,
      "grad_norm": 1.7384103085904508,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8876210451126099,
      "num_tokens": 472425295.0,
      "step": 13580
    },
    {
      "epoch": 2.5220055710306406,
      "grad_norm": 1.571852459157528,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8829817771911621,
      "num_tokens": 472461381.0,
      "step": 13581
    },
    {
      "epoch": 2.5221912720519963,
      "grad_norm": 1.4495825879653237,
      "learning_rate": 1e-06,
      "loss": 0.301,
      "mean_token_accuracy": 0.8933802843093872,
      "num_tokens": 472501239.0,
      "step": 13582
    },
    {
      "epoch": 2.522376973073352,
      "grad_norm": 1.5148251678013593,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8843077421188354,
      "num_tokens": 472536978.0,
      "step": 13583
    },
    {
      "epoch": 2.5225626740947074,
      "grad_norm": 1.5929082317716836,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8807787895202637,
      "num_tokens": 472574185.0,
      "step": 13584
    },
    {
      "epoch": 2.522748375116063,
      "grad_norm": 1.523635542090295,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.8944380283355713,
      "num_tokens": 472607599.0,
      "step": 13585
    },
    {
      "epoch": 2.522934076137419,
      "grad_norm": 1.4717860901250095,
      "learning_rate": 1e-06,
      "loss": 0.2968,
      "mean_token_accuracy": 0.8949592709541321,
      "num_tokens": 472643121.0,
      "step": 13586
    },
    {
      "epoch": 2.523119777158774,
      "grad_norm": 1.5101859488419829,
      "learning_rate": 1e-06,
      "loss": 0.2903,
      "mean_token_accuracy": 0.8964256048202515,
      "num_tokens": 472674511.0,
      "step": 13587
    },
    {
      "epoch": 2.52330547818013,
      "grad_norm": 1.6399962067800282,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8912431001663208,
      "num_tokens": 472712282.0,
      "step": 13588
    },
    {
      "epoch": 2.5234911792014856,
      "grad_norm": 1.503145267227208,
      "learning_rate": 1e-06,
      "loss": 0.2737,
      "mean_token_accuracy": 0.9020426273345947,
      "num_tokens": 472746536.0,
      "step": 13589
    },
    {
      "epoch": 2.5236768802228413,
      "grad_norm": 1.5189848900058782,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8904301524162292,
      "num_tokens": 472781759.0,
      "step": 13590
    },
    {
      "epoch": 2.523862581244197,
      "grad_norm": 1.6417128101909757,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8783688545227051,
      "num_tokens": 472814596.0,
      "step": 13591
    },
    {
      "epoch": 2.5240482822655523,
      "grad_norm": 1.599488350614163,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8733991384506226,
      "num_tokens": 472849983.0,
      "step": 13592
    },
    {
      "epoch": 2.524233983286908,
      "grad_norm": 1.6883901747544705,
      "learning_rate": 1e-06,
      "loss": 0.2773,
      "mean_token_accuracy": 0.8994295001029968,
      "num_tokens": 472878708.0,
      "step": 13593
    },
    {
      "epoch": 2.524419684308264,
      "grad_norm": 1.5984765879991767,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.888862669467926,
      "num_tokens": 472912785.0,
      "step": 13594
    },
    {
      "epoch": 2.524605385329619,
      "grad_norm": 1.5834931211952494,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8936800360679626,
      "num_tokens": 472951960.0,
      "step": 13595
    },
    {
      "epoch": 2.524791086350975,
      "grad_norm": 1.6986577259725852,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8921584486961365,
      "num_tokens": 472982005.0,
      "step": 13596
    },
    {
      "epoch": 2.5249767873723306,
      "grad_norm": 1.4620066067116022,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8906888961791992,
      "num_tokens": 473018506.0,
      "step": 13597
    },
    {
      "epoch": 2.5251624883936863,
      "grad_norm": 1.6296092590226061,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8857645392417908,
      "num_tokens": 473053227.0,
      "step": 13598
    },
    {
      "epoch": 2.525348189415042,
      "grad_norm": 1.7943142733454778,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8856053352355957,
      "num_tokens": 473077448.0,
      "step": 13599
    },
    {
      "epoch": 2.5255338904363973,
      "grad_norm": 1.5384308105800737,
      "learning_rate": 1e-06,
      "loss": 0.2824,
      "mean_token_accuracy": 0.9012241363525391,
      "num_tokens": 473109252.0,
      "step": 13600
    },
    {
      "epoch": 2.525719591457753,
      "grad_norm": 1.6139896310149142,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.872555136680603,
      "num_tokens": 473144631.0,
      "step": 13601
    },
    {
      "epoch": 2.5259052924791088,
      "grad_norm": 1.4829013991898958,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8943420052528381,
      "num_tokens": 473179710.0,
      "step": 13602
    },
    {
      "epoch": 2.526090993500464,
      "grad_norm": 1.5386513466922895,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8881250619888306,
      "num_tokens": 473214324.0,
      "step": 13603
    },
    {
      "epoch": 2.52627669452182,
      "grad_norm": 1.4044570597900303,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8952031135559082,
      "num_tokens": 473255349.0,
      "step": 13604
    },
    {
      "epoch": 2.5264623955431755,
      "grad_norm": 1.5734501818162743,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8835400342941284,
      "num_tokens": 473291983.0,
      "step": 13605
    },
    {
      "epoch": 2.5266480965645313,
      "grad_norm": 1.5477680591037908,
      "learning_rate": 1e-06,
      "loss": 0.3128,
      "mean_token_accuracy": 0.8911072611808777,
      "num_tokens": 473326684.0,
      "step": 13606
    },
    {
      "epoch": 2.526833797585887,
      "grad_norm": 1.4145895028471005,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.8925870060920715,
      "num_tokens": 473367205.0,
      "step": 13607
    },
    {
      "epoch": 2.5270194986072423,
      "grad_norm": 1.718392676309201,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8816951513290405,
      "num_tokens": 473401002.0,
      "step": 13608
    },
    {
      "epoch": 2.527205199628598,
      "grad_norm": 1.706653631945988,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8770028948783875,
      "num_tokens": 473434334.0,
      "step": 13609
    },
    {
      "epoch": 2.5273909006499533,
      "grad_norm": 1.5291468789918279,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8945589661598206,
      "num_tokens": 473469372.0,
      "step": 13610
    },
    {
      "epoch": 2.527576601671309,
      "grad_norm": 1.4847287143474894,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8834459781646729,
      "num_tokens": 473508731.0,
      "step": 13611
    },
    {
      "epoch": 2.5277623026926648,
      "grad_norm": 1.5864267741842655,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8775365352630615,
      "num_tokens": 473545301.0,
      "step": 13612
    },
    {
      "epoch": 2.5279480037140205,
      "grad_norm": 1.5485221237673619,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8790364861488342,
      "num_tokens": 473578769.0,
      "step": 13613
    },
    {
      "epoch": 2.5281337047353762,
      "grad_norm": 1.5654751582102822,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8681837320327759,
      "num_tokens": 473621467.0,
      "step": 13614
    },
    {
      "epoch": 2.5283194057567315,
      "grad_norm": 1.66460128429932,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8803224563598633,
      "num_tokens": 473654200.0,
      "step": 13615
    },
    {
      "epoch": 2.5285051067780873,
      "grad_norm": 1.5571872085135297,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8794285655021667,
      "num_tokens": 473692722.0,
      "step": 13616
    },
    {
      "epoch": 2.528690807799443,
      "grad_norm": 1.5541410777818352,
      "learning_rate": 1e-06,
      "loss": 0.299,
      "mean_token_accuracy": 0.8913635611534119,
      "num_tokens": 473726834.0,
      "step": 13617
    },
    {
      "epoch": 2.5288765088207983,
      "grad_norm": 1.4857793234562702,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8877349495887756,
      "num_tokens": 473766903.0,
      "step": 13618
    },
    {
      "epoch": 2.529062209842154,
      "grad_norm": 1.576663314056658,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8873172998428345,
      "num_tokens": 473802152.0,
      "step": 13619
    },
    {
      "epoch": 2.5292479108635098,
      "grad_norm": 1.7867694988935032,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8907256126403809,
      "num_tokens": 473830878.0,
      "step": 13620
    },
    {
      "epoch": 2.5294336118848655,
      "grad_norm": 1.5616025167746115,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8789767026901245,
      "num_tokens": 473867242.0,
      "step": 13621
    },
    {
      "epoch": 2.529619312906221,
      "grad_norm": 1.5464737641133632,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8968397378921509,
      "num_tokens": 473900404.0,
      "step": 13622
    },
    {
      "epoch": 2.5298050139275765,
      "grad_norm": 1.547005862525705,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8862252235412598,
      "num_tokens": 473939343.0,
      "step": 13623
    },
    {
      "epoch": 2.5299907149489322,
      "grad_norm": 1.732863238261496,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8796989917755127,
      "num_tokens": 473971465.0,
      "step": 13624
    },
    {
      "epoch": 2.530176415970288,
      "grad_norm": 1.535383464549579,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8840024471282959,
      "num_tokens": 474007185.0,
      "step": 13625
    },
    {
      "epoch": 2.5303621169916433,
      "grad_norm": 1.563568974257852,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8944039940834045,
      "num_tokens": 474039257.0,
      "step": 13626
    },
    {
      "epoch": 2.530547818012999,
      "grad_norm": 1.4762153888864757,
      "learning_rate": 1e-06,
      "loss": 0.2864,
      "mean_token_accuracy": 0.9006205201148987,
      "num_tokens": 474075966.0,
      "step": 13627
    },
    {
      "epoch": 2.5307335190343547,
      "grad_norm": 1.8219205389917439,
      "learning_rate": 1e-06,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8921573758125305,
      "num_tokens": 474103223.0,
      "step": 13628
    },
    {
      "epoch": 2.5309192200557105,
      "grad_norm": 1.454367742831921,
      "learning_rate": 1e-06,
      "loss": 0.256,
      "mean_token_accuracy": 0.907866358757019,
      "num_tokens": 474138166.0,
      "step": 13629
    },
    {
      "epoch": 2.531104921077066,
      "grad_norm": 1.7017500521402678,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8806909322738647,
      "num_tokens": 474167675.0,
      "step": 13630
    },
    {
      "epoch": 2.5312906220984215,
      "grad_norm": 1.5434851296731749,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8903517127037048,
      "num_tokens": 474206121.0,
      "step": 13631
    },
    {
      "epoch": 2.531476323119777,
      "grad_norm": 1.6017751789136945,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8898698687553406,
      "num_tokens": 474238914.0,
      "step": 13632
    },
    {
      "epoch": 2.5316620241411325,
      "grad_norm": 1.422096763836192,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8878917694091797,
      "num_tokens": 474283522.0,
      "step": 13633
    },
    {
      "epoch": 2.5318477251624882,
      "grad_norm": 1.538839709212677,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.8931771516799927,
      "num_tokens": 474316129.0,
      "step": 13634
    },
    {
      "epoch": 2.532033426183844,
      "grad_norm": 1.754901899494865,
      "learning_rate": 1e-06,
      "loss": 0.4002,
      "mean_token_accuracy": 0.866428017616272,
      "num_tokens": 474346359.0,
      "step": 13635
    },
    {
      "epoch": 2.5322191272051997,
      "grad_norm": 1.4820285212495738,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8885260820388794,
      "num_tokens": 474383143.0,
      "step": 13636
    },
    {
      "epoch": 2.5324048282265554,
      "grad_norm": 1.6396950744011813,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8881363868713379,
      "num_tokens": 474421870.0,
      "step": 13637
    },
    {
      "epoch": 2.5325905292479107,
      "grad_norm": 1.7103551395733316,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8803290128707886,
      "num_tokens": 474452060.0,
      "step": 13638
    },
    {
      "epoch": 2.5327762302692665,
      "grad_norm": 1.4393384536379294,
      "learning_rate": 1e-06,
      "loss": 0.2914,
      "mean_token_accuracy": 0.8935927748680115,
      "num_tokens": 474488697.0,
      "step": 13639
    },
    {
      "epoch": 2.532961931290622,
      "grad_norm": 1.4383907288127034,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8894139528274536,
      "num_tokens": 474527212.0,
      "step": 13640
    },
    {
      "epoch": 2.5331476323119775,
      "grad_norm": 1.508471529204003,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8943457007408142,
      "num_tokens": 474560697.0,
      "step": 13641
    },
    {
      "epoch": 2.533333333333333,
      "grad_norm": 1.5686434748851257,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8901054859161377,
      "num_tokens": 474595241.0,
      "step": 13642
    },
    {
      "epoch": 2.533519034354689,
      "grad_norm": 1.5101755739932683,
      "learning_rate": 1e-06,
      "loss": 0.2945,
      "mean_token_accuracy": 0.8957316279411316,
      "num_tokens": 474628137.0,
      "step": 13643
    },
    {
      "epoch": 2.5337047353760447,
      "grad_norm": 1.5509360183745255,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8761557340621948,
      "num_tokens": 474668168.0,
      "step": 13644
    },
    {
      "epoch": 2.5338904363974004,
      "grad_norm": 1.8165967706440724,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8733401298522949,
      "num_tokens": 474696846.0,
      "step": 13645
    },
    {
      "epoch": 2.5340761374187557,
      "grad_norm": 1.5784498664336781,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8854547142982483,
      "num_tokens": 474735840.0,
      "step": 13646
    },
    {
      "epoch": 2.5342618384401114,
      "grad_norm": 1.697395962229471,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.8961782455444336,
      "num_tokens": 474767481.0,
      "step": 13647
    },
    {
      "epoch": 2.534447539461467,
      "grad_norm": 1.7659370189613814,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8806971311569214,
      "num_tokens": 474799186.0,
      "step": 13648
    },
    {
      "epoch": 2.5346332404828225,
      "grad_norm": 1.5703115006511967,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8951963782310486,
      "num_tokens": 474832196.0,
      "step": 13649
    },
    {
      "epoch": 2.534818941504178,
      "grad_norm": 1.6269054930759705,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.889326274394989,
      "num_tokens": 474867347.0,
      "step": 13650
    },
    {
      "epoch": 2.535004642525534,
      "grad_norm": 1.5193561875436,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8849624395370483,
      "num_tokens": 474908609.0,
      "step": 13651
    },
    {
      "epoch": 2.5351903435468897,
      "grad_norm": 1.758244484251398,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8936148881912231,
      "num_tokens": 474939089.0,
      "step": 13652
    },
    {
      "epoch": 2.5353760445682454,
      "grad_norm": 1.7352096293735293,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8993090391159058,
      "num_tokens": 474965811.0,
      "step": 13653
    },
    {
      "epoch": 2.5355617455896007,
      "grad_norm": 1.6285943334202875,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.884710967540741,
      "num_tokens": 474998390.0,
      "step": 13654
    },
    {
      "epoch": 2.5357474466109564,
      "grad_norm": 1.6136386102548486,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8806724548339844,
      "num_tokens": 475036207.0,
      "step": 13655
    },
    {
      "epoch": 2.5359331476323117,
      "grad_norm": 1.6230722082178184,
      "learning_rate": 1e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8697075247764587,
      "num_tokens": 475072219.0,
      "step": 13656
    },
    {
      "epoch": 2.5361188486536674,
      "grad_norm": 1.8475557811355123,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8855365514755249,
      "num_tokens": 475102135.0,
      "step": 13657
    },
    {
      "epoch": 2.536304549675023,
      "grad_norm": 1.5445002099187548,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8924479484558105,
      "num_tokens": 475132132.0,
      "step": 13658
    },
    {
      "epoch": 2.536490250696379,
      "grad_norm": 1.579332045377239,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8843016624450684,
      "num_tokens": 475166517.0,
      "step": 13659
    },
    {
      "epoch": 2.5366759517177346,
      "grad_norm": 1.5414898981791794,
      "learning_rate": 1e-06,
      "loss": 0.2774,
      "mean_token_accuracy": 0.9014606475830078,
      "num_tokens": 475199770.0,
      "step": 13660
    },
    {
      "epoch": 2.53686165273909,
      "grad_norm": 1.692985486029145,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8782414197921753,
      "num_tokens": 475229824.0,
      "step": 13661
    },
    {
      "epoch": 2.5370473537604457,
      "grad_norm": 1.543437088134051,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8827030062675476,
      "num_tokens": 475263880.0,
      "step": 13662
    },
    {
      "epoch": 2.5372330547818014,
      "grad_norm": 1.5890333782933885,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8848171234130859,
      "num_tokens": 475298711.0,
      "step": 13663
    },
    {
      "epoch": 2.5374187558031567,
      "grad_norm": 1.6345641964858884,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.873934805393219,
      "num_tokens": 475333763.0,
      "step": 13664
    },
    {
      "epoch": 2.5376044568245124,
      "grad_norm": 1.4537269918669775,
      "learning_rate": 1e-06,
      "loss": 0.2589,
      "mean_token_accuracy": 0.9056757688522339,
      "num_tokens": 475367682.0,
      "step": 13665
    },
    {
      "epoch": 2.537790157845868,
      "grad_norm": 1.5257545016170402,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8899900913238525,
      "num_tokens": 475403348.0,
      "step": 13666
    },
    {
      "epoch": 2.537975858867224,
      "grad_norm": 1.651506014541471,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8813602924346924,
      "num_tokens": 475436673.0,
      "step": 13667
    },
    {
      "epoch": 2.5381615598885796,
      "grad_norm": 1.5467940147144237,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8828022480010986,
      "num_tokens": 475470722.0,
      "step": 13668
    },
    {
      "epoch": 2.538347260909935,
      "grad_norm": 1.49190959686395,
      "learning_rate": 1e-06,
      "loss": 0.3911,
      "mean_token_accuracy": 0.8694665431976318,
      "num_tokens": 475515260.0,
      "step": 13669
    },
    {
      "epoch": 2.5385329619312906,
      "grad_norm": 1.5116572007336704,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.8971657752990723,
      "num_tokens": 475549357.0,
      "step": 13670
    },
    {
      "epoch": 2.5387186629526464,
      "grad_norm": 1.5361767537331743,
      "learning_rate": 1e-06,
      "loss": 0.3475,
      "mean_token_accuracy": 0.8811558485031128,
      "num_tokens": 475587262.0,
      "step": 13671
    },
    {
      "epoch": 2.5389043639740017,
      "grad_norm": 1.7286625437280965,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8863950371742249,
      "num_tokens": 475614906.0,
      "step": 13672
    },
    {
      "epoch": 2.5390900649953574,
      "grad_norm": 1.460820953575143,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8913524150848389,
      "num_tokens": 475655257.0,
      "step": 13673
    },
    {
      "epoch": 2.539275766016713,
      "grad_norm": 1.6235708374892395,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8839042782783508,
      "num_tokens": 475688650.0,
      "step": 13674
    },
    {
      "epoch": 2.539461467038069,
      "grad_norm": 1.5990120175588471,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8877798914909363,
      "num_tokens": 475719416.0,
      "step": 13675
    },
    {
      "epoch": 2.5396471680594246,
      "grad_norm": 1.6363436472627626,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8810657262802124,
      "num_tokens": 475753230.0,
      "step": 13676
    },
    {
      "epoch": 2.53983286908078,
      "grad_norm": 1.7646170116336588,
      "learning_rate": 1e-06,
      "loss": 0.3705,
      "mean_token_accuracy": 0.8729560971260071,
      "num_tokens": 475788241.0,
      "step": 13677
    },
    {
      "epoch": 2.5400185701021356,
      "grad_norm": 1.6745868061441271,
      "learning_rate": 1e-06,
      "loss": 0.3419,
      "mean_token_accuracy": 0.8858569264411926,
      "num_tokens": 475822017.0,
      "step": 13678
    },
    {
      "epoch": 2.5402042711234913,
      "grad_norm": 1.678718347084098,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8902580738067627,
      "num_tokens": 475853427.0,
      "step": 13679
    },
    {
      "epoch": 2.5403899721448466,
      "grad_norm": 1.6440784647475102,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8795769214630127,
      "num_tokens": 475888173.0,
      "step": 13680
    },
    {
      "epoch": 2.5405756731662024,
      "grad_norm": 1.5708710821822283,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8825969696044922,
      "num_tokens": 475924104.0,
      "step": 13681
    },
    {
      "epoch": 2.540761374187558,
      "grad_norm": 1.6913098921897678,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8923681974411011,
      "num_tokens": 475957142.0,
      "step": 13682
    },
    {
      "epoch": 2.540947075208914,
      "grad_norm": 1.5626437857428401,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8876849412918091,
      "num_tokens": 475992244.0,
      "step": 13683
    },
    {
      "epoch": 2.541132776230269,
      "grad_norm": 1.6478246836112778,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.888735830783844,
      "num_tokens": 476023422.0,
      "step": 13684
    },
    {
      "epoch": 2.541318477251625,
      "grad_norm": 1.5376058471832073,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8850967884063721,
      "num_tokens": 476059724.0,
      "step": 13685
    },
    {
      "epoch": 2.5415041782729806,
      "grad_norm": 1.421105593661227,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8940024971961975,
      "num_tokens": 476099733.0,
      "step": 13686
    },
    {
      "epoch": 2.541689879294336,
      "grad_norm": 1.4525167803274615,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.869479775428772,
      "num_tokens": 476143340.0,
      "step": 13687
    },
    {
      "epoch": 2.5418755803156916,
      "grad_norm": 1.4944164582281843,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8873939514160156,
      "num_tokens": 476184850.0,
      "step": 13688
    },
    {
      "epoch": 2.5420612813370473,
      "grad_norm": 1.6001038930781353,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.8840539455413818,
      "num_tokens": 476218791.0,
      "step": 13689
    },
    {
      "epoch": 2.542246982358403,
      "grad_norm": 1.603773446952163,
      "learning_rate": 1e-06,
      "loss": 0.3174,
      "mean_token_accuracy": 0.8932861089706421,
      "num_tokens": 476252131.0,
      "step": 13690
    },
    {
      "epoch": 2.542432683379759,
      "grad_norm": 1.489153307969241,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8797522783279419,
      "num_tokens": 476290137.0,
      "step": 13691
    },
    {
      "epoch": 2.542618384401114,
      "grad_norm": 1.6138443595575147,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8846145272254944,
      "num_tokens": 476322957.0,
      "step": 13692
    },
    {
      "epoch": 2.54280408542247,
      "grad_norm": 1.4322265626691189,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8814138770103455,
      "num_tokens": 476363585.0,
      "step": 13693
    },
    {
      "epoch": 2.5429897864438256,
      "grad_norm": 1.6306930501887833,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8898763060569763,
      "num_tokens": 476392957.0,
      "step": 13694
    },
    {
      "epoch": 2.543175487465181,
      "grad_norm": 1.5561958701362277,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8860505819320679,
      "num_tokens": 476429014.0,
      "step": 13695
    },
    {
      "epoch": 2.5433611884865366,
      "grad_norm": 1.5613102605724931,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8929048180580139,
      "num_tokens": 476459559.0,
      "step": 13696
    },
    {
      "epoch": 2.5435468895078923,
      "grad_norm": 1.6775713020658958,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8750805258750916,
      "num_tokens": 476494772.0,
      "step": 13697
    },
    {
      "epoch": 2.543732590529248,
      "grad_norm": 1.5514969078902805,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8894118070602417,
      "num_tokens": 476530602.0,
      "step": 13698
    },
    {
      "epoch": 2.543918291550604,
      "grad_norm": 1.586200309029837,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8994836211204529,
      "num_tokens": 476563059.0,
      "step": 13699
    },
    {
      "epoch": 2.544103992571959,
      "grad_norm": 1.6020865558889976,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8826683759689331,
      "num_tokens": 476598599.0,
      "step": 13700
    },
    {
      "epoch": 2.544289693593315,
      "grad_norm": 1.6018859020800942,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8878316879272461,
      "num_tokens": 476630823.0,
      "step": 13701
    },
    {
      "epoch": 2.5444753946146705,
      "grad_norm": 1.7191498690925322,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8828500509262085,
      "num_tokens": 476667206.0,
      "step": 13702
    },
    {
      "epoch": 2.544661095636026,
      "grad_norm": 1.7141222283700237,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8847216963768005,
      "num_tokens": 476700109.0,
      "step": 13703
    },
    {
      "epoch": 2.5448467966573816,
      "grad_norm": 1.8121312576023683,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8964323997497559,
      "num_tokens": 476725379.0,
      "step": 13704
    },
    {
      "epoch": 2.5450324976787373,
      "grad_norm": 1.5142401370540974,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8912774324417114,
      "num_tokens": 476762024.0,
      "step": 13705
    },
    {
      "epoch": 2.545218198700093,
      "grad_norm": 1.728392854110492,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8860481977462769,
      "num_tokens": 476795804.0,
      "step": 13706
    },
    {
      "epoch": 2.5454038997214483,
      "grad_norm": 1.5041290262300075,
      "learning_rate": 1e-06,
      "loss": 0.2934,
      "mean_token_accuracy": 0.8973326683044434,
      "num_tokens": 476828933.0,
      "step": 13707
    },
    {
      "epoch": 2.545589600742804,
      "grad_norm": 1.4254459629079306,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8839130401611328,
      "num_tokens": 476872281.0,
      "step": 13708
    },
    {
      "epoch": 2.54577530176416,
      "grad_norm": 1.6730397868602884,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8743147253990173,
      "num_tokens": 476904277.0,
      "step": 13709
    },
    {
      "epoch": 2.545961002785515,
      "grad_norm": 1.6840752889402528,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8764507174491882,
      "num_tokens": 476937228.0,
      "step": 13710
    },
    {
      "epoch": 2.546146703806871,
      "grad_norm": 1.577107832928635,
      "learning_rate": 1e-06,
      "loss": 0.2902,
      "mean_token_accuracy": 0.8960564732551575,
      "num_tokens": 476971645.0,
      "step": 13711
    },
    {
      "epoch": 2.5463324048282265,
      "grad_norm": 1.6065222635471248,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8891435861587524,
      "num_tokens": 477001269.0,
      "step": 13712
    },
    {
      "epoch": 2.5465181058495823,
      "grad_norm": 1.7576788160868186,
      "learning_rate": 1e-06,
      "loss": 0.4061,
      "mean_token_accuracy": 0.8701505064964294,
      "num_tokens": 477037997.0,
      "step": 13713
    },
    {
      "epoch": 2.546703806870938,
      "grad_norm": 1.510305162533426,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8813822269439697,
      "num_tokens": 477073582.0,
      "step": 13714
    },
    {
      "epoch": 2.5468895078922933,
      "grad_norm": 1.5303804735468662,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8812367916107178,
      "num_tokens": 477109088.0,
      "step": 13715
    },
    {
      "epoch": 2.547075208913649,
      "grad_norm": 1.5461053255129185,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8893258571624756,
      "num_tokens": 477143195.0,
      "step": 13716
    },
    {
      "epoch": 2.5472609099350048,
      "grad_norm": 1.594205466731452,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.8891187906265259,
      "num_tokens": 477176293.0,
      "step": 13717
    },
    {
      "epoch": 2.54744661095636,
      "grad_norm": 1.6709993183647998,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.878383219242096,
      "num_tokens": 477206446.0,
      "step": 13718
    },
    {
      "epoch": 2.547632311977716,
      "grad_norm": 1.5307211511882988,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8905572891235352,
      "num_tokens": 477240070.0,
      "step": 13719
    },
    {
      "epoch": 2.5478180129990715,
      "grad_norm": 1.490370612626719,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.886284351348877,
      "num_tokens": 477277817.0,
      "step": 13720
    },
    {
      "epoch": 2.5480037140204272,
      "grad_norm": 1.5097035068663751,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8837730288505554,
      "num_tokens": 477318531.0,
      "step": 13721
    },
    {
      "epoch": 2.548189415041783,
      "grad_norm": 1.513139266518382,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8925655484199524,
      "num_tokens": 477355688.0,
      "step": 13722
    },
    {
      "epoch": 2.5483751160631383,
      "grad_norm": 1.623947777432967,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8966736197471619,
      "num_tokens": 477390789.0,
      "step": 13723
    },
    {
      "epoch": 2.548560817084494,
      "grad_norm": 1.6674872858897853,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8789982795715332,
      "num_tokens": 477423087.0,
      "step": 13724
    },
    {
      "epoch": 2.5487465181058497,
      "grad_norm": 1.4618371022979575,
      "learning_rate": 1e-06,
      "loss": 0.2907,
      "mean_token_accuracy": 0.8982940912246704,
      "num_tokens": 477459019.0,
      "step": 13725
    },
    {
      "epoch": 2.548932219127205,
      "grad_norm": 1.6235812958456366,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8752074241638184,
      "num_tokens": 477494122.0,
      "step": 13726
    },
    {
      "epoch": 2.5491179201485608,
      "grad_norm": 1.5922771247651235,
      "learning_rate": 1e-06,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8906301259994507,
      "num_tokens": 477524994.0,
      "step": 13727
    },
    {
      "epoch": 2.5493036211699165,
      "grad_norm": 1.4787517202899734,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.896034836769104,
      "num_tokens": 477558094.0,
      "step": 13728
    },
    {
      "epoch": 2.5494893221912722,
      "grad_norm": 1.6088080900045882,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8921701908111572,
      "num_tokens": 477591799.0,
      "step": 13729
    },
    {
      "epoch": 2.5496750232126275,
      "grad_norm": 1.5682869403860733,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8862127065658569,
      "num_tokens": 477628635.0,
      "step": 13730
    },
    {
      "epoch": 2.5498607242339832,
      "grad_norm": 1.5428022908681647,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8847178220748901,
      "num_tokens": 477666887.0,
      "step": 13731
    },
    {
      "epoch": 2.550046425255339,
      "grad_norm": 1.5188116915822105,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8799800872802734,
      "num_tokens": 477705781.0,
      "step": 13732
    },
    {
      "epoch": 2.5502321262766943,
      "grad_norm": 1.575688271586033,
      "learning_rate": 1e-06,
      "loss": 0.2852,
      "mean_token_accuracy": 0.9029827117919922,
      "num_tokens": 477736789.0,
      "step": 13733
    },
    {
      "epoch": 2.55041782729805,
      "grad_norm": 1.6663311189474532,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8858405351638794,
      "num_tokens": 477765172.0,
      "step": 13734
    },
    {
      "epoch": 2.5506035283194057,
      "grad_norm": 1.5477228920746882,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8969288468360901,
      "num_tokens": 477799938.0,
      "step": 13735
    },
    {
      "epoch": 2.5507892293407615,
      "grad_norm": 1.4941046180352497,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8967238664627075,
      "num_tokens": 477837158.0,
      "step": 13736
    },
    {
      "epoch": 2.550974930362117,
      "grad_norm": 1.5840319204978481,
      "learning_rate": 1e-06,
      "loss": 0.2843,
      "mean_token_accuracy": 0.8988244533538818,
      "num_tokens": 477865984.0,
      "step": 13737
    },
    {
      "epoch": 2.5511606313834725,
      "grad_norm": 1.632730261073255,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.879664957523346,
      "num_tokens": 477898002.0,
      "step": 13738
    },
    {
      "epoch": 2.551346332404828,
      "grad_norm": 1.6271083392722936,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.886161208152771,
      "num_tokens": 477935826.0,
      "step": 13739
    },
    {
      "epoch": 2.551532033426184,
      "grad_norm": 1.609389144804377,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8909685611724854,
      "num_tokens": 477968274.0,
      "step": 13740
    },
    {
      "epoch": 2.5517177344475392,
      "grad_norm": 1.5481939531708422,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.882420539855957,
      "num_tokens": 478004805.0,
      "step": 13741
    },
    {
      "epoch": 2.551903435468895,
      "grad_norm": 1.7039069099889155,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8874099254608154,
      "num_tokens": 478035157.0,
      "step": 13742
    },
    {
      "epoch": 2.5520891364902507,
      "grad_norm": 1.5845806851913413,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8852196931838989,
      "num_tokens": 478068959.0,
      "step": 13743
    },
    {
      "epoch": 2.5522748375116064,
      "grad_norm": 1.5138372972864333,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8904075622558594,
      "num_tokens": 478103497.0,
      "step": 13744
    },
    {
      "epoch": 2.552460538532962,
      "grad_norm": 1.6008621375878727,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.885952889919281,
      "num_tokens": 478139404.0,
      "step": 13745
    },
    {
      "epoch": 2.5526462395543175,
      "grad_norm": 1.6459414576950044,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8834611177444458,
      "num_tokens": 478169443.0,
      "step": 13746
    },
    {
      "epoch": 2.552831940575673,
      "grad_norm": 1.5464055746983068,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8901320099830627,
      "num_tokens": 478204464.0,
      "step": 13747
    },
    {
      "epoch": 2.553017641597029,
      "grad_norm": 1.5729583646561007,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8812829256057739,
      "num_tokens": 478238613.0,
      "step": 13748
    },
    {
      "epoch": 2.553203342618384,
      "grad_norm": 1.46554260087707,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8952317237854004,
      "num_tokens": 478272610.0,
      "step": 13749
    },
    {
      "epoch": 2.55338904363974,
      "grad_norm": 1.5983371115360272,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.880402147769928,
      "num_tokens": 478304225.0,
      "step": 13750
    },
    {
      "epoch": 2.5535747446610957,
      "grad_norm": 1.587332332163851,
      "learning_rate": 1e-06,
      "loss": 0.3073,
      "mean_token_accuracy": 0.8933228254318237,
      "num_tokens": 478336349.0,
      "step": 13751
    },
    {
      "epoch": 2.5537604456824514,
      "grad_norm": 1.360664694794473,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8940623998641968,
      "num_tokens": 478380533.0,
      "step": 13752
    },
    {
      "epoch": 2.5539461467038067,
      "grad_norm": 1.5725496149351144,
      "learning_rate": 1e-06,
      "loss": 0.2853,
      "mean_token_accuracy": 0.8983920216560364,
      "num_tokens": 478415680.0,
      "step": 13753
    },
    {
      "epoch": 2.5541318477251624,
      "grad_norm": 1.5001525285575061,
      "learning_rate": 1e-06,
      "loss": 0.269,
      "mean_token_accuracy": 0.9045220613479614,
      "num_tokens": 478450544.0,
      "step": 13754
    },
    {
      "epoch": 2.554317548746518,
      "grad_norm": 1.6730084873034152,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8834956884384155,
      "num_tokens": 478484482.0,
      "step": 13755
    },
    {
      "epoch": 2.5545032497678735,
      "grad_norm": 1.3866936612406904,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8903762102127075,
      "num_tokens": 478528842.0,
      "step": 13756
    },
    {
      "epoch": 2.554688950789229,
      "grad_norm": 1.6897753997177278,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8825232982635498,
      "num_tokens": 478559859.0,
      "step": 13757
    },
    {
      "epoch": 2.554874651810585,
      "grad_norm": 1.693898520931453,
      "learning_rate": 1e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8655860424041748,
      "num_tokens": 478597296.0,
      "step": 13758
    },
    {
      "epoch": 2.5550603528319407,
      "grad_norm": 1.491689693883866,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8845958113670349,
      "num_tokens": 478634344.0,
      "step": 13759
    },
    {
      "epoch": 2.5552460538532964,
      "grad_norm": 1.5640136609056137,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8912302851676941,
      "num_tokens": 478667670.0,
      "step": 13760
    },
    {
      "epoch": 2.5554317548746517,
      "grad_norm": 1.5383922466511855,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8809190988540649,
      "num_tokens": 478707393.0,
      "step": 13761
    },
    {
      "epoch": 2.5556174558960074,
      "grad_norm": 1.5670960699710026,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8765806555747986,
      "num_tokens": 478744040.0,
      "step": 13762
    },
    {
      "epoch": 2.555803156917363,
      "grad_norm": 1.6096316483346065,
      "learning_rate": 1e-06,
      "loss": 0.2954,
      "mean_token_accuracy": 0.8955127000808716,
      "num_tokens": 478775185.0,
      "step": 13763
    },
    {
      "epoch": 2.5559888579387184,
      "grad_norm": 1.550475098061735,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8873841762542725,
      "num_tokens": 478808411.0,
      "step": 13764
    },
    {
      "epoch": 2.556174558960074,
      "grad_norm": 1.5506472092541805,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8913126587867737,
      "num_tokens": 478845286.0,
      "step": 13765
    },
    {
      "epoch": 2.55636025998143,
      "grad_norm": 1.6012549980402047,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.8941502571105957,
      "num_tokens": 478876357.0,
      "step": 13766
    },
    {
      "epoch": 2.5565459610027856,
      "grad_norm": 1.8254312483584125,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8885776996612549,
      "num_tokens": 478901383.0,
      "step": 13767
    },
    {
      "epoch": 2.5567316620241414,
      "grad_norm": 1.64211838319452,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.880979597568512,
      "num_tokens": 478936523.0,
      "step": 13768
    },
    {
      "epoch": 2.5569173630454967,
      "grad_norm": 1.5354458781277367,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8849081993103027,
      "num_tokens": 478972427.0,
      "step": 13769
    },
    {
      "epoch": 2.5571030640668524,
      "grad_norm": 1.5141647547660544,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8896633386611938,
      "num_tokens": 479006257.0,
      "step": 13770
    },
    {
      "epoch": 2.557288765088208,
      "grad_norm": 1.4230545968746997,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8854490518569946,
      "num_tokens": 479045907.0,
      "step": 13771
    },
    {
      "epoch": 2.5574744661095634,
      "grad_norm": 1.6364033044080208,
      "learning_rate": 1e-06,
      "loss": 0.3654,
      "mean_token_accuracy": 0.8748346567153931,
      "num_tokens": 479080247.0,
      "step": 13772
    },
    {
      "epoch": 2.557660167130919,
      "grad_norm": 1.5862052746291446,
      "learning_rate": 1e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8745711445808411,
      "num_tokens": 479114995.0,
      "step": 13773
    },
    {
      "epoch": 2.557845868152275,
      "grad_norm": 1.8656137703326559,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8854356408119202,
      "num_tokens": 479140789.0,
      "step": 13774
    },
    {
      "epoch": 2.5580315691736306,
      "grad_norm": 1.6090478272627784,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8868216872215271,
      "num_tokens": 479173396.0,
      "step": 13775
    },
    {
      "epoch": 2.5582172701949863,
      "grad_norm": 1.5814609600737064,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8870556354522705,
      "num_tokens": 479211546.0,
      "step": 13776
    },
    {
      "epoch": 2.5584029712163416,
      "grad_norm": 1.5532093802427671,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.890485942363739,
      "num_tokens": 479249457.0,
      "step": 13777
    },
    {
      "epoch": 2.5585886722376974,
      "grad_norm": 1.5584893611430317,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8814334273338318,
      "num_tokens": 479285537.0,
      "step": 13778
    },
    {
      "epoch": 2.5587743732590527,
      "grad_norm": 1.5260509240739064,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8902896046638489,
      "num_tokens": 479321747.0,
      "step": 13779
    },
    {
      "epoch": 2.5589600742804084,
      "grad_norm": 1.5045177053553205,
      "learning_rate": 1e-06,
      "loss": 0.2714,
      "mean_token_accuracy": 0.902106761932373,
      "num_tokens": 479355715.0,
      "step": 13780
    },
    {
      "epoch": 2.559145775301764,
      "grad_norm": 1.6376514402330908,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8798338174819946,
      "num_tokens": 479387802.0,
      "step": 13781
    },
    {
      "epoch": 2.55933147632312,
      "grad_norm": 1.6385804627161715,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8783867955207825,
      "num_tokens": 479422809.0,
      "step": 13782
    },
    {
      "epoch": 2.5595171773444756,
      "grad_norm": 1.4663998743939677,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.8930096626281738,
      "num_tokens": 479459400.0,
      "step": 13783
    },
    {
      "epoch": 2.559702878365831,
      "grad_norm": 1.4624243555456262,
      "learning_rate": 1e-06,
      "loss": 0.2916,
      "mean_token_accuracy": 0.8947827816009521,
      "num_tokens": 479497593.0,
      "step": 13784
    },
    {
      "epoch": 2.5598885793871866,
      "grad_norm": 1.6283824423585778,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.879324197769165,
      "num_tokens": 479531563.0,
      "step": 13785
    },
    {
      "epoch": 2.5600742804085423,
      "grad_norm": 1.5403233854490257,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8951429128646851,
      "num_tokens": 479566876.0,
      "step": 13786
    },
    {
      "epoch": 2.5602599814298976,
      "grad_norm": 1.444627515445658,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8824359178543091,
      "num_tokens": 479607473.0,
      "step": 13787
    },
    {
      "epoch": 2.5604456824512534,
      "grad_norm": 1.3776739933204272,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8854009509086609,
      "num_tokens": 479651455.0,
      "step": 13788
    },
    {
      "epoch": 2.560631383472609,
      "grad_norm": 1.4259506194606428,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8822028636932373,
      "num_tokens": 479694320.0,
      "step": 13789
    },
    {
      "epoch": 2.560817084493965,
      "grad_norm": 1.4381464462081908,
      "learning_rate": 1e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8970721960067749,
      "num_tokens": 479736237.0,
      "step": 13790
    },
    {
      "epoch": 2.5610027855153206,
      "grad_norm": 1.59779487756226,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8919159173965454,
      "num_tokens": 479771752.0,
      "step": 13791
    },
    {
      "epoch": 2.561188486536676,
      "grad_norm": 1.7009929144037244,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8701579570770264,
      "num_tokens": 479803149.0,
      "step": 13792
    },
    {
      "epoch": 2.5613741875580316,
      "grad_norm": 1.5648485337969085,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8890343904495239,
      "num_tokens": 479838627.0,
      "step": 13793
    },
    {
      "epoch": 2.5615598885793873,
      "grad_norm": 1.5391075247713526,
      "learning_rate": 1e-06,
      "loss": 0.3114,
      "mean_token_accuracy": 0.8899338841438293,
      "num_tokens": 479871478.0,
      "step": 13794
    },
    {
      "epoch": 2.5617455896007426,
      "grad_norm": 1.748910142218584,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8877745866775513,
      "num_tokens": 479898791.0,
      "step": 13795
    },
    {
      "epoch": 2.5619312906220983,
      "grad_norm": 1.7585542284917224,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8870922327041626,
      "num_tokens": 479929572.0,
      "step": 13796
    },
    {
      "epoch": 2.562116991643454,
      "grad_norm": 1.4863663151947701,
      "learning_rate": 1e-06,
      "loss": 0.2845,
      "mean_token_accuracy": 0.9009944200515747,
      "num_tokens": 479964554.0,
      "step": 13797
    },
    {
      "epoch": 2.56230269266481,
      "grad_norm": 1.627223925519662,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8836429715156555,
      "num_tokens": 479995116.0,
      "step": 13798
    },
    {
      "epoch": 2.5624883936861655,
      "grad_norm": 1.608372073362766,
      "learning_rate": 1e-06,
      "loss": 0.2891,
      "mean_token_accuracy": 0.8970212340354919,
      "num_tokens": 480030530.0,
      "step": 13799
    },
    {
      "epoch": 2.562674094707521,
      "grad_norm": 1.5501435652161388,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8934069275856018,
      "num_tokens": 480063631.0,
      "step": 13800
    },
    {
      "epoch": 2.5628597957288766,
      "grad_norm": 1.5316435643848825,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.885257363319397,
      "num_tokens": 480102432.0,
      "step": 13801
    },
    {
      "epoch": 2.563045496750232,
      "grad_norm": 1.7540548060799401,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8826289176940918,
      "num_tokens": 480132795.0,
      "step": 13802
    },
    {
      "epoch": 2.5632311977715876,
      "grad_norm": 1.4764566127119159,
      "learning_rate": 1e-06,
      "loss": 0.2865,
      "mean_token_accuracy": 0.894246518611908,
      "num_tokens": 480169322.0,
      "step": 13803
    },
    {
      "epoch": 2.5634168987929433,
      "grad_norm": 1.797495971997413,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8792975544929504,
      "num_tokens": 480199200.0,
      "step": 13804
    },
    {
      "epoch": 2.563602599814299,
      "grad_norm": 1.5882639676497041,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8931503891944885,
      "num_tokens": 480230308.0,
      "step": 13805
    },
    {
      "epoch": 2.563788300835655,
      "grad_norm": 1.6954925995665269,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8913456201553345,
      "num_tokens": 480260796.0,
      "step": 13806
    },
    {
      "epoch": 2.56397400185701,
      "grad_norm": 1.4349084750664924,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.8984935283660889,
      "num_tokens": 480299165.0,
      "step": 13807
    },
    {
      "epoch": 2.564159702878366,
      "grad_norm": 1.5099229783466022,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8910609483718872,
      "num_tokens": 480332781.0,
      "step": 13808
    },
    {
      "epoch": 2.5643454038997215,
      "grad_norm": 1.5854513721532753,
      "learning_rate": 1e-06,
      "loss": 0.2984,
      "mean_token_accuracy": 0.8963651657104492,
      "num_tokens": 480366789.0,
      "step": 13809
    },
    {
      "epoch": 2.564531104921077,
      "grad_norm": 1.6506256645100033,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8812335133552551,
      "num_tokens": 480401004.0,
      "step": 13810
    },
    {
      "epoch": 2.5647168059424326,
      "grad_norm": 1.5403722463030927,
      "learning_rate": 1e-06,
      "loss": 0.3266,
      "mean_token_accuracy": 0.8838964700698853,
      "num_tokens": 480441529.0,
      "step": 13811
    },
    {
      "epoch": 2.5649025069637883,
      "grad_norm": 1.607054936915162,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8851608037948608,
      "num_tokens": 480474705.0,
      "step": 13812
    },
    {
      "epoch": 2.565088207985144,
      "grad_norm": 1.5060127679131872,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.878777265548706,
      "num_tokens": 480514373.0,
      "step": 13813
    },
    {
      "epoch": 2.5652739090064998,
      "grad_norm": 1.6204146559666894,
      "learning_rate": 1e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.8747464418411255,
      "num_tokens": 480551288.0,
      "step": 13814
    },
    {
      "epoch": 2.565459610027855,
      "grad_norm": 1.706338886406105,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8807148933410645,
      "num_tokens": 480583057.0,
      "step": 13815
    },
    {
      "epoch": 2.565645311049211,
      "grad_norm": 1.72654860814481,
      "learning_rate": 1e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8780717253684998,
      "num_tokens": 480615325.0,
      "step": 13816
    },
    {
      "epoch": 2.5658310120705665,
      "grad_norm": 1.5367872215176117,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.887992262840271,
      "num_tokens": 480649329.0,
      "step": 13817
    },
    {
      "epoch": 2.566016713091922,
      "grad_norm": 1.510739808818621,
      "learning_rate": 1e-06,
      "loss": 0.3026,
      "mean_token_accuracy": 0.8955053091049194,
      "num_tokens": 480686296.0,
      "step": 13818
    },
    {
      "epoch": 2.5662024141132775,
      "grad_norm": 1.6373235751657007,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.8973759412765503,
      "num_tokens": 480714397.0,
      "step": 13819
    },
    {
      "epoch": 2.5663881151346333,
      "grad_norm": 1.5301918624776811,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.889938235282898,
      "num_tokens": 480748440.0,
      "step": 13820
    },
    {
      "epoch": 2.566573816155989,
      "grad_norm": 1.3993713565685313,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8807880282402039,
      "num_tokens": 480790097.0,
      "step": 13821
    },
    {
      "epoch": 2.5667595171773447,
      "grad_norm": 1.610588577278695,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8782100677490234,
      "num_tokens": 480823862.0,
      "step": 13822
    },
    {
      "epoch": 2.5669452181987,
      "grad_norm": 1.5396599590239701,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8864970803260803,
      "num_tokens": 480860520.0,
      "step": 13823
    },
    {
      "epoch": 2.5671309192200558,
      "grad_norm": 1.7949556278732746,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8859107494354248,
      "num_tokens": 480892282.0,
      "step": 13824
    },
    {
      "epoch": 2.567316620241411,
      "grad_norm": 1.6691561922475067,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8874654769897461,
      "num_tokens": 480922809.0,
      "step": 13825
    },
    {
      "epoch": 2.567502321262767,
      "grad_norm": 1.6699396133961708,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8909381628036499,
      "num_tokens": 480951540.0,
      "step": 13826
    },
    {
      "epoch": 2.5676880222841225,
      "grad_norm": 1.6918643049345037,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8800137042999268,
      "num_tokens": 480984776.0,
      "step": 13827
    },
    {
      "epoch": 2.5678737233054783,
      "grad_norm": 1.612115181875463,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8809705376625061,
      "num_tokens": 481018506.0,
      "step": 13828
    },
    {
      "epoch": 2.568059424326834,
      "grad_norm": 1.7247613652067129,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8870181441307068,
      "num_tokens": 481048170.0,
      "step": 13829
    },
    {
      "epoch": 2.5682451253481893,
      "grad_norm": 1.5713625198018308,
      "learning_rate": 1e-06,
      "loss": 0.3404,
      "mean_token_accuracy": 0.8818066120147705,
      "num_tokens": 481083519.0,
      "step": 13830
    },
    {
      "epoch": 2.568430826369545,
      "grad_norm": 1.7273428422448895,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8899452090263367,
      "num_tokens": 481114278.0,
      "step": 13831
    },
    {
      "epoch": 2.5686165273909007,
      "grad_norm": 1.6938502455338822,
      "learning_rate": 1e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8708312511444092,
      "num_tokens": 481148062.0,
      "step": 13832
    },
    {
      "epoch": 2.568802228412256,
      "grad_norm": 1.6353413264388845,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8849921822547913,
      "num_tokens": 481178721.0,
      "step": 13833
    },
    {
      "epoch": 2.5689879294336118,
      "grad_norm": 1.6745569579026935,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8760037422180176,
      "num_tokens": 481211173.0,
      "step": 13834
    },
    {
      "epoch": 2.5691736304549675,
      "grad_norm": 1.6267130731779793,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.887273907661438,
      "num_tokens": 481240798.0,
      "step": 13835
    },
    {
      "epoch": 2.5693593314763232,
      "grad_norm": 1.5256650737631705,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8875547647476196,
      "num_tokens": 481277278.0,
      "step": 13836
    },
    {
      "epoch": 2.569545032497679,
      "grad_norm": 1.7505858675836952,
      "learning_rate": 1e-06,
      "loss": 0.3545,
      "mean_token_accuracy": 0.8791449666023254,
      "num_tokens": 481309660.0,
      "step": 13837
    },
    {
      "epoch": 2.5697307335190342,
      "grad_norm": 1.6182774391206016,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8783621788024902,
      "num_tokens": 481341998.0,
      "step": 13838
    },
    {
      "epoch": 2.56991643454039,
      "grad_norm": 1.545223476268457,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8862590789794922,
      "num_tokens": 481375777.0,
      "step": 13839
    },
    {
      "epoch": 2.5701021355617457,
      "grad_norm": 1.7431383514318939,
      "learning_rate": 1e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8769574165344238,
      "num_tokens": 481404991.0,
      "step": 13840
    },
    {
      "epoch": 2.570287836583101,
      "grad_norm": 1.6103587384562221,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8924821019172668,
      "num_tokens": 481441861.0,
      "step": 13841
    },
    {
      "epoch": 2.5704735376044567,
      "grad_norm": 1.4108504922421063,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.88481605052948,
      "num_tokens": 481483046.0,
      "step": 13842
    },
    {
      "epoch": 2.5706592386258125,
      "grad_norm": 1.6264434566996195,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8898118734359741,
      "num_tokens": 481517516.0,
      "step": 13843
    },
    {
      "epoch": 2.570844939647168,
      "grad_norm": 1.4081425455724474,
      "learning_rate": 1e-06,
      "loss": 0.2925,
      "mean_token_accuracy": 0.8957343101501465,
      "num_tokens": 481555960.0,
      "step": 13844
    },
    {
      "epoch": 2.571030640668524,
      "grad_norm": 1.6883155078131842,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8853839039802551,
      "num_tokens": 481586855.0,
      "step": 13845
    },
    {
      "epoch": 2.5712163416898792,
      "grad_norm": 1.7885574767154722,
      "learning_rate": 1e-06,
      "loss": 0.3387,
      "mean_token_accuracy": 0.8792179822921753,
      "num_tokens": 481616979.0,
      "step": 13846
    },
    {
      "epoch": 2.571402042711235,
      "grad_norm": 1.6394610035164807,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8861449360847473,
      "num_tokens": 481647998.0,
      "step": 13847
    },
    {
      "epoch": 2.5715877437325907,
      "grad_norm": 1.6361220466746607,
      "learning_rate": 1e-06,
      "loss": 0.3652,
      "mean_token_accuracy": 0.8727717399597168,
      "num_tokens": 481684645.0,
      "step": 13848
    },
    {
      "epoch": 2.571773444753946,
      "grad_norm": 1.6264944723489998,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8946576118469238,
      "num_tokens": 481719289.0,
      "step": 13849
    },
    {
      "epoch": 2.5719591457753017,
      "grad_norm": 1.5374440181212174,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8751457929611206,
      "num_tokens": 481757884.0,
      "step": 13850
    },
    {
      "epoch": 2.5721448467966574,
      "grad_norm": 1.7691196244318,
      "learning_rate": 1e-06,
      "loss": 0.3713,
      "mean_token_accuracy": 0.8697007894515991,
      "num_tokens": 481788606.0,
      "step": 13851
    },
    {
      "epoch": 2.572330547818013,
      "grad_norm": 1.8722525333506705,
      "learning_rate": 1e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8771872520446777,
      "num_tokens": 481814970.0,
      "step": 13852
    },
    {
      "epoch": 2.5725162488393685,
      "grad_norm": 1.6920653688264566,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8702879548072815,
      "num_tokens": 481849647.0,
      "step": 13853
    },
    {
      "epoch": 2.572701949860724,
      "grad_norm": 1.5772971430143121,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8892999291419983,
      "num_tokens": 481882535.0,
      "step": 13854
    },
    {
      "epoch": 2.57288765088208,
      "grad_norm": 1.5604078416084541,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8868911266326904,
      "num_tokens": 481918153.0,
      "step": 13855
    },
    {
      "epoch": 2.5730733519034352,
      "grad_norm": 1.5565928930510746,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8855792284011841,
      "num_tokens": 481955758.0,
      "step": 13856
    },
    {
      "epoch": 2.573259052924791,
      "grad_norm": 1.5316097158643085,
      "learning_rate": 1e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8725546598434448,
      "num_tokens": 481995414.0,
      "step": 13857
    },
    {
      "epoch": 2.5734447539461467,
      "grad_norm": 1.6099630309904716,
      "learning_rate": 1e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8635247945785522,
      "num_tokens": 482032656.0,
      "step": 13858
    },
    {
      "epoch": 2.5736304549675024,
      "grad_norm": 1.4818412388611484,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8948434591293335,
      "num_tokens": 482070516.0,
      "step": 13859
    },
    {
      "epoch": 2.573816155988858,
      "grad_norm": 1.7210219024967746,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.889130175113678,
      "num_tokens": 482097840.0,
      "step": 13860
    },
    {
      "epoch": 2.5740018570102134,
      "grad_norm": 1.5665044355300233,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8832278251647949,
      "num_tokens": 482132536.0,
      "step": 13861
    },
    {
      "epoch": 2.574187558031569,
      "grad_norm": 1.6983605847948122,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8837293386459351,
      "num_tokens": 482163281.0,
      "step": 13862
    },
    {
      "epoch": 2.574373259052925,
      "grad_norm": 1.4362011991198105,
      "learning_rate": 1e-06,
      "loss": 0.2903,
      "mean_token_accuracy": 0.8981372117996216,
      "num_tokens": 482197849.0,
      "step": 13863
    },
    {
      "epoch": 2.57455896007428,
      "grad_norm": 1.5616515971851825,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8907260298728943,
      "num_tokens": 482231378.0,
      "step": 13864
    },
    {
      "epoch": 2.574744661095636,
      "grad_norm": 1.5500503554556637,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8897496461868286,
      "num_tokens": 482265412.0,
      "step": 13865
    },
    {
      "epoch": 2.5749303621169917,
      "grad_norm": 1.5210346643763644,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.883478045463562,
      "num_tokens": 482305308.0,
      "step": 13866
    },
    {
      "epoch": 2.5751160631383474,
      "grad_norm": 1.5695100528257009,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8735569715499878,
      "num_tokens": 482342480.0,
      "step": 13867
    },
    {
      "epoch": 2.575301764159703,
      "grad_norm": 1.4485678484956757,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8947945833206177,
      "num_tokens": 482378330.0,
      "step": 13868
    },
    {
      "epoch": 2.5754874651810584,
      "grad_norm": 1.5683354793258182,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8802437782287598,
      "num_tokens": 482414949.0,
      "step": 13869
    },
    {
      "epoch": 2.575673166202414,
      "grad_norm": 1.3626299114238767,
      "learning_rate": 1e-06,
      "loss": 0.2816,
      "mean_token_accuracy": 0.8977572917938232,
      "num_tokens": 482457280.0,
      "step": 13870
    },
    {
      "epoch": 2.57585886722377,
      "grad_norm": 1.6258816920663832,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8866342306137085,
      "num_tokens": 482490253.0,
      "step": 13871
    },
    {
      "epoch": 2.576044568245125,
      "grad_norm": 1.5901108642500916,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8838546872138977,
      "num_tokens": 482523152.0,
      "step": 13872
    },
    {
      "epoch": 2.576230269266481,
      "grad_norm": 1.6258004812480715,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8934834003448486,
      "num_tokens": 482552941.0,
      "step": 13873
    },
    {
      "epoch": 2.5764159702878366,
      "grad_norm": 1.553854696979302,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8744395971298218,
      "num_tokens": 482591861.0,
      "step": 13874
    },
    {
      "epoch": 2.5766016713091924,
      "grad_norm": 1.5631311159845749,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8873955607414246,
      "num_tokens": 482625424.0,
      "step": 13875
    },
    {
      "epoch": 2.5767873723305477,
      "grad_norm": 1.6337631981755334,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8883883357048035,
      "num_tokens": 482655018.0,
      "step": 13876
    },
    {
      "epoch": 2.5769730733519034,
      "grad_norm": 1.584358158383404,
      "learning_rate": 1e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.870895266532898,
      "num_tokens": 482687980.0,
      "step": 13877
    },
    {
      "epoch": 2.577158774373259,
      "grad_norm": 1.5367672207537861,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8773347735404968,
      "num_tokens": 482727572.0,
      "step": 13878
    },
    {
      "epoch": 2.5773444753946144,
      "grad_norm": 1.5760717560369852,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.891215443611145,
      "num_tokens": 482762631.0,
      "step": 13879
    },
    {
      "epoch": 2.57753017641597,
      "grad_norm": 1.5095981550247881,
      "learning_rate": 1e-06,
      "loss": 0.3099,
      "mean_token_accuracy": 0.8898544311523438,
      "num_tokens": 482798266.0,
      "step": 13880
    },
    {
      "epoch": 2.577715877437326,
      "grad_norm": 1.6163748538789962,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8812474012374878,
      "num_tokens": 482829592.0,
      "step": 13881
    },
    {
      "epoch": 2.5779015784586816,
      "grad_norm": 1.5203484849840065,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8898523449897766,
      "num_tokens": 482863815.0,
      "step": 13882
    },
    {
      "epoch": 2.5780872794800374,
      "grad_norm": 1.6298554273608854,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8949196338653564,
      "num_tokens": 482893840.0,
      "step": 13883
    },
    {
      "epoch": 2.5782729805013926,
      "grad_norm": 1.5831605794150374,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.879911482334137,
      "num_tokens": 482931830.0,
      "step": 13884
    },
    {
      "epoch": 2.5784586815227484,
      "grad_norm": 1.6235488448813404,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8960919380187988,
      "num_tokens": 482962383.0,
      "step": 13885
    },
    {
      "epoch": 2.578644382544104,
      "grad_norm": 1.553171329044824,
      "learning_rate": 1e-06,
      "loss": 0.2899,
      "mean_token_accuracy": 0.8952677845954895,
      "num_tokens": 482997312.0,
      "step": 13886
    },
    {
      "epoch": 2.5788300835654594,
      "grad_norm": 1.5329621554392225,
      "learning_rate": 1e-06,
      "loss": 0.3286,
      "mean_token_accuracy": 0.8850185871124268,
      "num_tokens": 483035947.0,
      "step": 13887
    },
    {
      "epoch": 2.579015784586815,
      "grad_norm": 1.7031755374503974,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.895239531993866,
      "num_tokens": 483065186.0,
      "step": 13888
    },
    {
      "epoch": 2.579201485608171,
      "grad_norm": 1.526628277951643,
      "learning_rate": 1e-06,
      "loss": 0.2745,
      "mean_token_accuracy": 0.9008137583732605,
      "num_tokens": 483096366.0,
      "step": 13889
    },
    {
      "epoch": 2.5793871866295266,
      "grad_norm": 1.6553362406868917,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8880560398101807,
      "num_tokens": 483130211.0,
      "step": 13890
    },
    {
      "epoch": 2.5795728876508823,
      "grad_norm": 1.5493165899265018,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8884665369987488,
      "num_tokens": 483165428.0,
      "step": 13891
    },
    {
      "epoch": 2.5797585886722376,
      "grad_norm": 1.5541384695094846,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8912969827651978,
      "num_tokens": 483199322.0,
      "step": 13892
    },
    {
      "epoch": 2.5799442896935934,
      "grad_norm": 1.6490722263665076,
      "learning_rate": 1e-06,
      "loss": 0.262,
      "mean_token_accuracy": 0.9067211151123047,
      "num_tokens": 483229328.0,
      "step": 13893
    },
    {
      "epoch": 2.580129990714949,
      "grad_norm": 1.55080790857419,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8832122683525085,
      "num_tokens": 483267791.0,
      "step": 13894
    },
    {
      "epoch": 2.5803156917363044,
      "grad_norm": 1.5834417115460655,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8861768245697021,
      "num_tokens": 483301704.0,
      "step": 13895
    },
    {
      "epoch": 2.58050139275766,
      "grad_norm": 1.4390546800646633,
      "learning_rate": 1e-06,
      "loss": 0.2911,
      "mean_token_accuracy": 0.8961314558982849,
      "num_tokens": 483343568.0,
      "step": 13896
    },
    {
      "epoch": 2.580687093779016,
      "grad_norm": 1.6461028023236555,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8873969316482544,
      "num_tokens": 483380404.0,
      "step": 13897
    },
    {
      "epoch": 2.5808727948003716,
      "grad_norm": 1.5661004276107007,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8898969888687134,
      "num_tokens": 483415895.0,
      "step": 13898
    },
    {
      "epoch": 2.581058495821727,
      "grad_norm": 1.5271085880746753,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8821367025375366,
      "num_tokens": 483457605.0,
      "step": 13899
    },
    {
      "epoch": 2.5812441968430826,
      "grad_norm": 1.6547590723738335,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8860222101211548,
      "num_tokens": 483489711.0,
      "step": 13900
    },
    {
      "epoch": 2.5814298978644383,
      "grad_norm": 1.4295128776350325,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.892774224281311,
      "num_tokens": 483526919.0,
      "step": 13901
    },
    {
      "epoch": 2.5816155988857936,
      "grad_norm": 1.638555813102416,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8879318237304688,
      "num_tokens": 483565706.0,
      "step": 13902
    },
    {
      "epoch": 2.5818012999071493,
      "grad_norm": 1.5387616390839707,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8887691497802734,
      "num_tokens": 483603693.0,
      "step": 13903
    },
    {
      "epoch": 2.581987000928505,
      "grad_norm": 1.7700419549625033,
      "learning_rate": 1e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8569035530090332,
      "num_tokens": 483634826.0,
      "step": 13904
    },
    {
      "epoch": 2.582172701949861,
      "grad_norm": 1.768773635304029,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8890898823738098,
      "num_tokens": 483669080.0,
      "step": 13905
    },
    {
      "epoch": 2.5823584029712165,
      "grad_norm": 1.9526593748702474,
      "learning_rate": 1e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8716760277748108,
      "num_tokens": 483697012.0,
      "step": 13906
    },
    {
      "epoch": 2.582544103992572,
      "grad_norm": 1.628993018920668,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8950470685958862,
      "num_tokens": 483728951.0,
      "step": 13907
    },
    {
      "epoch": 2.5827298050139276,
      "grad_norm": 1.504918951621052,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8883541822433472,
      "num_tokens": 483763964.0,
      "step": 13908
    },
    {
      "epoch": 2.5829155060352833,
      "grad_norm": 1.7276959874694753,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8812981843948364,
      "num_tokens": 483796770.0,
      "step": 13909
    },
    {
      "epoch": 2.5831012070566386,
      "grad_norm": 1.635911387252689,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8920738697052002,
      "num_tokens": 483826817.0,
      "step": 13910
    },
    {
      "epoch": 2.5832869080779943,
      "grad_norm": 1.5729156574817533,
      "learning_rate": 1e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.876663863658905,
      "num_tokens": 483861623.0,
      "step": 13911
    },
    {
      "epoch": 2.58347260909935,
      "grad_norm": 1.738897176102773,
      "learning_rate": 1e-06,
      "loss": 0.305,
      "mean_token_accuracy": 0.8931145071983337,
      "num_tokens": 483890406.0,
      "step": 13912
    },
    {
      "epoch": 2.583658310120706,
      "grad_norm": 1.6271548310704738,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8801275491714478,
      "num_tokens": 483920728.0,
      "step": 13913
    },
    {
      "epoch": 2.5838440111420615,
      "grad_norm": 1.6206665925416983,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8730835914611816,
      "num_tokens": 483953636.0,
      "step": 13914
    },
    {
      "epoch": 2.584029712163417,
      "grad_norm": 1.776169335447879,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8926111459732056,
      "num_tokens": 483977686.0,
      "step": 13915
    },
    {
      "epoch": 2.5842154131847725,
      "grad_norm": 1.6025075092884642,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8859760761260986,
      "num_tokens": 484013678.0,
      "step": 13916
    },
    {
      "epoch": 2.5844011142061283,
      "grad_norm": 1.7521343067906558,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8771328926086426,
      "num_tokens": 484043868.0,
      "step": 13917
    },
    {
      "epoch": 2.5845868152274836,
      "grad_norm": 1.6283889521052564,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8825581073760986,
      "num_tokens": 484076518.0,
      "step": 13918
    },
    {
      "epoch": 2.5847725162488393,
      "grad_norm": 1.4334369468384218,
      "learning_rate": 1e-06,
      "loss": 0.3102,
      "mean_token_accuracy": 0.8918497562408447,
      "num_tokens": 484113776.0,
      "step": 13919
    },
    {
      "epoch": 2.584958217270195,
      "grad_norm": 1.5692105305538355,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8860839009284973,
      "num_tokens": 484144853.0,
      "step": 13920
    },
    {
      "epoch": 2.5851439182915508,
      "grad_norm": 1.7604051230197555,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8747403621673584,
      "num_tokens": 484175423.0,
      "step": 13921
    },
    {
      "epoch": 2.585329619312906,
      "grad_norm": 1.4327967453356636,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8919165134429932,
      "num_tokens": 484216065.0,
      "step": 13922
    },
    {
      "epoch": 2.585515320334262,
      "grad_norm": 1.5602877053165018,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8863959908485413,
      "num_tokens": 484251505.0,
      "step": 13923
    },
    {
      "epoch": 2.5857010213556175,
      "grad_norm": 1.4830924714377987,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8809617757797241,
      "num_tokens": 484292796.0,
      "step": 13924
    },
    {
      "epoch": 2.585886722376973,
      "grad_norm": 1.5729095450211243,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.892483115196228,
      "num_tokens": 484324708.0,
      "step": 13925
    },
    {
      "epoch": 2.5860724233983285,
      "grad_norm": 1.6376410851804295,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8934008479118347,
      "num_tokens": 484358505.0,
      "step": 13926
    },
    {
      "epoch": 2.5862581244196843,
      "grad_norm": 1.5089525846121523,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.889156699180603,
      "num_tokens": 484397403.0,
      "step": 13927
    },
    {
      "epoch": 2.58644382544104,
      "grad_norm": 1.646949990241157,
      "learning_rate": 1e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.8898635506629944,
      "num_tokens": 484426838.0,
      "step": 13928
    },
    {
      "epoch": 2.5866295264623957,
      "grad_norm": 1.4982582261826953,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.8948171138763428,
      "num_tokens": 484463245.0,
      "step": 13929
    },
    {
      "epoch": 2.586815227483751,
      "grad_norm": 1.597946948300045,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.890601634979248,
      "num_tokens": 484493939.0,
      "step": 13930
    },
    {
      "epoch": 2.5870009285051068,
      "grad_norm": 1.5538986954506817,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.889578104019165,
      "num_tokens": 484527227.0,
      "step": 13931
    },
    {
      "epoch": 2.5871866295264625,
      "grad_norm": 1.4783551269474404,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8915843367576599,
      "num_tokens": 484566415.0,
      "step": 13932
    },
    {
      "epoch": 2.587372330547818,
      "grad_norm": 1.4343072457994723,
      "learning_rate": 1e-06,
      "loss": 0.2831,
      "mean_token_accuracy": 0.8980841636657715,
      "num_tokens": 484602701.0,
      "step": 13933
    },
    {
      "epoch": 2.5875580315691735,
      "grad_norm": 1.678907187081584,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8875844478607178,
      "num_tokens": 484634237.0,
      "step": 13934
    },
    {
      "epoch": 2.5877437325905293,
      "grad_norm": 1.5833575187764821,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8822993040084839,
      "num_tokens": 484669693.0,
      "step": 13935
    },
    {
      "epoch": 2.587929433611885,
      "grad_norm": 1.6309199307035858,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8785330653190613,
      "num_tokens": 484705330.0,
      "step": 13936
    },
    {
      "epoch": 2.5881151346332407,
      "grad_norm": 1.5645203473450848,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8901451826095581,
      "num_tokens": 484738931.0,
      "step": 13937
    },
    {
      "epoch": 2.588300835654596,
      "grad_norm": 1.5831740042186127,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8875278830528259,
      "num_tokens": 484774276.0,
      "step": 13938
    },
    {
      "epoch": 2.5884865366759517,
      "grad_norm": 1.6742424767279103,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8762508630752563,
      "num_tokens": 484812172.0,
      "step": 13939
    },
    {
      "epoch": 2.5886722376973075,
      "grad_norm": 1.59928745206785,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.8764451742172241,
      "num_tokens": 484843284.0,
      "step": 13940
    },
    {
      "epoch": 2.5888579387186628,
      "grad_norm": 1.5863395002120935,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8850517868995667,
      "num_tokens": 484879098.0,
      "step": 13941
    },
    {
      "epoch": 2.5890436397400185,
      "grad_norm": 1.5634697591766569,
      "learning_rate": 1e-06,
      "loss": 0.2889,
      "mean_token_accuracy": 0.8962415456771851,
      "num_tokens": 484912910.0,
      "step": 13942
    },
    {
      "epoch": 2.5892293407613742,
      "grad_norm": 1.492131349710115,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8888847231864929,
      "num_tokens": 484951661.0,
      "step": 13943
    },
    {
      "epoch": 2.58941504178273,
      "grad_norm": 1.5062896006623114,
      "learning_rate": 1e-06,
      "loss": 0.295,
      "mean_token_accuracy": 0.8949539661407471,
      "num_tokens": 484987760.0,
      "step": 13944
    },
    {
      "epoch": 2.5896007428040857,
      "grad_norm": 1.4262603202809843,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8887792825698853,
      "num_tokens": 485026093.0,
      "step": 13945
    },
    {
      "epoch": 2.589786443825441,
      "grad_norm": 1.6678616411424227,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8690484166145325,
      "num_tokens": 485060226.0,
      "step": 13946
    },
    {
      "epoch": 2.5899721448467967,
      "grad_norm": 1.5988009216597006,
      "learning_rate": 1e-06,
      "loss": 0.2805,
      "mean_token_accuracy": 0.9003692865371704,
      "num_tokens": 485090237.0,
      "step": 13947
    },
    {
      "epoch": 2.590157845868152,
      "grad_norm": 1.4926346139739124,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8835693597793579,
      "num_tokens": 485126559.0,
      "step": 13948
    },
    {
      "epoch": 2.5903435468895077,
      "grad_norm": 1.5620113683193564,
      "learning_rate": 1e-06,
      "loss": 0.3565,
      "mean_token_accuracy": 0.876549243927002,
      "num_tokens": 485164377.0,
      "step": 13949
    },
    {
      "epoch": 2.5905292479108635,
      "grad_norm": 1.6614635504638127,
      "learning_rate": 1e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8675229549407959,
      "num_tokens": 485199471.0,
      "step": 13950
    },
    {
      "epoch": 2.590714948932219,
      "grad_norm": 1.481820724748297,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8877774477005005,
      "num_tokens": 485237065.0,
      "step": 13951
    },
    {
      "epoch": 2.590900649953575,
      "grad_norm": 1.7693687889901044,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8700034022331238,
      "num_tokens": 485267983.0,
      "step": 13952
    },
    {
      "epoch": 2.5910863509749302,
      "grad_norm": 1.4181199808036604,
      "learning_rate": 1e-06,
      "loss": 0.2869,
      "mean_token_accuracy": 0.8946201801300049,
      "num_tokens": 485307734.0,
      "step": 13953
    },
    {
      "epoch": 2.591272051996286,
      "grad_norm": 1.7262147645202435,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8887115120887756,
      "num_tokens": 485340326.0,
      "step": 13954
    },
    {
      "epoch": 2.5914577530176417,
      "grad_norm": 1.4760551858907025,
      "learning_rate": 1e-06,
      "loss": 0.305,
      "mean_token_accuracy": 0.8892534375190735,
      "num_tokens": 485377160.0,
      "step": 13955
    },
    {
      "epoch": 2.591643454038997,
      "grad_norm": 1.6107845862935664,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8923212885856628,
      "num_tokens": 485408677.0,
      "step": 13956
    },
    {
      "epoch": 2.5918291550603527,
      "grad_norm": 1.5444570380148113,
      "learning_rate": 1e-06,
      "loss": 0.2965,
      "mean_token_accuracy": 0.896202564239502,
      "num_tokens": 485441041.0,
      "step": 13957
    },
    {
      "epoch": 2.5920148560817085,
      "grad_norm": 1.5552635832060726,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8839521408081055,
      "num_tokens": 485475875.0,
      "step": 13958
    },
    {
      "epoch": 2.592200557103064,
      "grad_norm": 1.60882717567319,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8771188855171204,
      "num_tokens": 485510708.0,
      "step": 13959
    },
    {
      "epoch": 2.59238625812442,
      "grad_norm": 1.586914298440903,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8888320922851562,
      "num_tokens": 485541551.0,
      "step": 13960
    },
    {
      "epoch": 2.592571959145775,
      "grad_norm": 1.5823383825043376,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8836656808853149,
      "num_tokens": 485578357.0,
      "step": 13961
    },
    {
      "epoch": 2.592757660167131,
      "grad_norm": 1.632328648724232,
      "learning_rate": 1e-06,
      "loss": 0.3412,
      "mean_token_accuracy": 0.8808795213699341,
      "num_tokens": 485609953.0,
      "step": 13962
    },
    {
      "epoch": 2.5929433611884867,
      "grad_norm": 1.599592445522634,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8833419680595398,
      "num_tokens": 485644660.0,
      "step": 13963
    },
    {
      "epoch": 2.593129062209842,
      "grad_norm": 1.4253289729838146,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8770883083343506,
      "num_tokens": 485689265.0,
      "step": 13964
    },
    {
      "epoch": 2.5933147632311977,
      "grad_norm": 1.6152824378177681,
      "learning_rate": 1e-06,
      "loss": 0.2915,
      "mean_token_accuracy": 0.897434413433075,
      "num_tokens": 485718838.0,
      "step": 13965
    },
    {
      "epoch": 2.5935004642525534,
      "grad_norm": 1.6001995861354146,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8784637451171875,
      "num_tokens": 485752965.0,
      "step": 13966
    },
    {
      "epoch": 2.593686165273909,
      "grad_norm": 1.4708565714119473,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8779215812683105,
      "num_tokens": 485793095.0,
      "step": 13967
    },
    {
      "epoch": 2.593871866295265,
      "grad_norm": 1.734602844110098,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.89174485206604,
      "num_tokens": 485820604.0,
      "step": 13968
    },
    {
      "epoch": 2.59405756731662,
      "grad_norm": 1.5990789934366711,
      "learning_rate": 1e-06,
      "loss": 0.2748,
      "mean_token_accuracy": 0.898746907711029,
      "num_tokens": 485851676.0,
      "step": 13969
    },
    {
      "epoch": 2.594243268337976,
      "grad_norm": 1.505995797863241,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8877214789390564,
      "num_tokens": 485887951.0,
      "step": 13970
    },
    {
      "epoch": 2.594428969359331,
      "grad_norm": 1.681230077206535,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8852088451385498,
      "num_tokens": 485919566.0,
      "step": 13971
    },
    {
      "epoch": 2.594614670380687,
      "grad_norm": 1.5034880278444984,
      "learning_rate": 1e-06,
      "loss": 0.283,
      "mean_token_accuracy": 0.9006853103637695,
      "num_tokens": 485958344.0,
      "step": 13972
    },
    {
      "epoch": 2.5948003714020427,
      "grad_norm": 1.494537906725082,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8846195936203003,
      "num_tokens": 485997471.0,
      "step": 13973
    },
    {
      "epoch": 2.5949860724233984,
      "grad_norm": 1.6470427923845559,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8834992051124573,
      "num_tokens": 486029469.0,
      "step": 13974
    },
    {
      "epoch": 2.595171773444754,
      "grad_norm": 1.5618846645024067,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8957536220550537,
      "num_tokens": 486061257.0,
      "step": 13975
    },
    {
      "epoch": 2.5953574744661094,
      "grad_norm": 1.729866452864282,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8883439302444458,
      "num_tokens": 486090038.0,
      "step": 13976
    },
    {
      "epoch": 2.595543175487465,
      "grad_norm": 1.5660411378273509,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8834503293037415,
      "num_tokens": 486129368.0,
      "step": 13977
    },
    {
      "epoch": 2.595728876508821,
      "grad_norm": 1.6080082769047646,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8931059837341309,
      "num_tokens": 486160443.0,
      "step": 13978
    },
    {
      "epoch": 2.595914577530176,
      "grad_norm": 1.6504399525840656,
      "learning_rate": 1e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.8731367588043213,
      "num_tokens": 486196253.0,
      "step": 13979
    },
    {
      "epoch": 2.596100278551532,
      "grad_norm": 1.718307356225345,
      "learning_rate": 1e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8825160264968872,
      "num_tokens": 486224454.0,
      "step": 13980
    },
    {
      "epoch": 2.5962859795728876,
      "grad_norm": 1.5559166502605932,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8867178559303284,
      "num_tokens": 486262971.0,
      "step": 13981
    },
    {
      "epoch": 2.5964716805942434,
      "grad_norm": 1.6145549680674254,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8809952735900879,
      "num_tokens": 486296460.0,
      "step": 13982
    },
    {
      "epoch": 2.596657381615599,
      "grad_norm": 1.5079596359278165,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8865805268287659,
      "num_tokens": 486333780.0,
      "step": 13983
    },
    {
      "epoch": 2.5968430826369544,
      "grad_norm": 1.557136490224497,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8896505236625671,
      "num_tokens": 486370144.0,
      "step": 13984
    },
    {
      "epoch": 2.59702878365831,
      "grad_norm": 1.4773330504996003,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8828592896461487,
      "num_tokens": 486407814.0,
      "step": 13985
    },
    {
      "epoch": 2.597214484679666,
      "grad_norm": 1.575018659357369,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8865578174591064,
      "num_tokens": 486439658.0,
      "step": 13986
    },
    {
      "epoch": 2.597400185701021,
      "grad_norm": 1.6961678282829689,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8892890214920044,
      "num_tokens": 486469194.0,
      "step": 13987
    },
    {
      "epoch": 2.597585886722377,
      "grad_norm": 1.605815013687313,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8894306421279907,
      "num_tokens": 486502892.0,
      "step": 13988
    },
    {
      "epoch": 2.5977715877437326,
      "grad_norm": 1.4253438436885106,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8925460577011108,
      "num_tokens": 486545090.0,
      "step": 13989
    },
    {
      "epoch": 2.5979572887650884,
      "grad_norm": 1.6029133799826216,
      "learning_rate": 1e-06,
      "loss": 0.3466,
      "mean_token_accuracy": 0.8790906667709351,
      "num_tokens": 486583340.0,
      "step": 13990
    },
    {
      "epoch": 2.598142989786444,
      "grad_norm": 1.517033535073679,
      "learning_rate": 1e-06,
      "loss": 0.3397,
      "mean_token_accuracy": 0.8826486468315125,
      "num_tokens": 486616906.0,
      "step": 13991
    },
    {
      "epoch": 2.5983286908077994,
      "grad_norm": 1.6227469590138004,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8827046155929565,
      "num_tokens": 486650569.0,
      "step": 13992
    },
    {
      "epoch": 2.598514391829155,
      "grad_norm": 1.8605656047920818,
      "learning_rate": 1e-06,
      "loss": 0.3502,
      "mean_token_accuracy": 0.8803629875183105,
      "num_tokens": 486677356.0,
      "step": 13993
    },
    {
      "epoch": 2.5987000928505104,
      "grad_norm": 1.5983152393716744,
      "learning_rate": 1e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8762756586074829,
      "num_tokens": 486715201.0,
      "step": 13994
    },
    {
      "epoch": 2.598885793871866,
      "grad_norm": 1.6762237806867,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8819774985313416,
      "num_tokens": 486747810.0,
      "step": 13995
    },
    {
      "epoch": 2.599071494893222,
      "grad_norm": 1.5882928247043646,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8863667249679565,
      "num_tokens": 486779703.0,
      "step": 13996
    },
    {
      "epoch": 2.5992571959145776,
      "grad_norm": 1.5000619054117823,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8802363872528076,
      "num_tokens": 486820225.0,
      "step": 13997
    },
    {
      "epoch": 2.5994428969359333,
      "grad_norm": 1.479674026928807,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8909018039703369,
      "num_tokens": 486855030.0,
      "step": 13998
    },
    {
      "epoch": 2.5996285979572886,
      "grad_norm": 1.5695232483371429,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8735540509223938,
      "num_tokens": 486891337.0,
      "step": 13999
    },
    {
      "epoch": 2.5998142989786444,
      "grad_norm": 1.5577232654264448,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8753421902656555,
      "num_tokens": 486932093.0,
      "step": 14000
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.604106442252079,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8742375373840332,
      "num_tokens": 486968865.0,
      "step": 14001
    },
    {
      "epoch": 2.6001857010213554,
      "grad_norm": 1.5342646929492936,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8951430916786194,
      "num_tokens": 487001881.0,
      "step": 14002
    },
    {
      "epoch": 2.600371402042711,
      "grad_norm": 1.6804618823669373,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8827574253082275,
      "num_tokens": 487036075.0,
      "step": 14003
    },
    {
      "epoch": 2.600557103064067,
      "grad_norm": 1.6594786202838043,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8869807720184326,
      "num_tokens": 487066116.0,
      "step": 14004
    },
    {
      "epoch": 2.6007428040854226,
      "grad_norm": 1.5102051218836225,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8855490684509277,
      "num_tokens": 487105691.0,
      "step": 14005
    },
    {
      "epoch": 2.6009285051067783,
      "grad_norm": 1.6242546762730083,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8833820819854736,
      "num_tokens": 487139253.0,
      "step": 14006
    },
    {
      "epoch": 2.6011142061281336,
      "grad_norm": 1.4602325297839853,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.88532954454422,
      "num_tokens": 487181060.0,
      "step": 14007
    },
    {
      "epoch": 2.6012999071494893,
      "grad_norm": 1.535351653893282,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8757634162902832,
      "num_tokens": 487223208.0,
      "step": 14008
    },
    {
      "epoch": 2.601485608170845,
      "grad_norm": 1.680517639283985,
      "learning_rate": 1e-06,
      "loss": 0.2886,
      "mean_token_accuracy": 0.8977587223052979,
      "num_tokens": 487253955.0,
      "step": 14009
    },
    {
      "epoch": 2.6016713091922004,
      "grad_norm": 1.7055810209088802,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8933715224266052,
      "num_tokens": 487282762.0,
      "step": 14010
    },
    {
      "epoch": 2.601857010213556,
      "grad_norm": 1.5723337739097354,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8745357990264893,
      "num_tokens": 487321448.0,
      "step": 14011
    },
    {
      "epoch": 2.602042711234912,
      "grad_norm": 1.6460011489766082,
      "learning_rate": 1e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8575932383537292,
      "num_tokens": 487355072.0,
      "step": 14012
    },
    {
      "epoch": 2.6022284122562676,
      "grad_norm": 1.5010903615754438,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.878125011920929,
      "num_tokens": 487395824.0,
      "step": 14013
    },
    {
      "epoch": 2.6024141132776233,
      "grad_norm": 1.6072540089675373,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8917491436004639,
      "num_tokens": 487428809.0,
      "step": 14014
    },
    {
      "epoch": 2.6025998142989786,
      "grad_norm": 1.6954057608008135,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8824042081832886,
      "num_tokens": 487462961.0,
      "step": 14015
    },
    {
      "epoch": 2.6027855153203343,
      "grad_norm": 1.8408165136455445,
      "learning_rate": 1e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.8761458396911621,
      "num_tokens": 487494357.0,
      "step": 14016
    },
    {
      "epoch": 2.60297121634169,
      "grad_norm": 1.513834073075014,
      "learning_rate": 1e-06,
      "loss": 0.2993,
      "mean_token_accuracy": 0.8918896913528442,
      "num_tokens": 487526209.0,
      "step": 14017
    },
    {
      "epoch": 2.6031569173630453,
      "grad_norm": 1.562843560097285,
      "learning_rate": 1e-06,
      "loss": 0.3107,
      "mean_token_accuracy": 0.8928707838058472,
      "num_tokens": 487559548.0,
      "step": 14018
    },
    {
      "epoch": 2.603342618384401,
      "grad_norm": 1.5013206887297696,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.8793070316314697,
      "num_tokens": 487598611.0,
      "step": 14019
    },
    {
      "epoch": 2.603528319405757,
      "grad_norm": 1.6344709186348483,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8893073797225952,
      "num_tokens": 487628241.0,
      "step": 14020
    },
    {
      "epoch": 2.6037140204271125,
      "grad_norm": 1.3987421086472218,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8963597416877747,
      "num_tokens": 487665258.0,
      "step": 14021
    },
    {
      "epoch": 2.603899721448468,
      "grad_norm": 1.5117799573353015,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8975151181221008,
      "num_tokens": 487700260.0,
      "step": 14022
    },
    {
      "epoch": 2.6040854224698236,
      "grad_norm": 1.5810828573140745,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8813053369522095,
      "num_tokens": 487734736.0,
      "step": 14023
    },
    {
      "epoch": 2.6042711234911793,
      "grad_norm": 1.640111853449644,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.878089189529419,
      "num_tokens": 487767782.0,
      "step": 14024
    },
    {
      "epoch": 2.6044568245125346,
      "grad_norm": 1.5120894747039566,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8885635137557983,
      "num_tokens": 487803126.0,
      "step": 14025
    },
    {
      "epoch": 2.6046425255338903,
      "grad_norm": 1.4492456401756801,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.8934425115585327,
      "num_tokens": 487843688.0,
      "step": 14026
    },
    {
      "epoch": 2.604828226555246,
      "grad_norm": 1.5044406526821557,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8830721378326416,
      "num_tokens": 487883035.0,
      "step": 14027
    },
    {
      "epoch": 2.6050139275766018,
      "grad_norm": 1.6845393706182499,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8945555686950684,
      "num_tokens": 487912048.0,
      "step": 14028
    },
    {
      "epoch": 2.6051996285979575,
      "grad_norm": 1.5095484083567543,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.903624951839447,
      "num_tokens": 487945525.0,
      "step": 14029
    },
    {
      "epoch": 2.605385329619313,
      "grad_norm": 1.6055040660214561,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8944551944732666,
      "num_tokens": 487980349.0,
      "step": 14030
    },
    {
      "epoch": 2.6055710306406685,
      "grad_norm": 1.7147917435121836,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8696591854095459,
      "num_tokens": 488016953.0,
      "step": 14031
    },
    {
      "epoch": 2.6057567316620243,
      "grad_norm": 1.6626409807759193,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.883399248123169,
      "num_tokens": 488051045.0,
      "step": 14032
    },
    {
      "epoch": 2.6059424326833796,
      "grad_norm": 1.5385604711863266,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8882901072502136,
      "num_tokens": 488084315.0,
      "step": 14033
    },
    {
      "epoch": 2.6061281337047353,
      "grad_norm": 1.6312034683114545,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8838935494422913,
      "num_tokens": 488118033.0,
      "step": 14034
    },
    {
      "epoch": 2.606313834726091,
      "grad_norm": 1.4768514116059501,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8882892727851868,
      "num_tokens": 488154467.0,
      "step": 14035
    },
    {
      "epoch": 2.6064995357474467,
      "grad_norm": 1.4876291373905233,
      "learning_rate": 1e-06,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8758102655410767,
      "num_tokens": 488194118.0,
      "step": 14036
    },
    {
      "epoch": 2.6066852367688025,
      "grad_norm": 1.5231478579218216,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8836687803268433,
      "num_tokens": 488230010.0,
      "step": 14037
    },
    {
      "epoch": 2.6068709377901578,
      "grad_norm": 1.4242781717384956,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8775651454925537,
      "num_tokens": 488271602.0,
      "step": 14038
    },
    {
      "epoch": 2.6070566388115135,
      "grad_norm": 1.4937760598706853,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8965615630149841,
      "num_tokens": 488306426.0,
      "step": 14039
    },
    {
      "epoch": 2.6072423398328692,
      "grad_norm": 1.451959634397009,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8980720639228821,
      "num_tokens": 488343285.0,
      "step": 14040
    },
    {
      "epoch": 2.6074280408542245,
      "grad_norm": 1.4543656463605232,
      "learning_rate": 1e-06,
      "loss": 0.3151,
      "mean_token_accuracy": 0.8901386260986328,
      "num_tokens": 488380555.0,
      "step": 14041
    },
    {
      "epoch": 2.6076137418755803,
      "grad_norm": 1.4688008124706426,
      "learning_rate": 1e-06,
      "loss": 0.2729,
      "mean_token_accuracy": 0.9029704928398132,
      "num_tokens": 488414835.0,
      "step": 14042
    },
    {
      "epoch": 2.607799442896936,
      "grad_norm": 1.5057981802146923,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8785170316696167,
      "num_tokens": 488453377.0,
      "step": 14043
    },
    {
      "epoch": 2.6079851439182917,
      "grad_norm": 1.5346992267934814,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8864452242851257,
      "num_tokens": 488489955.0,
      "step": 14044
    },
    {
      "epoch": 2.608170844939647,
      "grad_norm": 1.585513432481429,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8879697322845459,
      "num_tokens": 488522900.0,
      "step": 14045
    },
    {
      "epoch": 2.6083565459610027,
      "grad_norm": 1.5416777885923798,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8845584392547607,
      "num_tokens": 488558114.0,
      "step": 14046
    },
    {
      "epoch": 2.6085422469823585,
      "grad_norm": 1.6467934789435332,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8890581727027893,
      "num_tokens": 488586270.0,
      "step": 14047
    },
    {
      "epoch": 2.6087279480037138,
      "grad_norm": 1.7091045115223193,
      "learning_rate": 1e-06,
      "loss": 0.3594,
      "mean_token_accuracy": 0.8741198182106018,
      "num_tokens": 488624786.0,
      "step": 14048
    },
    {
      "epoch": 2.6089136490250695,
      "grad_norm": 1.620796321214138,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8925714492797852,
      "num_tokens": 488655593.0,
      "step": 14049
    },
    {
      "epoch": 2.6090993500464252,
      "grad_norm": 1.785986956617848,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.886254072189331,
      "num_tokens": 488684097.0,
      "step": 14050
    },
    {
      "epoch": 2.609285051067781,
      "grad_norm": 1.5257958291270637,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8921465873718262,
      "num_tokens": 488717627.0,
      "step": 14051
    },
    {
      "epoch": 2.6094707520891367,
      "grad_norm": 1.4595808244975785,
      "learning_rate": 1e-06,
      "loss": 0.2967,
      "mean_token_accuracy": 0.8942515850067139,
      "num_tokens": 488757194.0,
      "step": 14052
    },
    {
      "epoch": 2.609656453110492,
      "grad_norm": 1.4068934928874113,
      "learning_rate": 1e-06,
      "loss": 0.2703,
      "mean_token_accuracy": 0.9041882157325745,
      "num_tokens": 488793335.0,
      "step": 14053
    },
    {
      "epoch": 2.6098421541318477,
      "grad_norm": 1.6495001821086839,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8783841133117676,
      "num_tokens": 488824809.0,
      "step": 14054
    },
    {
      "epoch": 2.6100278551532035,
      "grad_norm": 1.5741916668880223,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8895179033279419,
      "num_tokens": 488858482.0,
      "step": 14055
    },
    {
      "epoch": 2.6102135561745587,
      "grad_norm": 1.6451097250212672,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8854491114616394,
      "num_tokens": 488890910.0,
      "step": 14056
    },
    {
      "epoch": 2.6103992571959145,
      "grad_norm": 1.6339527806257548,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8881047368049622,
      "num_tokens": 488921487.0,
      "step": 14057
    },
    {
      "epoch": 2.61058495821727,
      "grad_norm": 1.5538443132793438,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8831026554107666,
      "num_tokens": 488955711.0,
      "step": 14058
    },
    {
      "epoch": 2.610770659238626,
      "grad_norm": 1.4034271137581305,
      "learning_rate": 1e-06,
      "loss": 0.273,
      "mean_token_accuracy": 0.9030892848968506,
      "num_tokens": 488994554.0,
      "step": 14059
    },
    {
      "epoch": 2.6109563602599817,
      "grad_norm": 1.5577514209780627,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8960354924201965,
      "num_tokens": 489027379.0,
      "step": 14060
    },
    {
      "epoch": 2.611142061281337,
      "grad_norm": 1.5536471369183236,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8776884078979492,
      "num_tokens": 489065170.0,
      "step": 14061
    },
    {
      "epoch": 2.6113277623026927,
      "grad_norm": 1.6047402603579042,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8856073617935181,
      "num_tokens": 489097416.0,
      "step": 14062
    },
    {
      "epoch": 2.6115134633240484,
      "grad_norm": 1.5759451980606358,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8909693360328674,
      "num_tokens": 489133135.0,
      "step": 14063
    },
    {
      "epoch": 2.6116991643454037,
      "grad_norm": 1.663644224108355,
      "learning_rate": 1e-06,
      "loss": 0.3645,
      "mean_token_accuracy": 0.8755310773849487,
      "num_tokens": 489168531.0,
      "step": 14064
    },
    {
      "epoch": 2.6118848653667595,
      "grad_norm": 1.6307034729902408,
      "learning_rate": 1e-06,
      "loss": 0.3688,
      "mean_token_accuracy": 0.8746076822280884,
      "num_tokens": 489204202.0,
      "step": 14065
    },
    {
      "epoch": 2.612070566388115,
      "grad_norm": 1.434585501634606,
      "learning_rate": 1e-06,
      "loss": 0.3324,
      "mean_token_accuracy": 0.8844872713088989,
      "num_tokens": 489248430.0,
      "step": 14066
    },
    {
      "epoch": 2.612256267409471,
      "grad_norm": 1.4931571160155255,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8813769817352295,
      "num_tokens": 489286418.0,
      "step": 14067
    },
    {
      "epoch": 2.612441968430826,
      "grad_norm": 1.429260629349158,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8825434446334839,
      "num_tokens": 489327797.0,
      "step": 14068
    },
    {
      "epoch": 2.612627669452182,
      "grad_norm": 1.591490328352719,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.8916811943054199,
      "num_tokens": 489358798.0,
      "step": 14069
    },
    {
      "epoch": 2.6128133704735377,
      "grad_norm": 1.6161635500999325,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8889422416687012,
      "num_tokens": 489393292.0,
      "step": 14070
    },
    {
      "epoch": 2.612999071494893,
      "grad_norm": 1.6395390027286543,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.880117654800415,
      "num_tokens": 489427098.0,
      "step": 14071
    },
    {
      "epoch": 2.6131847725162487,
      "grad_norm": 1.493667447586655,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8923132419586182,
      "num_tokens": 489465121.0,
      "step": 14072
    },
    {
      "epoch": 2.6133704735376044,
      "grad_norm": 1.3672992498593919,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.8943303823471069,
      "num_tokens": 489508491.0,
      "step": 14073
    },
    {
      "epoch": 2.61355617455896,
      "grad_norm": 1.5849785996037529,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8796826601028442,
      "num_tokens": 489543056.0,
      "step": 14074
    },
    {
      "epoch": 2.613741875580316,
      "grad_norm": 1.5590527648521442,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8919941186904907,
      "num_tokens": 489578294.0,
      "step": 14075
    },
    {
      "epoch": 2.613927576601671,
      "grad_norm": 1.5196290414109015,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8841742277145386,
      "num_tokens": 489620578.0,
      "step": 14076
    },
    {
      "epoch": 2.614113277623027,
      "grad_norm": 1.5522976130137198,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8955751657485962,
      "num_tokens": 489655587.0,
      "step": 14077
    },
    {
      "epoch": 2.6142989786443827,
      "grad_norm": 1.5767375057596407,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8835354447364807,
      "num_tokens": 489690147.0,
      "step": 14078
    },
    {
      "epoch": 2.614484679665738,
      "grad_norm": 1.6607316809415358,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8783963918685913,
      "num_tokens": 489725804.0,
      "step": 14079
    },
    {
      "epoch": 2.6146703806870937,
      "grad_norm": 1.6147327714283297,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8882551193237305,
      "num_tokens": 489754965.0,
      "step": 14080
    },
    {
      "epoch": 2.6148560817084494,
      "grad_norm": 1.609478946093281,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.887546956539154,
      "num_tokens": 489788929.0,
      "step": 14081
    },
    {
      "epoch": 2.615041782729805,
      "grad_norm": 1.638059948356434,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8765758275985718,
      "num_tokens": 489824028.0,
      "step": 14082
    },
    {
      "epoch": 2.615227483751161,
      "grad_norm": 1.4380109471723355,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8965846300125122,
      "num_tokens": 489861596.0,
      "step": 14083
    },
    {
      "epoch": 2.615413184772516,
      "grad_norm": 1.6394146410619108,
      "learning_rate": 1e-06,
      "loss": 0.2981,
      "mean_token_accuracy": 0.893519401550293,
      "num_tokens": 489891560.0,
      "step": 14084
    },
    {
      "epoch": 2.615598885793872,
      "grad_norm": 1.568776631995774,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8846328258514404,
      "num_tokens": 489927002.0,
      "step": 14085
    },
    {
      "epoch": 2.6157845868152276,
      "grad_norm": 1.582007447727133,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8846465945243835,
      "num_tokens": 489963390.0,
      "step": 14086
    },
    {
      "epoch": 2.615970287836583,
      "grad_norm": 1.6055386130319562,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8928934335708618,
      "num_tokens": 489995853.0,
      "step": 14087
    },
    {
      "epoch": 2.6161559888579387,
      "grad_norm": 1.625700640383849,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8851526379585266,
      "num_tokens": 490029240.0,
      "step": 14088
    },
    {
      "epoch": 2.6163416898792944,
      "grad_norm": 1.6421268655153036,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.887069046497345,
      "num_tokens": 490064952.0,
      "step": 14089
    },
    {
      "epoch": 2.61652739090065,
      "grad_norm": 1.6063415277045243,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8872970342636108,
      "num_tokens": 490097453.0,
      "step": 14090
    },
    {
      "epoch": 2.6167130919220054,
      "grad_norm": 1.3477522016278147,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.891028881072998,
      "num_tokens": 490137822.0,
      "step": 14091
    },
    {
      "epoch": 2.616898792943361,
      "grad_norm": 1.5767066934161493,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8714675903320312,
      "num_tokens": 490175591.0,
      "step": 14092
    },
    {
      "epoch": 2.617084493964717,
      "grad_norm": 1.5321042615649214,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.886142909526825,
      "num_tokens": 490211319.0,
      "step": 14093
    },
    {
      "epoch": 2.617270194986072,
      "grad_norm": 1.5918328970135238,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.88514244556427,
      "num_tokens": 490245459.0,
      "step": 14094
    },
    {
      "epoch": 2.617455896007428,
      "grad_norm": 1.451853376370069,
      "learning_rate": 1e-06,
      "loss": 0.2571,
      "mean_token_accuracy": 0.9068317413330078,
      "num_tokens": 490277123.0,
      "step": 14095
    },
    {
      "epoch": 2.6176415970287836,
      "grad_norm": 1.5853684293967156,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8906680345535278,
      "num_tokens": 490308583.0,
      "step": 14096
    },
    {
      "epoch": 2.6178272980501394,
      "grad_norm": 1.8453938838954644,
      "learning_rate": 1e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8630337715148926,
      "num_tokens": 490337624.0,
      "step": 14097
    },
    {
      "epoch": 2.618012999071495,
      "grad_norm": 1.6994546969519462,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8757339715957642,
      "num_tokens": 490372174.0,
      "step": 14098
    },
    {
      "epoch": 2.6181987000928504,
      "grad_norm": 1.4627553692653499,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8852324485778809,
      "num_tokens": 490411615.0,
      "step": 14099
    },
    {
      "epoch": 2.618384401114206,
      "grad_norm": 1.5910560533689189,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8773472309112549,
      "num_tokens": 490448278.0,
      "step": 14100
    },
    {
      "epoch": 2.618570102135562,
      "grad_norm": 1.5736268972077805,
      "learning_rate": 1e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8764071464538574,
      "num_tokens": 490487240.0,
      "step": 14101
    },
    {
      "epoch": 2.618755803156917,
      "grad_norm": 1.5787284711239318,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.896752655506134,
      "num_tokens": 490518587.0,
      "step": 14102
    },
    {
      "epoch": 2.618941504178273,
      "grad_norm": 1.6928474722297366,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8872455954551697,
      "num_tokens": 490550835.0,
      "step": 14103
    },
    {
      "epoch": 2.6191272051996286,
      "grad_norm": 1.4698635415948487,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8962244987487793,
      "num_tokens": 490586589.0,
      "step": 14104
    },
    {
      "epoch": 2.6193129062209843,
      "grad_norm": 1.5294616993808245,
      "learning_rate": 1e-06,
      "loss": 0.2926,
      "mean_token_accuracy": 0.8950010538101196,
      "num_tokens": 490622075.0,
      "step": 14105
    },
    {
      "epoch": 2.61949860724234,
      "grad_norm": 1.6135072839595728,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8909518718719482,
      "num_tokens": 490652341.0,
      "step": 14106
    },
    {
      "epoch": 2.6196843082636954,
      "grad_norm": 1.907237253597755,
      "learning_rate": 1e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.866187572479248,
      "num_tokens": 490680700.0,
      "step": 14107
    },
    {
      "epoch": 2.619870009285051,
      "grad_norm": 1.5855218242007085,
      "learning_rate": 1e-06,
      "loss": 0.2786,
      "mean_token_accuracy": 0.898654580116272,
      "num_tokens": 490710296.0,
      "step": 14108
    },
    {
      "epoch": 2.620055710306407,
      "grad_norm": 1.5385137435085694,
      "learning_rate": 1e-06,
      "loss": 0.2894,
      "mean_token_accuracy": 0.8981837630271912,
      "num_tokens": 490740145.0,
      "step": 14109
    },
    {
      "epoch": 2.620241411327762,
      "grad_norm": 1.7068137251367292,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8777248859405518,
      "num_tokens": 490769540.0,
      "step": 14110
    },
    {
      "epoch": 2.620427112349118,
      "grad_norm": 1.4782268316020055,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8897008895874023,
      "num_tokens": 490805977.0,
      "step": 14111
    },
    {
      "epoch": 2.6206128133704736,
      "grad_norm": 1.524828015389876,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8783154487609863,
      "num_tokens": 490842911.0,
      "step": 14112
    },
    {
      "epoch": 2.6207985143918293,
      "grad_norm": 1.5015502821043865,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8734973073005676,
      "num_tokens": 490882517.0,
      "step": 14113
    },
    {
      "epoch": 2.620984215413185,
      "grad_norm": 1.6577712426347189,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.887397289276123,
      "num_tokens": 490911652.0,
      "step": 14114
    },
    {
      "epoch": 2.6211699164345403,
      "grad_norm": 1.476830430557552,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.889435887336731,
      "num_tokens": 490946730.0,
      "step": 14115
    },
    {
      "epoch": 2.621355617455896,
      "grad_norm": 1.3694211143529762,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8871471881866455,
      "num_tokens": 490987936.0,
      "step": 14116
    },
    {
      "epoch": 2.6215413184772514,
      "grad_norm": 1.7246233207139234,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8842004537582397,
      "num_tokens": 491016511.0,
      "step": 14117
    },
    {
      "epoch": 2.621727019498607,
      "grad_norm": 1.714515396722078,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8714449405670166,
      "num_tokens": 491049205.0,
      "step": 14118
    },
    {
      "epoch": 2.621912720519963,
      "grad_norm": 1.5034797565529072,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8860968351364136,
      "num_tokens": 491087043.0,
      "step": 14119
    },
    {
      "epoch": 2.6220984215413186,
      "grad_norm": 1.6027284167071787,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.879231333732605,
      "num_tokens": 491120916.0,
      "step": 14120
    },
    {
      "epoch": 2.6222841225626743,
      "grad_norm": 1.4422154629745196,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8908265829086304,
      "num_tokens": 491163177.0,
      "step": 14121
    },
    {
      "epoch": 2.6224698235840296,
      "grad_norm": 1.5708912399067971,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.89250248670578,
      "num_tokens": 491199242.0,
      "step": 14122
    },
    {
      "epoch": 2.6226555246053853,
      "grad_norm": 1.6063875554715261,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.891433596611023,
      "num_tokens": 491231032.0,
      "step": 14123
    },
    {
      "epoch": 2.622841225626741,
      "grad_norm": 1.6034880062209853,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8773902654647827,
      "num_tokens": 491266782.0,
      "step": 14124
    },
    {
      "epoch": 2.6230269266480963,
      "grad_norm": 1.5370600443177371,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8913838267326355,
      "num_tokens": 491302515.0,
      "step": 14125
    },
    {
      "epoch": 2.623212627669452,
      "grad_norm": 1.5108478083026096,
      "learning_rate": 1e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8759094476699829,
      "num_tokens": 491343463.0,
      "step": 14126
    },
    {
      "epoch": 2.623398328690808,
      "grad_norm": 1.654410988146399,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8954286575317383,
      "num_tokens": 491372432.0,
      "step": 14127
    },
    {
      "epoch": 2.6235840297121635,
      "grad_norm": 1.5399729645738331,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8781399726867676,
      "num_tokens": 491414170.0,
      "step": 14128
    },
    {
      "epoch": 2.6237697307335193,
      "grad_norm": 1.6672250645938655,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8782161474227905,
      "num_tokens": 491448386.0,
      "step": 14129
    },
    {
      "epoch": 2.6239554317548746,
      "grad_norm": 1.5746706077429125,
      "learning_rate": 1e-06,
      "loss": 0.291,
      "mean_token_accuracy": 0.8947118520736694,
      "num_tokens": 491484127.0,
      "step": 14130
    },
    {
      "epoch": 2.6241411327762303,
      "grad_norm": 1.5918698948718089,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8963795304298401,
      "num_tokens": 491516746.0,
      "step": 14131
    },
    {
      "epoch": 2.624326833797586,
      "grad_norm": 1.6293402596573157,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8813130855560303,
      "num_tokens": 491550605.0,
      "step": 14132
    },
    {
      "epoch": 2.6245125348189413,
      "grad_norm": 1.5480120886480782,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8833026885986328,
      "num_tokens": 491584381.0,
      "step": 14133
    },
    {
      "epoch": 2.624698235840297,
      "grad_norm": 1.6009025799380103,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8742035627365112,
      "num_tokens": 491623840.0,
      "step": 14134
    },
    {
      "epoch": 2.6248839368616528,
      "grad_norm": 1.6044054647594015,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8944852352142334,
      "num_tokens": 491655364.0,
      "step": 14135
    },
    {
      "epoch": 2.6250696378830085,
      "grad_norm": 1.4506192373649462,
      "learning_rate": 1e-06,
      "loss": 0.2841,
      "mean_token_accuracy": 0.899211049079895,
      "num_tokens": 491691389.0,
      "step": 14136
    },
    {
      "epoch": 2.6252553389043642,
      "grad_norm": 1.5847225717784035,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8841690421104431,
      "num_tokens": 491727231.0,
      "step": 14137
    },
    {
      "epoch": 2.6254410399257195,
      "grad_norm": 1.4531263602279862,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.893958330154419,
      "num_tokens": 491764119.0,
      "step": 14138
    },
    {
      "epoch": 2.6256267409470753,
      "grad_norm": 1.6608088259522484,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8824360966682434,
      "num_tokens": 491797098.0,
      "step": 14139
    },
    {
      "epoch": 2.6258124419684306,
      "grad_norm": 1.7892196866410348,
      "learning_rate": 1e-06,
      "loss": 0.3864,
      "mean_token_accuracy": 0.8696867227554321,
      "num_tokens": 491831719.0,
      "step": 14140
    },
    {
      "epoch": 2.6259981429897863,
      "grad_norm": 1.6670858850582226,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8780907392501831,
      "num_tokens": 491865839.0,
      "step": 14141
    },
    {
      "epoch": 2.626183844011142,
      "grad_norm": 1.6394282997450949,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8771617412567139,
      "num_tokens": 491904426.0,
      "step": 14142
    },
    {
      "epoch": 2.6263695450324978,
      "grad_norm": 1.6551311865584064,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.889816164970398,
      "num_tokens": 491936234.0,
      "step": 14143
    },
    {
      "epoch": 2.6265552460538535,
      "grad_norm": 1.5436937181310033,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.894441545009613,
      "num_tokens": 491972015.0,
      "step": 14144
    },
    {
      "epoch": 2.6267409470752088,
      "grad_norm": 1.6447945842070821,
      "learning_rate": 1e-06,
      "loss": 0.3408,
      "mean_token_accuracy": 0.8807712197303772,
      "num_tokens": 492006954.0,
      "step": 14145
    },
    {
      "epoch": 2.6269266480965645,
      "grad_norm": 1.5494321494095868,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.884247362613678,
      "num_tokens": 492046849.0,
      "step": 14146
    },
    {
      "epoch": 2.6271123491179202,
      "grad_norm": 1.5478678356241102,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8762108087539673,
      "num_tokens": 492085855.0,
      "step": 14147
    },
    {
      "epoch": 2.6272980501392755,
      "grad_norm": 1.5883785135345665,
      "learning_rate": 1e-06,
      "loss": 0.3309,
      "mean_token_accuracy": 0.8828762769699097,
      "num_tokens": 492122045.0,
      "step": 14148
    },
    {
      "epoch": 2.6274837511606313,
      "grad_norm": 1.5935936128091124,
      "learning_rate": 1e-06,
      "loss": 0.365,
      "mean_token_accuracy": 0.8728697896003723,
      "num_tokens": 492159248.0,
      "step": 14149
    },
    {
      "epoch": 2.627669452181987,
      "grad_norm": 1.4744470899365076,
      "learning_rate": 1e-06,
      "loss": 0.2904,
      "mean_token_accuracy": 0.8967746496200562,
      "num_tokens": 492198189.0,
      "step": 14150
    },
    {
      "epoch": 2.6278551532033427,
      "grad_norm": 1.5347444467475324,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8813787698745728,
      "num_tokens": 492235543.0,
      "step": 14151
    },
    {
      "epoch": 2.6280408542246985,
      "grad_norm": 1.4823015769954753,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8894864916801453,
      "num_tokens": 492271141.0,
      "step": 14152
    },
    {
      "epoch": 2.6282265552460538,
      "grad_norm": 1.5868812761932012,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8878764510154724,
      "num_tokens": 492304196.0,
      "step": 14153
    },
    {
      "epoch": 2.6284122562674095,
      "grad_norm": 1.6017786888359475,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8796591758728027,
      "num_tokens": 492342531.0,
      "step": 14154
    },
    {
      "epoch": 2.628597957288765,
      "grad_norm": 1.6304512997513276,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.889647364616394,
      "num_tokens": 492373564.0,
      "step": 14155
    },
    {
      "epoch": 2.6287836583101205,
      "grad_norm": 1.464921126726617,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8875505924224854,
      "num_tokens": 492412491.0,
      "step": 14156
    },
    {
      "epoch": 2.6289693593314762,
      "grad_norm": 1.6271264651774024,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8821860551834106,
      "num_tokens": 492444338.0,
      "step": 14157
    },
    {
      "epoch": 2.629155060352832,
      "grad_norm": 1.4496519773536332,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8909444808959961,
      "num_tokens": 492482218.0,
      "step": 14158
    },
    {
      "epoch": 2.6293407613741877,
      "grad_norm": 1.5107164145440026,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.8830645084381104,
      "num_tokens": 492521483.0,
      "step": 14159
    },
    {
      "epoch": 2.6295264623955434,
      "grad_norm": 1.5478269830968217,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8773755431175232,
      "num_tokens": 492559530.0,
      "step": 14160
    },
    {
      "epoch": 2.6297121634168987,
      "grad_norm": 1.4450682269997488,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8883757591247559,
      "num_tokens": 492596851.0,
      "step": 14161
    },
    {
      "epoch": 2.6298978644382545,
      "grad_norm": 1.6269725145106035,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8816994428634644,
      "num_tokens": 492631033.0,
      "step": 14162
    },
    {
      "epoch": 2.6300835654596098,
      "grad_norm": 1.6250718604155197,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8807119131088257,
      "num_tokens": 492664679.0,
      "step": 14163
    },
    {
      "epoch": 2.6302692664809655,
      "grad_norm": 1.6030489097934244,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8908185958862305,
      "num_tokens": 492696287.0,
      "step": 14164
    },
    {
      "epoch": 2.630454967502321,
      "grad_norm": 1.76065169739796,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.884583592414856,
      "num_tokens": 492725486.0,
      "step": 14165
    },
    {
      "epoch": 2.630640668523677,
      "grad_norm": 1.7234579087180224,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8870953321456909,
      "num_tokens": 492755977.0,
      "step": 14166
    },
    {
      "epoch": 2.6308263695450327,
      "grad_norm": 1.7490174747252734,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8715718388557434,
      "num_tokens": 492788641.0,
      "step": 14167
    },
    {
      "epoch": 2.631012070566388,
      "grad_norm": 1.6352532709386045,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8893136978149414,
      "num_tokens": 492820676.0,
      "step": 14168
    },
    {
      "epoch": 2.6311977715877437,
      "grad_norm": 1.5055849519863986,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8864718675613403,
      "num_tokens": 492856713.0,
      "step": 14169
    },
    {
      "epoch": 2.6313834726090994,
      "grad_norm": 1.6532085842742157,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8829339742660522,
      "num_tokens": 492892441.0,
      "step": 14170
    },
    {
      "epoch": 2.6315691736304547,
      "grad_norm": 1.5833814017958592,
      "learning_rate": 1e-06,
      "loss": 0.294,
      "mean_token_accuracy": 0.8949226140975952,
      "num_tokens": 492927745.0,
      "step": 14171
    },
    {
      "epoch": 2.6317548746518105,
      "grad_norm": 1.43667104698948,
      "learning_rate": 1e-06,
      "loss": 0.2792,
      "mean_token_accuracy": 0.9021956920623779,
      "num_tokens": 492962604.0,
      "step": 14172
    },
    {
      "epoch": 2.631940575673166,
      "grad_norm": 1.5566787085613263,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8914896845817566,
      "num_tokens": 492997535.0,
      "step": 14173
    },
    {
      "epoch": 2.632126276694522,
      "grad_norm": 1.6341791798091938,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.889488935470581,
      "num_tokens": 493028592.0,
      "step": 14174
    },
    {
      "epoch": 2.6323119777158777,
      "grad_norm": 1.5503459066985237,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8820238709449768,
      "num_tokens": 493064832.0,
      "step": 14175
    },
    {
      "epoch": 2.632497678737233,
      "grad_norm": 1.622716764274563,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.882583737373352,
      "num_tokens": 493098746.0,
      "step": 14176
    },
    {
      "epoch": 2.6326833797585887,
      "grad_norm": 1.5475103899667941,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8802844285964966,
      "num_tokens": 493133927.0,
      "step": 14177
    },
    {
      "epoch": 2.6328690807799444,
      "grad_norm": 1.5031797990154074,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8793199062347412,
      "num_tokens": 493173636.0,
      "step": 14178
    },
    {
      "epoch": 2.6330547818012997,
      "grad_norm": 1.3529415140749537,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8908514976501465,
      "num_tokens": 493221391.0,
      "step": 14179
    },
    {
      "epoch": 2.6332404828226554,
      "grad_norm": 1.5152289720118077,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8833224773406982,
      "num_tokens": 493259003.0,
      "step": 14180
    },
    {
      "epoch": 2.633426183844011,
      "grad_norm": 1.583146089675335,
      "learning_rate": 1e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.8664501905441284,
      "num_tokens": 493296301.0,
      "step": 14181
    },
    {
      "epoch": 2.633611884865367,
      "grad_norm": 1.6484959428879224,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.896536111831665,
      "num_tokens": 493324634.0,
      "step": 14182
    },
    {
      "epoch": 2.6337975858867226,
      "grad_norm": 1.5926715260981965,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8906593918800354,
      "num_tokens": 493358940.0,
      "step": 14183
    },
    {
      "epoch": 2.633983286908078,
      "grad_norm": 1.4485379575699138,
      "learning_rate": 1e-06,
      "loss": 0.2792,
      "mean_token_accuracy": 0.8998394012451172,
      "num_tokens": 493394971.0,
      "step": 14184
    },
    {
      "epoch": 2.6341689879294337,
      "grad_norm": 1.4603566406438364,
      "learning_rate": 1e-06,
      "loss": 0.2825,
      "mean_token_accuracy": 0.8986208438873291,
      "num_tokens": 493429704.0,
      "step": 14185
    },
    {
      "epoch": 2.6343546889507894,
      "grad_norm": 1.7063901051160826,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8772682547569275,
      "num_tokens": 493460990.0,
      "step": 14186
    },
    {
      "epoch": 2.6345403899721447,
      "grad_norm": 1.4978936359522028,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8896203637123108,
      "num_tokens": 493501644.0,
      "step": 14187
    },
    {
      "epoch": 2.6347260909935004,
      "grad_norm": 1.4531953819636805,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.884769082069397,
      "num_tokens": 493539976.0,
      "step": 14188
    },
    {
      "epoch": 2.634911792014856,
      "grad_norm": 1.5733028218321898,
      "learning_rate": 1e-06,
      "loss": 0.269,
      "mean_token_accuracy": 0.9050673842430115,
      "num_tokens": 493571572.0,
      "step": 14189
    },
    {
      "epoch": 2.635097493036212,
      "grad_norm": 1.5342379562972042,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8820179104804993,
      "num_tokens": 493608732.0,
      "step": 14190
    },
    {
      "epoch": 2.635283194057567,
      "grad_norm": 1.6583994979979457,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8781694173812866,
      "num_tokens": 493640015.0,
      "step": 14191
    },
    {
      "epoch": 2.635468895078923,
      "grad_norm": 1.7261282235222863,
      "learning_rate": 1e-06,
      "loss": 0.3396,
      "mean_token_accuracy": 0.8811769485473633,
      "num_tokens": 493669773.0,
      "step": 14192
    },
    {
      "epoch": 2.6356545961002786,
      "grad_norm": 1.4889733267576415,
      "learning_rate": 1e-06,
      "loss": 0.2871,
      "mean_token_accuracy": 0.8991310000419617,
      "num_tokens": 493707720.0,
      "step": 14193
    },
    {
      "epoch": 2.635840297121634,
      "grad_norm": 1.5242346353980263,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.892754316329956,
      "num_tokens": 493741714.0,
      "step": 14194
    },
    {
      "epoch": 2.6360259981429897,
      "grad_norm": 1.465090171767217,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8837398290634155,
      "num_tokens": 493778522.0,
      "step": 14195
    },
    {
      "epoch": 2.6362116991643454,
      "grad_norm": 1.829885157511721,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8741927146911621,
      "num_tokens": 493805005.0,
      "step": 14196
    },
    {
      "epoch": 2.636397400185701,
      "grad_norm": 1.5211811247668674,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8757438659667969,
      "num_tokens": 493841316.0,
      "step": 14197
    },
    {
      "epoch": 2.636583101207057,
      "grad_norm": 1.5076203128165022,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8884730339050293,
      "num_tokens": 493877783.0,
      "step": 14198
    },
    {
      "epoch": 2.636768802228412,
      "grad_norm": 1.542729365060978,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8732038736343384,
      "num_tokens": 493914088.0,
      "step": 14199
    },
    {
      "epoch": 2.636954503249768,
      "grad_norm": 1.5302840481226025,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8932479619979858,
      "num_tokens": 493950264.0,
      "step": 14200
    },
    {
      "epoch": 2.6371402042711236,
      "grad_norm": 1.5349947728421383,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.8840799331665039,
      "num_tokens": 493986759.0,
      "step": 14201
    },
    {
      "epoch": 2.637325905292479,
      "grad_norm": 1.4548107234170473,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.883607029914856,
      "num_tokens": 494024589.0,
      "step": 14202
    },
    {
      "epoch": 2.6375116063138346,
      "grad_norm": 1.4906799164255389,
      "learning_rate": 1e-06,
      "loss": 0.2849,
      "mean_token_accuracy": 0.8983892202377319,
      "num_tokens": 494058524.0,
      "step": 14203
    },
    {
      "epoch": 2.6376973073351904,
      "grad_norm": 1.6399131517473777,
      "learning_rate": 1e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.862321138381958,
      "num_tokens": 494095766.0,
      "step": 14204
    },
    {
      "epoch": 2.637883008356546,
      "grad_norm": 1.5612387206536789,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8859754204750061,
      "num_tokens": 494128935.0,
      "step": 14205
    },
    {
      "epoch": 2.638068709377902,
      "grad_norm": 1.6326346128568265,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8820515275001526,
      "num_tokens": 494160641.0,
      "step": 14206
    },
    {
      "epoch": 2.638254410399257,
      "grad_norm": 1.5614406565513876,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8952593803405762,
      "num_tokens": 494190714.0,
      "step": 14207
    },
    {
      "epoch": 2.638440111420613,
      "grad_norm": 1.6173983120370545,
      "learning_rate": 1e-06,
      "loss": 0.2845,
      "mean_token_accuracy": 0.9039673805236816,
      "num_tokens": 494220884.0,
      "step": 14208
    },
    {
      "epoch": 2.6386258124419686,
      "grad_norm": 1.4770958315376583,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8808466792106628,
      "num_tokens": 494262268.0,
      "step": 14209
    },
    {
      "epoch": 2.638811513463324,
      "grad_norm": 1.475457373025467,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8833231329917908,
      "num_tokens": 494307718.0,
      "step": 14210
    },
    {
      "epoch": 2.6389972144846796,
      "grad_norm": 1.597143585435716,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8837581872940063,
      "num_tokens": 494343931.0,
      "step": 14211
    },
    {
      "epoch": 2.6391829155060353,
      "grad_norm": 1.5887954145396614,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8805835247039795,
      "num_tokens": 494383298.0,
      "step": 14212
    },
    {
      "epoch": 2.639368616527391,
      "grad_norm": 1.7048652957341397,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8932687044143677,
      "num_tokens": 494417661.0,
      "step": 14213
    },
    {
      "epoch": 2.6395543175487464,
      "grad_norm": 1.4988459274448238,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.8976924419403076,
      "num_tokens": 494454423.0,
      "step": 14214
    },
    {
      "epoch": 2.639740018570102,
      "grad_norm": 1.5920516941274643,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8857021927833557,
      "num_tokens": 494484466.0,
      "step": 14215
    },
    {
      "epoch": 2.639925719591458,
      "grad_norm": 1.5062329931177605,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8861784934997559,
      "num_tokens": 494522956.0,
      "step": 14216
    },
    {
      "epoch": 2.640111420612813,
      "grad_norm": 1.5909069894191026,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8853168487548828,
      "num_tokens": 494559061.0,
      "step": 14217
    },
    {
      "epoch": 2.640297121634169,
      "grad_norm": 1.6271644018634184,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8936803936958313,
      "num_tokens": 494593670.0,
      "step": 14218
    },
    {
      "epoch": 2.6404828226555246,
      "grad_norm": 1.6892328610801153,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.872795045375824,
      "num_tokens": 494626905.0,
      "step": 14219
    },
    {
      "epoch": 2.6406685236768803,
      "grad_norm": 1.5004487898449335,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8797037601470947,
      "num_tokens": 494666059.0,
      "step": 14220
    },
    {
      "epoch": 2.640854224698236,
      "grad_norm": 1.4626030012860372,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8895012140274048,
      "num_tokens": 494706763.0,
      "step": 14221
    },
    {
      "epoch": 2.6410399257195913,
      "grad_norm": 1.4839429178406491,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8843721151351929,
      "num_tokens": 494743052.0,
      "step": 14222
    },
    {
      "epoch": 2.641225626740947,
      "grad_norm": 1.5656288473127755,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.891880989074707,
      "num_tokens": 494781199.0,
      "step": 14223
    },
    {
      "epoch": 2.641411327762303,
      "grad_norm": 1.721273148684055,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8904035091400146,
      "num_tokens": 494813118.0,
      "step": 14224
    },
    {
      "epoch": 2.641597028783658,
      "grad_norm": 1.5756280088642705,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8905256986618042,
      "num_tokens": 494848786.0,
      "step": 14225
    },
    {
      "epoch": 2.641782729805014,
      "grad_norm": 1.576000381878292,
      "learning_rate": 1e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8689218759536743,
      "num_tokens": 494892951.0,
      "step": 14226
    },
    {
      "epoch": 2.6419684308263696,
      "grad_norm": 1.626885115784697,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8920651078224182,
      "num_tokens": 494924981.0,
      "step": 14227
    },
    {
      "epoch": 2.6421541318477253,
      "grad_norm": 1.577729967579332,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8995969891548157,
      "num_tokens": 494958879.0,
      "step": 14228
    },
    {
      "epoch": 2.642339832869081,
      "grad_norm": 1.535057518454924,
      "learning_rate": 1e-06,
      "loss": 0.2937,
      "mean_token_accuracy": 0.895431637763977,
      "num_tokens": 494992251.0,
      "step": 14229
    },
    {
      "epoch": 2.6425255338904363,
      "grad_norm": 1.5308508656606419,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8880752921104431,
      "num_tokens": 495030053.0,
      "step": 14230
    },
    {
      "epoch": 2.642711234911792,
      "grad_norm": 1.5747312384377334,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8871477842330933,
      "num_tokens": 495064758.0,
      "step": 14231
    },
    {
      "epoch": 2.642896935933148,
      "grad_norm": 1.6488272495301646,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.8783243894577026,
      "num_tokens": 495095276.0,
      "step": 14232
    },
    {
      "epoch": 2.643082636954503,
      "grad_norm": 1.6613876096591849,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8887988924980164,
      "num_tokens": 495130112.0,
      "step": 14233
    },
    {
      "epoch": 2.643268337975859,
      "grad_norm": 1.5242092558482185,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8817439675331116,
      "num_tokens": 495166483.0,
      "step": 14234
    },
    {
      "epoch": 2.6434540389972145,
      "grad_norm": 1.6192823155539007,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8889504671096802,
      "num_tokens": 495199201.0,
      "step": 14235
    },
    {
      "epoch": 2.6436397400185703,
      "grad_norm": 1.6515817670762065,
      "learning_rate": 1e-06,
      "loss": 0.2805,
      "mean_token_accuracy": 0.8984094858169556,
      "num_tokens": 495228673.0,
      "step": 14236
    },
    {
      "epoch": 2.6438254410399256,
      "grad_norm": 1.5639254161661662,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8832041025161743,
      "num_tokens": 495262527.0,
      "step": 14237
    },
    {
      "epoch": 2.6440111420612813,
      "grad_norm": 2.2670693239220703,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.885075569152832,
      "num_tokens": 495297407.0,
      "step": 14238
    },
    {
      "epoch": 2.644196843082637,
      "grad_norm": 1.6004684853726732,
      "learning_rate": 1e-06,
      "loss": 0.3091,
      "mean_token_accuracy": 0.889966607093811,
      "num_tokens": 495332794.0,
      "step": 14239
    },
    {
      "epoch": 2.6443825441039923,
      "grad_norm": 1.5325336736632684,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8895708322525024,
      "num_tokens": 495371450.0,
      "step": 14240
    },
    {
      "epoch": 2.644568245125348,
      "grad_norm": 1.5744766098818008,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8948056697845459,
      "num_tokens": 495405310.0,
      "step": 14241
    },
    {
      "epoch": 2.644753946146704,
      "grad_norm": 1.453533250373347,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8747318983078003,
      "num_tokens": 495446303.0,
      "step": 14242
    },
    {
      "epoch": 2.6449396471680595,
      "grad_norm": 1.4750402153742788,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8853750824928284,
      "num_tokens": 495487042.0,
      "step": 14243
    },
    {
      "epoch": 2.6451253481894152,
      "grad_norm": 1.5708342838539309,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8933517932891846,
      "num_tokens": 495520874.0,
      "step": 14244
    },
    {
      "epoch": 2.6453110492107705,
      "grad_norm": 1.5060028458475714,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8779277801513672,
      "num_tokens": 495557237.0,
      "step": 14245
    },
    {
      "epoch": 2.6454967502321263,
      "grad_norm": 1.7097713435068003,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8783975839614868,
      "num_tokens": 495591030.0,
      "step": 14246
    },
    {
      "epoch": 2.645682451253482,
      "grad_norm": 1.645630132178611,
      "learning_rate": 1e-06,
      "loss": 0.2995,
      "mean_token_accuracy": 0.89293372631073,
      "num_tokens": 495619619.0,
      "step": 14247
    },
    {
      "epoch": 2.6458681522748373,
      "grad_norm": 1.4567808845482142,
      "learning_rate": 1e-06,
      "loss": 0.2727,
      "mean_token_accuracy": 0.9026139378547668,
      "num_tokens": 495654671.0,
      "step": 14248
    },
    {
      "epoch": 2.646053853296193,
      "grad_norm": 1.6413796691830083,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8949939012527466,
      "num_tokens": 495689856.0,
      "step": 14249
    },
    {
      "epoch": 2.6462395543175488,
      "grad_norm": 1.563060650640058,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.8768973350524902,
      "num_tokens": 495725308.0,
      "step": 14250
    },
    {
      "epoch": 2.6464252553389045,
      "grad_norm": 1.6722042665404486,
      "learning_rate": 1e-06,
      "loss": 0.3509,
      "mean_token_accuracy": 0.8754359483718872,
      "num_tokens": 495759559.0,
      "step": 14251
    },
    {
      "epoch": 2.6466109563602602,
      "grad_norm": 1.513024898962001,
      "learning_rate": 1e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8749536275863647,
      "num_tokens": 495797151.0,
      "step": 14252
    },
    {
      "epoch": 2.6467966573816155,
      "grad_norm": 1.4798254668693227,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.892546534538269,
      "num_tokens": 495832832.0,
      "step": 14253
    },
    {
      "epoch": 2.6469823584029712,
      "grad_norm": 1.4936615932211035,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8894498348236084,
      "num_tokens": 495868279.0,
      "step": 14254
    },
    {
      "epoch": 2.647168059424327,
      "grad_norm": 1.7036634751003945,
      "learning_rate": 1e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8718355894088745,
      "num_tokens": 495899863.0,
      "step": 14255
    },
    {
      "epoch": 2.6473537604456823,
      "grad_norm": 1.7451144601218018,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8797668218612671,
      "num_tokens": 495928378.0,
      "step": 14256
    },
    {
      "epoch": 2.647539461467038,
      "grad_norm": 1.5302557784874347,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8859769105911255,
      "num_tokens": 495962552.0,
      "step": 14257
    },
    {
      "epoch": 2.6477251624883937,
      "grad_norm": 1.686904541455152,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8788367509841919,
      "num_tokens": 495996657.0,
      "step": 14258
    },
    {
      "epoch": 2.6479108635097495,
      "grad_norm": 1.5209095590875628,
      "learning_rate": 1e-06,
      "loss": 0.3048,
      "mean_token_accuracy": 0.8912632465362549,
      "num_tokens": 496035762.0,
      "step": 14259
    },
    {
      "epoch": 2.6480965645311048,
      "grad_norm": 1.575875337584889,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8797829747200012,
      "num_tokens": 496071682.0,
      "step": 14260
    },
    {
      "epoch": 2.6482822655524605,
      "grad_norm": 1.6779807026575504,
      "learning_rate": 1e-06,
      "loss": 0.3009,
      "mean_token_accuracy": 0.894608199596405,
      "num_tokens": 496101960.0,
      "step": 14261
    },
    {
      "epoch": 2.6484679665738162,
      "grad_norm": 1.4822159471190943,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8748735189437866,
      "num_tokens": 496140552.0,
      "step": 14262
    },
    {
      "epoch": 2.6486536675951715,
      "grad_norm": 1.3946299371170323,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.885729193687439,
      "num_tokens": 496178469.0,
      "step": 14263
    },
    {
      "epoch": 2.6488393686165272,
      "grad_norm": 1.602183952812612,
      "learning_rate": 1e-06,
      "loss": 0.377,
      "mean_token_accuracy": 0.8703548908233643,
      "num_tokens": 496214824.0,
      "step": 14264
    },
    {
      "epoch": 2.649025069637883,
      "grad_norm": 1.5714638325051682,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8831171989440918,
      "num_tokens": 496253609.0,
      "step": 14265
    },
    {
      "epoch": 2.6492107706592387,
      "grad_norm": 1.5385803047794357,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.895858883857727,
      "num_tokens": 496286580.0,
      "step": 14266
    },
    {
      "epoch": 2.6493964716805944,
      "grad_norm": 1.6784402610633447,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8824763298034668,
      "num_tokens": 496319099.0,
      "step": 14267
    },
    {
      "epoch": 2.6495821727019497,
      "grad_norm": 1.4957711474253081,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8823675513267517,
      "num_tokens": 496358338.0,
      "step": 14268
    },
    {
      "epoch": 2.6497678737233055,
      "grad_norm": 1.6552666906092073,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8872519731521606,
      "num_tokens": 496393186.0,
      "step": 14269
    },
    {
      "epoch": 2.649953574744661,
      "grad_norm": 1.6135979792425335,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8727227449417114,
      "num_tokens": 496425456.0,
      "step": 14270
    },
    {
      "epoch": 2.6501392757660165,
      "grad_norm": 1.4518905021213875,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8930284380912781,
      "num_tokens": 496462424.0,
      "step": 14271
    },
    {
      "epoch": 2.650324976787372,
      "grad_norm": 1.7513337785695646,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8699816465377808,
      "num_tokens": 496492979.0,
      "step": 14272
    },
    {
      "epoch": 2.650510677808728,
      "grad_norm": 1.8333593290259436,
      "learning_rate": 1e-06,
      "loss": 0.3641,
      "mean_token_accuracy": 0.8741596937179565,
      "num_tokens": 496519523.0,
      "step": 14273
    },
    {
      "epoch": 2.6506963788300837,
      "grad_norm": 1.5750892138192047,
      "learning_rate": 1e-06,
      "loss": 0.2729,
      "mean_token_accuracy": 0.9020276069641113,
      "num_tokens": 496548345.0,
      "step": 14274
    },
    {
      "epoch": 2.6508820798514394,
      "grad_norm": 1.4064600352109804,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8948822021484375,
      "num_tokens": 496592287.0,
      "step": 14275
    },
    {
      "epoch": 2.6510677808727947,
      "grad_norm": 1.5660855752110152,
      "learning_rate": 1e-06,
      "loss": 0.3513,
      "mean_token_accuracy": 0.8751277923583984,
      "num_tokens": 496628484.0,
      "step": 14276
    },
    {
      "epoch": 2.6512534818941504,
      "grad_norm": 1.5336372538134027,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8892116546630859,
      "num_tokens": 496660758.0,
      "step": 14277
    },
    {
      "epoch": 2.651439182915506,
      "grad_norm": 1.5212881649888632,
      "learning_rate": 1e-06,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9163118600845337,
      "num_tokens": 496690204.0,
      "step": 14278
    },
    {
      "epoch": 2.6516248839368615,
      "grad_norm": 1.5480571009104278,
      "learning_rate": 1e-06,
      "loss": 0.2984,
      "mean_token_accuracy": 0.8935500383377075,
      "num_tokens": 496721868.0,
      "step": 14279
    },
    {
      "epoch": 2.651810584958217,
      "grad_norm": 1.52524465767362,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8841274380683899,
      "num_tokens": 496759074.0,
      "step": 14280
    },
    {
      "epoch": 2.651996285979573,
      "grad_norm": 1.5046380368102426,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8822256326675415,
      "num_tokens": 496798495.0,
      "step": 14281
    },
    {
      "epoch": 2.6521819870009287,
      "grad_norm": 1.49439044407227,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8835899829864502,
      "num_tokens": 496838129.0,
      "step": 14282
    },
    {
      "epoch": 2.6523676880222844,
      "grad_norm": 1.552045614663662,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8856565356254578,
      "num_tokens": 496875122.0,
      "step": 14283
    },
    {
      "epoch": 2.6525533890436397,
      "grad_norm": 1.5351085144976282,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8915623426437378,
      "num_tokens": 496910666.0,
      "step": 14284
    },
    {
      "epoch": 2.6527390900649954,
      "grad_norm": 1.5139435031563484,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8783079385757446,
      "num_tokens": 496951517.0,
      "step": 14285
    },
    {
      "epoch": 2.6529247910863507,
      "grad_norm": 1.5508340933121345,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.881446897983551,
      "num_tokens": 496986933.0,
      "step": 14286
    },
    {
      "epoch": 2.6531104921077064,
      "grad_norm": 1.6278344241325067,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8856481909751892,
      "num_tokens": 497022385.0,
      "step": 14287
    },
    {
      "epoch": 2.653296193129062,
      "grad_norm": 1.7054098730775364,
      "learning_rate": 1e-06,
      "loss": 0.3622,
      "mean_token_accuracy": 0.8750263452529907,
      "num_tokens": 497054801.0,
      "step": 14288
    },
    {
      "epoch": 2.653481894150418,
      "grad_norm": 1.6604019438982178,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8897786140441895,
      "num_tokens": 497089302.0,
      "step": 14289
    },
    {
      "epoch": 2.6536675951717736,
      "grad_norm": 1.6466938863783331,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8855459690093994,
      "num_tokens": 497121897.0,
      "step": 14290
    },
    {
      "epoch": 2.653853296193129,
      "grad_norm": 1.5442531625422,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8867796659469604,
      "num_tokens": 497157097.0,
      "step": 14291
    },
    {
      "epoch": 2.6540389972144847,
      "grad_norm": 1.6570902575307491,
      "learning_rate": 1e-06,
      "loss": 0.3272,
      "mean_token_accuracy": 0.8848307728767395,
      "num_tokens": 497189329.0,
      "step": 14292
    },
    {
      "epoch": 2.6542246982358404,
      "grad_norm": 1.4266680575450066,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8891716599464417,
      "num_tokens": 497226743.0,
      "step": 14293
    },
    {
      "epoch": 2.6544103992571957,
      "grad_norm": 1.488495809314343,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8759269714355469,
      "num_tokens": 497264764.0,
      "step": 14294
    },
    {
      "epoch": 2.6545961002785514,
      "grad_norm": 1.7083698402393255,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8795679807662964,
      "num_tokens": 497296495.0,
      "step": 14295
    },
    {
      "epoch": 2.654781801299907,
      "grad_norm": 1.6204244566693027,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8833067417144775,
      "num_tokens": 497331057.0,
      "step": 14296
    },
    {
      "epoch": 2.654967502321263,
      "grad_norm": 1.5912838590864427,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.8938829898834229,
      "num_tokens": 497365110.0,
      "step": 14297
    },
    {
      "epoch": 2.6551532033426186,
      "grad_norm": 1.669845790019575,
      "learning_rate": 1e-06,
      "loss": 0.2832,
      "mean_token_accuracy": 0.9010371565818787,
      "num_tokens": 497395822.0,
      "step": 14298
    },
    {
      "epoch": 2.655338904363974,
      "grad_norm": 1.5472842747743154,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8803995847702026,
      "num_tokens": 497434270.0,
      "step": 14299
    },
    {
      "epoch": 2.6555246053853296,
      "grad_norm": 1.4515254859231979,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8887361288070679,
      "num_tokens": 497473127.0,
      "step": 14300
    },
    {
      "epoch": 2.6557103064066854,
      "grad_norm": 1.632961060470131,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8775737285614014,
      "num_tokens": 497504076.0,
      "step": 14301
    },
    {
      "epoch": 2.6558960074280407,
      "grad_norm": 1.6732198525636341,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8863220810890198,
      "num_tokens": 497532692.0,
      "step": 14302
    },
    {
      "epoch": 2.6560817084493964,
      "grad_norm": 1.879080844640428,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8861361742019653,
      "num_tokens": 497559642.0,
      "step": 14303
    },
    {
      "epoch": 2.656267409470752,
      "grad_norm": 1.642253805430725,
      "learning_rate": 1e-06,
      "loss": 0.2969,
      "mean_token_accuracy": 0.8937199115753174,
      "num_tokens": 497586519.0,
      "step": 14304
    },
    {
      "epoch": 2.656453110492108,
      "grad_norm": 1.756751815418734,
      "learning_rate": 1e-06,
      "loss": 0.3196,
      "mean_token_accuracy": 0.8861593008041382,
      "num_tokens": 497615029.0,
      "step": 14305
    },
    {
      "epoch": 2.6566388115134636,
      "grad_norm": 1.5561603598953633,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8938348889350891,
      "num_tokens": 497648574.0,
      "step": 14306
    },
    {
      "epoch": 2.656824512534819,
      "grad_norm": 1.549527034667061,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.876812756061554,
      "num_tokens": 497687480.0,
      "step": 14307
    },
    {
      "epoch": 2.6570102135561746,
      "grad_norm": 1.5549730674971705,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8862252831459045,
      "num_tokens": 497720679.0,
      "step": 14308
    },
    {
      "epoch": 2.65719591457753,
      "grad_norm": 1.59519453612085,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8815764784812927,
      "num_tokens": 497753936.0,
      "step": 14309
    },
    {
      "epoch": 2.6573816155988856,
      "grad_norm": 1.643282206260235,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.890310525894165,
      "num_tokens": 497787085.0,
      "step": 14310
    },
    {
      "epoch": 2.6575673166202414,
      "grad_norm": 1.6402959420307255,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8913732171058655,
      "num_tokens": 497818006.0,
      "step": 14311
    },
    {
      "epoch": 2.657753017641597,
      "grad_norm": 1.5248945716875175,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8834731578826904,
      "num_tokens": 497855507.0,
      "step": 14312
    },
    {
      "epoch": 2.657938718662953,
      "grad_norm": 1.372535864140086,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8947165608406067,
      "num_tokens": 497897767.0,
      "step": 14313
    },
    {
      "epoch": 2.658124419684308,
      "grad_norm": 1.6266356640831017,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.8834287524223328,
      "num_tokens": 497932715.0,
      "step": 14314
    },
    {
      "epoch": 2.658310120705664,
      "grad_norm": 1.6969151707945893,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8842650055885315,
      "num_tokens": 497963780.0,
      "step": 14315
    },
    {
      "epoch": 2.6584958217270196,
      "grad_norm": 1.5122245430812846,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8804753422737122,
      "num_tokens": 498005207.0,
      "step": 14316
    },
    {
      "epoch": 2.658681522748375,
      "grad_norm": 1.5503079632987817,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8895093202590942,
      "num_tokens": 498042509.0,
      "step": 14317
    },
    {
      "epoch": 2.6588672237697306,
      "grad_norm": 1.5746583297557053,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.878083348274231,
      "num_tokens": 498078877.0,
      "step": 14318
    },
    {
      "epoch": 2.6590529247910863,
      "grad_norm": 1.5810161299949428,
      "learning_rate": 1e-06,
      "loss": 0.2815,
      "mean_token_accuracy": 0.9004641175270081,
      "num_tokens": 498114777.0,
      "step": 14319
    },
    {
      "epoch": 2.659238625812442,
      "grad_norm": 1.4897467975571126,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8845232725143433,
      "num_tokens": 498154240.0,
      "step": 14320
    },
    {
      "epoch": 2.659424326833798,
      "grad_norm": 1.8994387344084789,
      "learning_rate": 1e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8641862869262695,
      "num_tokens": 498184349.0,
      "step": 14321
    },
    {
      "epoch": 2.659610027855153,
      "grad_norm": 1.6811742794330824,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8756269216537476,
      "num_tokens": 498220891.0,
      "step": 14322
    },
    {
      "epoch": 2.659795728876509,
      "grad_norm": 1.693028059145795,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.8829403519630432,
      "num_tokens": 498251818.0,
      "step": 14323
    },
    {
      "epoch": 2.6599814298978646,
      "grad_norm": 1.5461177569792286,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8883829116821289,
      "num_tokens": 498290229.0,
      "step": 14324
    },
    {
      "epoch": 2.66016713091922,
      "grad_norm": 1.5399674718318912,
      "learning_rate": 1e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.8791422843933105,
      "num_tokens": 498328299.0,
      "step": 14325
    },
    {
      "epoch": 2.6603528319405756,
      "grad_norm": 1.4922818917045149,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8885802626609802,
      "num_tokens": 498364427.0,
      "step": 14326
    },
    {
      "epoch": 2.6605385329619313,
      "grad_norm": 1.4651333924439554,
      "learning_rate": 1e-06,
      "loss": 0.2849,
      "mean_token_accuracy": 0.8969133496284485,
      "num_tokens": 498399619.0,
      "step": 14327
    },
    {
      "epoch": 2.660724233983287,
      "grad_norm": 1.5403269100633779,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8848810195922852,
      "num_tokens": 498435795.0,
      "step": 14328
    },
    {
      "epoch": 2.660909935004643,
      "grad_norm": 1.5811201087805096,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8772315979003906,
      "num_tokens": 498470997.0,
      "step": 14329
    },
    {
      "epoch": 2.661095636025998,
      "grad_norm": 1.577558768348074,
      "learning_rate": 1e-06,
      "loss": 0.2869,
      "mean_token_accuracy": 0.8937398195266724,
      "num_tokens": 498501096.0,
      "step": 14330
    },
    {
      "epoch": 2.661281337047354,
      "grad_norm": 1.4630840592992982,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8896461129188538,
      "num_tokens": 498538726.0,
      "step": 14331
    },
    {
      "epoch": 2.661467038068709,
      "grad_norm": 1.521416495431197,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8921221494674683,
      "num_tokens": 498572535.0,
      "step": 14332
    },
    {
      "epoch": 2.661652739090065,
      "grad_norm": 1.5024283488709231,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8817495703697205,
      "num_tokens": 498610973.0,
      "step": 14333
    },
    {
      "epoch": 2.6618384401114206,
      "grad_norm": 1.7008642784133527,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8922438621520996,
      "num_tokens": 498640299.0,
      "step": 14334
    },
    {
      "epoch": 2.6620241411327763,
      "grad_norm": 1.5025669554523178,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.887558102607727,
      "num_tokens": 498680114.0,
      "step": 14335
    },
    {
      "epoch": 2.662209842154132,
      "grad_norm": 1.8222811675900388,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.882554829120636,
      "num_tokens": 498713158.0,
      "step": 14336
    },
    {
      "epoch": 2.6623955431754873,
      "grad_norm": 1.6657890581284407,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8859929442405701,
      "num_tokens": 498743550.0,
      "step": 14337
    },
    {
      "epoch": 2.662581244196843,
      "grad_norm": 1.575790315415321,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8825204968452454,
      "num_tokens": 498781151.0,
      "step": 14338
    },
    {
      "epoch": 2.662766945218199,
      "grad_norm": 1.5942446241837196,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8841552138328552,
      "num_tokens": 498814973.0,
      "step": 14339
    },
    {
      "epoch": 2.662952646239554,
      "grad_norm": 1.529514332239165,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8866187334060669,
      "num_tokens": 498851287.0,
      "step": 14340
    },
    {
      "epoch": 2.66313834726091,
      "grad_norm": 1.4968498152172467,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8908075094223022,
      "num_tokens": 498889021.0,
      "step": 14341
    },
    {
      "epoch": 2.6633240482822655,
      "grad_norm": 1.627053943744661,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8756220936775208,
      "num_tokens": 498921892.0,
      "step": 14342
    },
    {
      "epoch": 2.6635097493036213,
      "grad_norm": 1.5290532246584567,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8871268630027771,
      "num_tokens": 498958683.0,
      "step": 14343
    },
    {
      "epoch": 2.663695450324977,
      "grad_norm": 1.3938960223211463,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8930889368057251,
      "num_tokens": 499001922.0,
      "step": 14344
    },
    {
      "epoch": 2.6638811513463323,
      "grad_norm": 1.5069212573261626,
      "learning_rate": 1e-06,
      "loss": 0.2739,
      "mean_token_accuracy": 0.9019746780395508,
      "num_tokens": 499035604.0,
      "step": 14345
    },
    {
      "epoch": 2.664066852367688,
      "grad_norm": 1.4310762003259057,
      "learning_rate": 1e-06,
      "loss": 0.3002,
      "mean_token_accuracy": 0.8957477807998657,
      "num_tokens": 499075833.0,
      "step": 14346
    },
    {
      "epoch": 2.6642525533890438,
      "grad_norm": 1.5737263875827683,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8863220810890198,
      "num_tokens": 499112916.0,
      "step": 14347
    },
    {
      "epoch": 2.664438254410399,
      "grad_norm": 1.4949949247604732,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8860340118408203,
      "num_tokens": 499150577.0,
      "step": 14348
    },
    {
      "epoch": 2.664623955431755,
      "grad_norm": 1.5559232148172313,
      "learning_rate": 1e-06,
      "loss": 0.3004,
      "mean_token_accuracy": 0.8918725848197937,
      "num_tokens": 499179872.0,
      "step": 14349
    },
    {
      "epoch": 2.6648096564531105,
      "grad_norm": 1.3900853471269945,
      "learning_rate": 1e-06,
      "loss": 0.277,
      "mean_token_accuracy": 0.8995869159698486,
      "num_tokens": 499216642.0,
      "step": 14350
    },
    {
      "epoch": 2.6649953574744663,
      "grad_norm": 1.5876277585708956,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8904247283935547,
      "num_tokens": 499253693.0,
      "step": 14351
    },
    {
      "epoch": 2.665181058495822,
      "grad_norm": 1.6826736180587718,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8872958421707153,
      "num_tokens": 499289230.0,
      "step": 14352
    },
    {
      "epoch": 2.6653667595171773,
      "grad_norm": 1.5715660553372084,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8928865194320679,
      "num_tokens": 499322695.0,
      "step": 14353
    },
    {
      "epoch": 2.665552460538533,
      "grad_norm": 1.604306595751663,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8884242177009583,
      "num_tokens": 499356135.0,
      "step": 14354
    },
    {
      "epoch": 2.6657381615598887,
      "grad_norm": 1.5813904362519422,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8820832967758179,
      "num_tokens": 499392294.0,
      "step": 14355
    },
    {
      "epoch": 2.665923862581244,
      "grad_norm": 1.5312860102578452,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8901538252830505,
      "num_tokens": 499425396.0,
      "step": 14356
    },
    {
      "epoch": 2.6661095636025998,
      "grad_norm": 1.7601325086988984,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8789317607879639,
      "num_tokens": 499455498.0,
      "step": 14357
    },
    {
      "epoch": 2.6662952646239555,
      "grad_norm": 1.5993650394583097,
      "learning_rate": 1e-06,
      "loss": 0.3177,
      "mean_token_accuracy": 0.8887848854064941,
      "num_tokens": 499489393.0,
      "step": 14358
    },
    {
      "epoch": 2.6664809656453112,
      "grad_norm": 1.4327158220484055,
      "learning_rate": 1e-06,
      "loss": 0.2856,
      "mean_token_accuracy": 0.8989695310592651,
      "num_tokens": 499528297.0,
      "step": 14359
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 1.4513044966333,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8904472589492798,
      "num_tokens": 499566345.0,
      "step": 14360
    },
    {
      "epoch": 2.6668523676880223,
      "grad_norm": 1.5963877772658623,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8759621381759644,
      "num_tokens": 499600218.0,
      "step": 14361
    },
    {
      "epoch": 2.667038068709378,
      "grad_norm": 1.5504279815572115,
      "learning_rate": 1e-06,
      "loss": 0.295,
      "mean_token_accuracy": 0.8937029838562012,
      "num_tokens": 499635285.0,
      "step": 14362
    },
    {
      "epoch": 2.6672237697307333,
      "grad_norm": 1.7485927120482074,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8832131624221802,
      "num_tokens": 499665740.0,
      "step": 14363
    },
    {
      "epoch": 2.667409470752089,
      "grad_norm": 1.7535743186026533,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8829248547554016,
      "num_tokens": 499693837.0,
      "step": 14364
    },
    {
      "epoch": 2.6675951717734447,
      "grad_norm": 1.511765910159016,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8878464102745056,
      "num_tokens": 499733567.0,
      "step": 14365
    },
    {
      "epoch": 2.6677808727948005,
      "grad_norm": 1.5439285724170542,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.8876715898513794,
      "num_tokens": 499772962.0,
      "step": 14366
    },
    {
      "epoch": 2.667966573816156,
      "grad_norm": 1.7528729348667378,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8942908048629761,
      "num_tokens": 499801319.0,
      "step": 14367
    },
    {
      "epoch": 2.6681522748375115,
      "grad_norm": 1.5001886968378269,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8896893262863159,
      "num_tokens": 499836495.0,
      "step": 14368
    },
    {
      "epoch": 2.6683379758588672,
      "grad_norm": 1.6093169497435789,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8919879794120789,
      "num_tokens": 499867451.0,
      "step": 14369
    },
    {
      "epoch": 2.668523676880223,
      "grad_norm": 1.5294676962078935,
      "learning_rate": 1e-06,
      "loss": 0.3138,
      "mean_token_accuracy": 0.8866966366767883,
      "num_tokens": 499901617.0,
      "step": 14370
    },
    {
      "epoch": 2.6687093779015783,
      "grad_norm": 1.5957774014526047,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8741046190261841,
      "num_tokens": 499939130.0,
      "step": 14371
    },
    {
      "epoch": 2.668895078922934,
      "grad_norm": 1.576754182776235,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.886320948600769,
      "num_tokens": 499974825.0,
      "step": 14372
    },
    {
      "epoch": 2.6690807799442897,
      "grad_norm": 1.5913854542937398,
      "learning_rate": 1e-06,
      "loss": 0.3042,
      "mean_token_accuracy": 0.8908545970916748,
      "num_tokens": 500008421.0,
      "step": 14373
    },
    {
      "epoch": 2.6692664809656454,
      "grad_norm": 1.6405611822346189,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8776171803474426,
      "num_tokens": 500040005.0,
      "step": 14374
    },
    {
      "epoch": 2.669452181987001,
      "grad_norm": 1.6156221845140912,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8825197219848633,
      "num_tokens": 500073670.0,
      "step": 14375
    },
    {
      "epoch": 2.6696378830083565,
      "grad_norm": 1.6045306805693251,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8877238035202026,
      "num_tokens": 500103990.0,
      "step": 14376
    },
    {
      "epoch": 2.669823584029712,
      "grad_norm": 1.5854906206980044,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8816449642181396,
      "num_tokens": 500136321.0,
      "step": 14377
    },
    {
      "epoch": 2.670009285051068,
      "grad_norm": 1.6971009544809788,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8718358874320984,
      "num_tokens": 500173166.0,
      "step": 14378
    },
    {
      "epoch": 2.6701949860724232,
      "grad_norm": 1.4779683969128081,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8850072622299194,
      "num_tokens": 500212477.0,
      "step": 14379
    },
    {
      "epoch": 2.670380687093779,
      "grad_norm": 1.494260677366502,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8935269117355347,
      "num_tokens": 500251265.0,
      "step": 14380
    },
    {
      "epoch": 2.6705663881151347,
      "grad_norm": 1.6529914257167524,
      "learning_rate": 1e-06,
      "loss": 0.2765,
      "mean_token_accuracy": 0.9036422371864319,
      "num_tokens": 500281881.0,
      "step": 14381
    },
    {
      "epoch": 2.6707520891364904,
      "grad_norm": 1.634813320183168,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8877475261688232,
      "num_tokens": 500312267.0,
      "step": 14382
    },
    {
      "epoch": 2.6709377901578457,
      "grad_norm": 1.621138208703903,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8811441659927368,
      "num_tokens": 500347522.0,
      "step": 14383
    },
    {
      "epoch": 2.6711234911792014,
      "grad_norm": 1.7023459134744183,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8806253671646118,
      "num_tokens": 500379689.0,
      "step": 14384
    },
    {
      "epoch": 2.671309192200557,
      "grad_norm": 1.4780100146897959,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8927930593490601,
      "num_tokens": 500414928.0,
      "step": 14385
    },
    {
      "epoch": 2.6714948932219125,
      "grad_norm": 1.5296954378551568,
      "learning_rate": 1e-06,
      "loss": 0.2692,
      "mean_token_accuracy": 0.9030993580818176,
      "num_tokens": 500448542.0,
      "step": 14386
    },
    {
      "epoch": 2.671680594243268,
      "grad_norm": 1.6941353138113715,
      "learning_rate": 1e-06,
      "loss": 0.339,
      "mean_token_accuracy": 0.8822150230407715,
      "num_tokens": 500478833.0,
      "step": 14387
    },
    {
      "epoch": 2.671866295264624,
      "grad_norm": 1.5680283540601712,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8890186548233032,
      "num_tokens": 500513072.0,
      "step": 14388
    },
    {
      "epoch": 2.6720519962859797,
      "grad_norm": 1.5020825727277938,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8899641036987305,
      "num_tokens": 500551703.0,
      "step": 14389
    },
    {
      "epoch": 2.6722376973073354,
      "grad_norm": 1.6217415197919618,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.8835185766220093,
      "num_tokens": 500585083.0,
      "step": 14390
    },
    {
      "epoch": 2.6724233983286907,
      "grad_norm": 1.5597626527056798,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8840073347091675,
      "num_tokens": 500620646.0,
      "step": 14391
    },
    {
      "epoch": 2.6726090993500464,
      "grad_norm": 1.6029967781377958,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8898126482963562,
      "num_tokens": 500653332.0,
      "step": 14392
    },
    {
      "epoch": 2.672794800371402,
      "grad_norm": 1.590686327058266,
      "learning_rate": 1e-06,
      "loss": 0.2678,
      "mean_token_accuracy": 0.903755784034729,
      "num_tokens": 500682935.0,
      "step": 14393
    },
    {
      "epoch": 2.6729805013927574,
      "grad_norm": 1.5065062125208546,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8843531608581543,
      "num_tokens": 500720187.0,
      "step": 14394
    },
    {
      "epoch": 2.673166202414113,
      "grad_norm": 1.5090542353769592,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.881988525390625,
      "num_tokens": 500759321.0,
      "step": 14395
    },
    {
      "epoch": 2.673351903435469,
      "grad_norm": 1.6812196124658285,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8920645713806152,
      "num_tokens": 500793245.0,
      "step": 14396
    },
    {
      "epoch": 2.6735376044568246,
      "grad_norm": 1.5565643437329144,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8919689059257507,
      "num_tokens": 500830107.0,
      "step": 14397
    },
    {
      "epoch": 2.6737233054781804,
      "grad_norm": 1.7318095954782327,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8905982971191406,
      "num_tokens": 500858879.0,
      "step": 14398
    },
    {
      "epoch": 2.6739090064995357,
      "grad_norm": 1.5465392654072685,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8822692036628723,
      "num_tokens": 500895586.0,
      "step": 14399
    },
    {
      "epoch": 2.6740947075208914,
      "grad_norm": 1.6491119132221341,
      "learning_rate": 1e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8706772327423096,
      "num_tokens": 500932139.0,
      "step": 14400
    },
    {
      "epoch": 2.674280408542247,
      "grad_norm": 1.6117000307080531,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8853768110275269,
      "num_tokens": 500966106.0,
      "step": 14401
    },
    {
      "epoch": 2.6744661095636024,
      "grad_norm": 1.480785492891583,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8886562585830688,
      "num_tokens": 501003591.0,
      "step": 14402
    },
    {
      "epoch": 2.674651810584958,
      "grad_norm": 1.7950875894731628,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8780475854873657,
      "num_tokens": 501034228.0,
      "step": 14403
    },
    {
      "epoch": 2.674837511606314,
      "grad_norm": 1.4810077423933568,
      "learning_rate": 1e-06,
      "loss": 0.2768,
      "mean_token_accuracy": 0.9022524952888489,
      "num_tokens": 501069235.0,
      "step": 14404
    },
    {
      "epoch": 2.6750232126276696,
      "grad_norm": 1.5938768923741196,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8941850662231445,
      "num_tokens": 501102688.0,
      "step": 14405
    },
    {
      "epoch": 2.675208913649025,
      "grad_norm": 1.7408712291016686,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8796841502189636,
      "num_tokens": 501132704.0,
      "step": 14406
    },
    {
      "epoch": 2.6753946146703806,
      "grad_norm": 1.5762100855426988,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8813213109970093,
      "num_tokens": 501167429.0,
      "step": 14407
    },
    {
      "epoch": 2.6755803156917364,
      "grad_norm": 1.5197296841385837,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8815245032310486,
      "num_tokens": 501203815.0,
      "step": 14408
    },
    {
      "epoch": 2.6757660167130917,
      "grad_norm": 1.4577500108531583,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8967627882957458,
      "num_tokens": 501244328.0,
      "step": 14409
    },
    {
      "epoch": 2.6759517177344474,
      "grad_norm": 1.6904854742170408,
      "learning_rate": 1e-06,
      "loss": 0.3741,
      "mean_token_accuracy": 0.8717951774597168,
      "num_tokens": 501278302.0,
      "step": 14410
    },
    {
      "epoch": 2.676137418755803,
      "grad_norm": 1.5272524121861837,
      "learning_rate": 1e-06,
      "loss": 0.2703,
      "mean_token_accuracy": 0.9008713960647583,
      "num_tokens": 501312399.0,
      "step": 14411
    },
    {
      "epoch": 2.676323119777159,
      "grad_norm": 1.5403337215005193,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.889572262763977,
      "num_tokens": 501346273.0,
      "step": 14412
    },
    {
      "epoch": 2.6765088207985146,
      "grad_norm": 1.5883190842635881,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8828051090240479,
      "num_tokens": 501377531.0,
      "step": 14413
    },
    {
      "epoch": 2.67669452181987,
      "grad_norm": 1.570639250870121,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.880679726600647,
      "num_tokens": 501411921.0,
      "step": 14414
    },
    {
      "epoch": 2.6768802228412256,
      "grad_norm": 1.7142533313150317,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8893119096755981,
      "num_tokens": 501441059.0,
      "step": 14415
    },
    {
      "epoch": 2.6770659238625814,
      "grad_norm": 1.6130839985216139,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8868271112442017,
      "num_tokens": 501473503.0,
      "step": 14416
    },
    {
      "epoch": 2.6772516248839366,
      "grad_norm": 1.4881519638268697,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8877689838409424,
      "num_tokens": 501512956.0,
      "step": 14417
    },
    {
      "epoch": 2.6774373259052924,
      "grad_norm": 1.6035395673506652,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8860882520675659,
      "num_tokens": 501547328.0,
      "step": 14418
    },
    {
      "epoch": 2.677623026926648,
      "grad_norm": 1.5948961161614013,
      "learning_rate": 1e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8651173114776611,
      "num_tokens": 501584974.0,
      "step": 14419
    },
    {
      "epoch": 2.677808727948004,
      "grad_norm": 1.4720208953269154,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8738997578620911,
      "num_tokens": 501627996.0,
      "step": 14420
    },
    {
      "epoch": 2.6779944289693596,
      "grad_norm": 1.4957492043368978,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8855552077293396,
      "num_tokens": 501665438.0,
      "step": 14421
    },
    {
      "epoch": 2.678180129990715,
      "grad_norm": 1.5497231985538915,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8939372301101685,
      "num_tokens": 501699749.0,
      "step": 14422
    },
    {
      "epoch": 2.6783658310120706,
      "grad_norm": 1.5763801350277666,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8941368460655212,
      "num_tokens": 501738240.0,
      "step": 14423
    },
    {
      "epoch": 2.6785515320334263,
      "grad_norm": 1.7097464354144467,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8946661949157715,
      "num_tokens": 501766242.0,
      "step": 14424
    },
    {
      "epoch": 2.6787372330547816,
      "grad_norm": 1.5731891128834004,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8915317058563232,
      "num_tokens": 501800632.0,
      "step": 14425
    },
    {
      "epoch": 2.6789229340761374,
      "grad_norm": 1.6253594557399231,
      "learning_rate": 1e-06,
      "loss": 0.2921,
      "mean_token_accuracy": 0.8956655859947205,
      "num_tokens": 501829245.0,
      "step": 14426
    },
    {
      "epoch": 2.679108635097493,
      "grad_norm": 1.704968558563376,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8845586180686951,
      "num_tokens": 501866316.0,
      "step": 14427
    },
    {
      "epoch": 2.679294336118849,
      "grad_norm": 1.476068491017205,
      "learning_rate": 1e-06,
      "loss": 0.2858,
      "mean_token_accuracy": 0.8967325091362,
      "num_tokens": 501904910.0,
      "step": 14428
    },
    {
      "epoch": 2.679480037140204,
      "grad_norm": 1.6045700643618668,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.8951661586761475,
      "num_tokens": 501936365.0,
      "step": 14429
    },
    {
      "epoch": 2.67966573816156,
      "grad_norm": 1.557670119700243,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8872858285903931,
      "num_tokens": 501971143.0,
      "step": 14430
    },
    {
      "epoch": 2.6798514391829156,
      "grad_norm": 1.662829262519436,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8956508636474609,
      "num_tokens": 502001833.0,
      "step": 14431
    },
    {
      "epoch": 2.680037140204271,
      "grad_norm": 1.5140492697894175,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.8929623365402222,
      "num_tokens": 502034744.0,
      "step": 14432
    },
    {
      "epoch": 2.6802228412256266,
      "grad_norm": 1.5101983676030444,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.8933085799217224,
      "num_tokens": 502072744.0,
      "step": 14433
    },
    {
      "epoch": 2.6804085422469823,
      "grad_norm": 1.5559628363389482,
      "learning_rate": 1e-06,
      "loss": 0.2962,
      "mean_token_accuracy": 0.8953334093093872,
      "num_tokens": 502110581.0,
      "step": 14434
    },
    {
      "epoch": 2.680594243268338,
      "grad_norm": 1.6973960844709164,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8896499872207642,
      "num_tokens": 502141012.0,
      "step": 14435
    },
    {
      "epoch": 2.680779944289694,
      "grad_norm": 1.5755799463659097,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8958324193954468,
      "num_tokens": 502173185.0,
      "step": 14436
    },
    {
      "epoch": 2.680965645311049,
      "grad_norm": 1.461685459589832,
      "learning_rate": 1e-06,
      "loss": 0.2921,
      "mean_token_accuracy": 0.8950881958007812,
      "num_tokens": 502211225.0,
      "step": 14437
    },
    {
      "epoch": 2.681151346332405,
      "grad_norm": 1.4296135438736703,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8800673484802246,
      "num_tokens": 502253515.0,
      "step": 14438
    },
    {
      "epoch": 2.6813370473537605,
      "grad_norm": 1.7894511446575645,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.8757994174957275,
      "num_tokens": 502281835.0,
      "step": 14439
    },
    {
      "epoch": 2.681522748375116,
      "grad_norm": 1.616357362437175,
      "learning_rate": 1e-06,
      "loss": 0.3174,
      "mean_token_accuracy": 0.8869593739509583,
      "num_tokens": 502317486.0,
      "step": 14440
    },
    {
      "epoch": 2.6817084493964716,
      "grad_norm": 1.6140325259600101,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8864502906799316,
      "num_tokens": 502353668.0,
      "step": 14441
    },
    {
      "epoch": 2.6818941504178273,
      "grad_norm": 1.5931192794803657,
      "learning_rate": 1e-06,
      "loss": 0.3659,
      "mean_token_accuracy": 0.8731538653373718,
      "num_tokens": 502391545.0,
      "step": 14442
    },
    {
      "epoch": 2.682079851439183,
      "grad_norm": 1.5218590386504591,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8802688121795654,
      "num_tokens": 502426548.0,
      "step": 14443
    },
    {
      "epoch": 2.6822655524605388,
      "grad_norm": 1.5111396655345697,
      "learning_rate": 1e-06,
      "loss": 0.3169,
      "mean_token_accuracy": 0.8894731998443604,
      "num_tokens": 502463223.0,
      "step": 14444
    },
    {
      "epoch": 2.682451253481894,
      "grad_norm": 1.6057537824047854,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8827100992202759,
      "num_tokens": 502496743.0,
      "step": 14445
    },
    {
      "epoch": 2.68263695450325,
      "grad_norm": 1.4991606747238533,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8796172142028809,
      "num_tokens": 502535128.0,
      "step": 14446
    },
    {
      "epoch": 2.6828226555246055,
      "grad_norm": 1.6254069603516397,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.8996708989143372,
      "num_tokens": 502564369.0,
      "step": 14447
    },
    {
      "epoch": 2.683008356545961,
      "grad_norm": 1.4939979331839734,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8851317167282104,
      "num_tokens": 502606881.0,
      "step": 14448
    },
    {
      "epoch": 2.6831940575673165,
      "grad_norm": 1.5622687235555563,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8832647800445557,
      "num_tokens": 502642997.0,
      "step": 14449
    },
    {
      "epoch": 2.6833797585886723,
      "grad_norm": 1.5480020778577717,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8898258209228516,
      "num_tokens": 502676368.0,
      "step": 14450
    },
    {
      "epoch": 2.683565459610028,
      "grad_norm": 1.3721363317122108,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8908593654632568,
      "num_tokens": 502720138.0,
      "step": 14451
    },
    {
      "epoch": 2.6837511606313837,
      "grad_norm": 1.497520779466188,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8921663761138916,
      "num_tokens": 502754492.0,
      "step": 14452
    },
    {
      "epoch": 2.683936861652739,
      "grad_norm": 1.5211146778160096,
      "learning_rate": 1e-06,
      "loss": 0.2937,
      "mean_token_accuracy": 0.8950529098510742,
      "num_tokens": 502790850.0,
      "step": 14453
    },
    {
      "epoch": 2.6841225626740948,
      "grad_norm": 1.6329714363940449,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8930246233940125,
      "num_tokens": 502819853.0,
      "step": 14454
    },
    {
      "epoch": 2.68430826369545,
      "grad_norm": 1.5416890511619952,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8779857754707336,
      "num_tokens": 502857767.0,
      "step": 14455
    },
    {
      "epoch": 2.684493964716806,
      "grad_norm": 1.7198112054522923,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8882378339767456,
      "num_tokens": 502889013.0,
      "step": 14456
    },
    {
      "epoch": 2.6846796657381615,
      "grad_norm": 1.7472383007112489,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.886048436164856,
      "num_tokens": 502916743.0,
      "step": 14457
    },
    {
      "epoch": 2.6848653667595173,
      "grad_norm": 1.748682341244689,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8844373226165771,
      "num_tokens": 502943852.0,
      "step": 14458
    },
    {
      "epoch": 2.685051067780873,
      "grad_norm": 1.4818717987095078,
      "learning_rate": 1e-06,
      "loss": 0.2634,
      "mean_token_accuracy": 0.9047911167144775,
      "num_tokens": 502978919.0,
      "step": 14459
    },
    {
      "epoch": 2.6852367688022283,
      "grad_norm": 1.6346554334081105,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8894780874252319,
      "num_tokens": 503008999.0,
      "step": 14460
    },
    {
      "epoch": 2.685422469823584,
      "grad_norm": 1.4561076714096564,
      "learning_rate": 1e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8846322298049927,
      "num_tokens": 503053414.0,
      "step": 14461
    },
    {
      "epoch": 2.6856081708449397,
      "grad_norm": 1.5172271657382406,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8885196447372437,
      "num_tokens": 503086581.0,
      "step": 14462
    },
    {
      "epoch": 2.685793871866295,
      "grad_norm": 1.680495315244291,
      "learning_rate": 1e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8768225908279419,
      "num_tokens": 503118319.0,
      "step": 14463
    },
    {
      "epoch": 2.6859795728876508,
      "grad_norm": 1.4885366474783757,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8891241550445557,
      "num_tokens": 503158942.0,
      "step": 14464
    },
    {
      "epoch": 2.6861652739090065,
      "grad_norm": 1.4146359483935642,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8943950533866882,
      "num_tokens": 503199231.0,
      "step": 14465
    },
    {
      "epoch": 2.6863509749303622,
      "grad_norm": 1.5403709764149331,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8800780177116394,
      "num_tokens": 503234931.0,
      "step": 14466
    },
    {
      "epoch": 2.686536675951718,
      "grad_norm": 1.4448252832016333,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8936376571655273,
      "num_tokens": 503272731.0,
      "step": 14467
    },
    {
      "epoch": 2.6867223769730733,
      "grad_norm": 1.6986753927843545,
      "learning_rate": 1e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8733689785003662,
      "num_tokens": 503307188.0,
      "step": 14468
    },
    {
      "epoch": 2.686908077994429,
      "grad_norm": 1.5508889382943598,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8923729658126831,
      "num_tokens": 503341763.0,
      "step": 14469
    },
    {
      "epoch": 2.6870937790157847,
      "grad_norm": 1.5034471216552818,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8937461972236633,
      "num_tokens": 503381138.0,
      "step": 14470
    },
    {
      "epoch": 2.68727948003714,
      "grad_norm": 1.4617252065791382,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8828930258750916,
      "num_tokens": 503420825.0,
      "step": 14471
    },
    {
      "epoch": 2.6874651810584957,
      "grad_norm": 1.538799997622756,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.888548731803894,
      "num_tokens": 503455653.0,
      "step": 14472
    },
    {
      "epoch": 2.6876508820798515,
      "grad_norm": 1.7063969315971173,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8828277587890625,
      "num_tokens": 503486347.0,
      "step": 14473
    },
    {
      "epoch": 2.687836583101207,
      "grad_norm": 1.6685150154370034,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8806011080741882,
      "num_tokens": 503519596.0,
      "step": 14474
    },
    {
      "epoch": 2.688022284122563,
      "grad_norm": 1.5815946768538405,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8822872042655945,
      "num_tokens": 503558242.0,
      "step": 14475
    },
    {
      "epoch": 2.6882079851439182,
      "grad_norm": 3.667541546526143,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8871281147003174,
      "num_tokens": 503594679.0,
      "step": 14476
    },
    {
      "epoch": 2.688393686165274,
      "grad_norm": 1.6772231289363932,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8826473951339722,
      "num_tokens": 503626347.0,
      "step": 14477
    },
    {
      "epoch": 2.6885793871866293,
      "grad_norm": 1.6321194908513292,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.8854070901870728,
      "num_tokens": 503654222.0,
      "step": 14478
    },
    {
      "epoch": 2.688765088207985,
      "grad_norm": 1.5237330442386305,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8935868740081787,
      "num_tokens": 503688921.0,
      "step": 14479
    },
    {
      "epoch": 2.6889507892293407,
      "grad_norm": 1.6830915433923344,
      "learning_rate": 1e-06,
      "loss": 0.2398,
      "mean_token_accuracy": 0.914445698261261,
      "num_tokens": 503714841.0,
      "step": 14480
    },
    {
      "epoch": 2.6891364902506965,
      "grad_norm": 1.575515849404465,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.876412034034729,
      "num_tokens": 503752645.0,
      "step": 14481
    },
    {
      "epoch": 2.689322191272052,
      "grad_norm": 1.4166399699385848,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.882766842842102,
      "num_tokens": 503794322.0,
      "step": 14482
    },
    {
      "epoch": 2.6895078922934075,
      "grad_norm": 1.5139960146806466,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8731147646903992,
      "num_tokens": 503835961.0,
      "step": 14483
    },
    {
      "epoch": 2.689693593314763,
      "grad_norm": 1.5206907980011148,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8814718723297119,
      "num_tokens": 503873497.0,
      "step": 14484
    },
    {
      "epoch": 2.689879294336119,
      "grad_norm": 1.5159151796046555,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8811540603637695,
      "num_tokens": 503912262.0,
      "step": 14485
    },
    {
      "epoch": 2.6900649953574742,
      "grad_norm": 1.642996156646574,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.88658607006073,
      "num_tokens": 503942335.0,
      "step": 14486
    },
    {
      "epoch": 2.69025069637883,
      "grad_norm": 1.4455217759728136,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.901528000831604,
      "num_tokens": 503979608.0,
      "step": 14487
    },
    {
      "epoch": 2.6904363974001857,
      "grad_norm": 1.5082456548643415,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.8957758545875549,
      "num_tokens": 504016325.0,
      "step": 14488
    },
    {
      "epoch": 2.6906220984215414,
      "grad_norm": 1.650752261543233,
      "learning_rate": 1e-06,
      "loss": 0.3038,
      "mean_token_accuracy": 0.8931901454925537,
      "num_tokens": 504046241.0,
      "step": 14489
    },
    {
      "epoch": 2.690807799442897,
      "grad_norm": 1.533995244287667,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.891059398651123,
      "num_tokens": 504077756.0,
      "step": 14490
    },
    {
      "epoch": 2.6909935004642525,
      "grad_norm": 1.6573076883337898,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8960819840431213,
      "num_tokens": 504105190.0,
      "step": 14491
    },
    {
      "epoch": 2.691179201485608,
      "grad_norm": 1.5072116394290955,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.884361982345581,
      "num_tokens": 504143280.0,
      "step": 14492
    },
    {
      "epoch": 2.691364902506964,
      "grad_norm": 1.705354472548485,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8893851637840271,
      "num_tokens": 504169822.0,
      "step": 14493
    },
    {
      "epoch": 2.691550603528319,
      "grad_norm": 1.481614305509425,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8843964338302612,
      "num_tokens": 504206264.0,
      "step": 14494
    },
    {
      "epoch": 2.691736304549675,
      "grad_norm": 1.514082021995741,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8874911069869995,
      "num_tokens": 504240888.0,
      "step": 14495
    },
    {
      "epoch": 2.6919220055710307,
      "grad_norm": 1.6335567488214187,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8920360803604126,
      "num_tokens": 504274499.0,
      "step": 14496
    },
    {
      "epoch": 2.6921077065923864,
      "grad_norm": 1.7639373086376382,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8891315460205078,
      "num_tokens": 504303257.0,
      "step": 14497
    },
    {
      "epoch": 2.692293407613742,
      "grad_norm": 1.6012141540245517,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8746326565742493,
      "num_tokens": 504339816.0,
      "step": 14498
    },
    {
      "epoch": 2.6924791086350974,
      "grad_norm": 1.4973166894762995,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8892964124679565,
      "num_tokens": 504374676.0,
      "step": 14499
    },
    {
      "epoch": 2.692664809656453,
      "grad_norm": 1.431255012086597,
      "learning_rate": 1e-06,
      "loss": 0.2859,
      "mean_token_accuracy": 0.8941853642463684,
      "num_tokens": 504410829.0,
      "step": 14500
    },
    {
      "epoch": 2.6928505106778085,
      "grad_norm": 1.5335280845745547,
      "learning_rate": 1e-06,
      "loss": 0.3544,
      "mean_token_accuracy": 0.8771907687187195,
      "num_tokens": 504450855.0,
      "step": 14501
    },
    {
      "epoch": 2.693036211699164,
      "grad_norm": 1.5974181977743984,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8870282173156738,
      "num_tokens": 504483959.0,
      "step": 14502
    },
    {
      "epoch": 2.69322191272052,
      "grad_norm": 1.5428156959712636,
      "learning_rate": 1e-06,
      "loss": 0.3,
      "mean_token_accuracy": 0.8945763111114502,
      "num_tokens": 504519417.0,
      "step": 14503
    },
    {
      "epoch": 2.6934076137418757,
      "grad_norm": 1.4937439496678535,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8909158706665039,
      "num_tokens": 504559440.0,
      "step": 14504
    },
    {
      "epoch": 2.6935933147632314,
      "grad_norm": 1.5208712187566689,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8955001831054688,
      "num_tokens": 504593226.0,
      "step": 14505
    },
    {
      "epoch": 2.6937790157845867,
      "grad_norm": 1.4863226392967928,
      "learning_rate": 1e-06,
      "loss": 0.2746,
      "mean_token_accuracy": 0.9001733064651489,
      "num_tokens": 504625603.0,
      "step": 14506
    },
    {
      "epoch": 2.6939647168059424,
      "grad_norm": 1.5050296113743524,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8913474082946777,
      "num_tokens": 504664125.0,
      "step": 14507
    },
    {
      "epoch": 2.694150417827298,
      "grad_norm": 1.5794075320830225,
      "learning_rate": 1e-06,
      "loss": 0.2905,
      "mean_token_accuracy": 0.8972676992416382,
      "num_tokens": 504697583.0,
      "step": 14508
    },
    {
      "epoch": 2.6943361188486534,
      "grad_norm": 1.5974104976370387,
      "learning_rate": 1e-06,
      "loss": 0.3006,
      "mean_token_accuracy": 0.8962356448173523,
      "num_tokens": 504729278.0,
      "step": 14509
    },
    {
      "epoch": 2.694521819870009,
      "grad_norm": 1.558660055850489,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8819433450698853,
      "num_tokens": 504767611.0,
      "step": 14510
    },
    {
      "epoch": 2.694707520891365,
      "grad_norm": 1.7013999014477574,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8877249956130981,
      "num_tokens": 504797678.0,
      "step": 14511
    },
    {
      "epoch": 2.6948932219127206,
      "grad_norm": 1.5319968103014177,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8911048173904419,
      "num_tokens": 504834389.0,
      "step": 14512
    },
    {
      "epoch": 2.6950789229340764,
      "grad_norm": 1.6198276504138733,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8769835233688354,
      "num_tokens": 504872115.0,
      "step": 14513
    },
    {
      "epoch": 2.6952646239554316,
      "grad_norm": 1.6720736872564268,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.886499285697937,
      "num_tokens": 504904128.0,
      "step": 14514
    },
    {
      "epoch": 2.6954503249767874,
      "grad_norm": 1.8509087865249056,
      "learning_rate": 1e-06,
      "loss": 0.3045,
      "mean_token_accuracy": 0.8906108736991882,
      "num_tokens": 504933389.0,
      "step": 14515
    },
    {
      "epoch": 2.695636025998143,
      "grad_norm": 1.5699070210846497,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8891795873641968,
      "num_tokens": 504970181.0,
      "step": 14516
    },
    {
      "epoch": 2.6958217270194984,
      "grad_norm": 1.526626742473697,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8803479671478271,
      "num_tokens": 505007919.0,
      "step": 14517
    },
    {
      "epoch": 2.696007428040854,
      "grad_norm": 1.6514104867057744,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8832988142967224,
      "num_tokens": 505038915.0,
      "step": 14518
    },
    {
      "epoch": 2.69619312906221,
      "grad_norm": 1.4463305315196133,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8923403024673462,
      "num_tokens": 505080009.0,
      "step": 14519
    },
    {
      "epoch": 2.6963788300835656,
      "grad_norm": 1.6973745357898915,
      "learning_rate": 1e-06,
      "loss": 0.3561,
      "mean_token_accuracy": 0.8745057582855225,
      "num_tokens": 505108560.0,
      "step": 14520
    },
    {
      "epoch": 2.6965645311049213,
      "grad_norm": 1.585527455740912,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8823403120040894,
      "num_tokens": 505143299.0,
      "step": 14521
    },
    {
      "epoch": 2.6967502321262766,
      "grad_norm": 1.4659739844073207,
      "learning_rate": 1e-06,
      "loss": 0.3467,
      "mean_token_accuracy": 0.8788275718688965,
      "num_tokens": 505184587.0,
      "step": 14522
    },
    {
      "epoch": 2.6969359331476324,
      "grad_norm": 1.7753728375345372,
      "learning_rate": 1e-06,
      "loss": 0.3277,
      "mean_token_accuracy": 0.8835572600364685,
      "num_tokens": 505213082.0,
      "step": 14523
    },
    {
      "epoch": 2.697121634168988,
      "grad_norm": 1.4840465419050384,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8982166647911072,
      "num_tokens": 505250224.0,
      "step": 14524
    },
    {
      "epoch": 2.6973073351903434,
      "grad_norm": 1.5066448666408707,
      "learning_rate": 1e-06,
      "loss": 0.2748,
      "mean_token_accuracy": 0.9014825820922852,
      "num_tokens": 505285975.0,
      "step": 14525
    },
    {
      "epoch": 2.697493036211699,
      "grad_norm": 1.686840734978736,
      "learning_rate": 1e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.8739809989929199,
      "num_tokens": 505321791.0,
      "step": 14526
    },
    {
      "epoch": 2.697678737233055,
      "grad_norm": 1.5165577828064085,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8826106786727905,
      "num_tokens": 505359214.0,
      "step": 14527
    },
    {
      "epoch": 2.6978644382544106,
      "grad_norm": 1.6303448323717342,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8910119533538818,
      "num_tokens": 505396133.0,
      "step": 14528
    },
    {
      "epoch": 2.698050139275766,
      "grad_norm": 1.7371904637017932,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.8971037864685059,
      "num_tokens": 505422302.0,
      "step": 14529
    },
    {
      "epoch": 2.6982358402971216,
      "grad_norm": 1.5985875052786953,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8809911012649536,
      "num_tokens": 505459501.0,
      "step": 14530
    },
    {
      "epoch": 2.6984215413184773,
      "grad_norm": 1.4449938746593167,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8946770429611206,
      "num_tokens": 505499767.0,
      "step": 14531
    },
    {
      "epoch": 2.6986072423398326,
      "grad_norm": 1.6289736432464992,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8886047005653381,
      "num_tokens": 505533779.0,
      "step": 14532
    },
    {
      "epoch": 2.6987929433611884,
      "grad_norm": 1.5096441680491632,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8797677755355835,
      "num_tokens": 505573583.0,
      "step": 14533
    },
    {
      "epoch": 2.698978644382544,
      "grad_norm": 1.4686926252132555,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8855512142181396,
      "num_tokens": 505613544.0,
      "step": 14534
    },
    {
      "epoch": 2.6991643454039,
      "grad_norm": 1.6308212887999676,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8824769854545593,
      "num_tokens": 505651839.0,
      "step": 14535
    },
    {
      "epoch": 2.6993500464252556,
      "grad_norm": 1.631629484057112,
      "learning_rate": 1e-06,
      "loss": 0.3634,
      "mean_token_accuracy": 0.8753329515457153,
      "num_tokens": 505687107.0,
      "step": 14536
    },
    {
      "epoch": 2.699535747446611,
      "grad_norm": 1.5345729882436272,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8921486139297485,
      "num_tokens": 505721884.0,
      "step": 14537
    },
    {
      "epoch": 2.6997214484679666,
      "grad_norm": 1.6349393257809048,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8800955414772034,
      "num_tokens": 505756920.0,
      "step": 14538
    },
    {
      "epoch": 2.6999071494893223,
      "grad_norm": 1.4259768234643782,
      "learning_rate": 1e-06,
      "loss": 0.2674,
      "mean_token_accuracy": 0.9052306413650513,
      "num_tokens": 505791361.0,
      "step": 14539
    },
    {
      "epoch": 2.7000928505106776,
      "grad_norm": 1.6988448403949763,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.8755994439125061,
      "num_tokens": 505826245.0,
      "step": 14540
    },
    {
      "epoch": 2.7002785515320333,
      "grad_norm": 1.5982538804357578,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8773874640464783,
      "num_tokens": 505860707.0,
      "step": 14541
    },
    {
      "epoch": 2.700464252553389,
      "grad_norm": 1.5640770414016634,
      "learning_rate": 1e-06,
      "loss": 0.304,
      "mean_token_accuracy": 0.890073835849762,
      "num_tokens": 505896968.0,
      "step": 14542
    },
    {
      "epoch": 2.700649953574745,
      "grad_norm": 1.7584926399733694,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8903152942657471,
      "num_tokens": 505925145.0,
      "step": 14543
    },
    {
      "epoch": 2.7008356545961005,
      "grad_norm": 1.589217022391135,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8844797611236572,
      "num_tokens": 505957102.0,
      "step": 14544
    },
    {
      "epoch": 2.701021355617456,
      "grad_norm": 1.7122440134096388,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8853499889373779,
      "num_tokens": 505985797.0,
      "step": 14545
    },
    {
      "epoch": 2.7012070566388116,
      "grad_norm": 1.5817745498560118,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8859610557556152,
      "num_tokens": 506027020.0,
      "step": 14546
    },
    {
      "epoch": 2.7013927576601673,
      "grad_norm": 1.5843572732319346,
      "learning_rate": 1e-06,
      "loss": 0.3537,
      "mean_token_accuracy": 0.8760166168212891,
      "num_tokens": 506061832.0,
      "step": 14547
    },
    {
      "epoch": 2.7015784586815226,
      "grad_norm": 1.6990275083780597,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8865221738815308,
      "num_tokens": 506092343.0,
      "step": 14548
    },
    {
      "epoch": 2.7017641597028783,
      "grad_norm": 1.5843856292663623,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.8930309414863586,
      "num_tokens": 506126324.0,
      "step": 14549
    },
    {
      "epoch": 2.701949860724234,
      "grad_norm": 1.5551360049884184,
      "learning_rate": 1e-06,
      "loss": 0.2872,
      "mean_token_accuracy": 0.8964213132858276,
      "num_tokens": 506160310.0,
      "step": 14550
    },
    {
      "epoch": 2.7021355617455898,
      "grad_norm": 1.5882270140965362,
      "learning_rate": 1e-06,
      "loss": 0.2979,
      "mean_token_accuracy": 0.8949492573738098,
      "num_tokens": 506192388.0,
      "step": 14551
    },
    {
      "epoch": 2.702321262766945,
      "grad_norm": 1.621278212584404,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8831275105476379,
      "num_tokens": 506227971.0,
      "step": 14552
    },
    {
      "epoch": 2.702506963788301,
      "grad_norm": 1.6374709076131586,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8884395360946655,
      "num_tokens": 506260347.0,
      "step": 14553
    },
    {
      "epoch": 2.7026926648096565,
      "grad_norm": 1.4934584708631273,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8834348917007446,
      "num_tokens": 506306266.0,
      "step": 14554
    },
    {
      "epoch": 2.702878365831012,
      "grad_norm": 1.477310919019406,
      "learning_rate": 1e-06,
      "loss": 0.3446,
      "mean_token_accuracy": 0.880649745464325,
      "num_tokens": 506347969.0,
      "step": 14555
    },
    {
      "epoch": 2.7030640668523676,
      "grad_norm": 1.5661868320494217,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8906148672103882,
      "num_tokens": 506384162.0,
      "step": 14556
    },
    {
      "epoch": 2.7032497678737233,
      "grad_norm": 1.6150349548687581,
      "learning_rate": 1e-06,
      "loss": 0.3573,
      "mean_token_accuracy": 0.8754732608795166,
      "num_tokens": 506420010.0,
      "step": 14557
    },
    {
      "epoch": 2.703435468895079,
      "grad_norm": 1.6132628342305448,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8894730806350708,
      "num_tokens": 506450617.0,
      "step": 14558
    },
    {
      "epoch": 2.7036211699164348,
      "grad_norm": 1.5691420054125418,
      "learning_rate": 1e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8656297326087952,
      "num_tokens": 506490334.0,
      "step": 14559
    },
    {
      "epoch": 2.70380687093779,
      "grad_norm": 1.6578451698556873,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8742977380752563,
      "num_tokens": 506522293.0,
      "step": 14560
    },
    {
      "epoch": 2.7039925719591458,
      "grad_norm": 1.4895041492950807,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8828564882278442,
      "num_tokens": 506560029.0,
      "step": 14561
    },
    {
      "epoch": 2.7041782729805015,
      "grad_norm": 1.7170281589035692,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8845393657684326,
      "num_tokens": 506592468.0,
      "step": 14562
    },
    {
      "epoch": 2.704363974001857,
      "grad_norm": 1.753917986154877,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8784095048904419,
      "num_tokens": 506623563.0,
      "step": 14563
    },
    {
      "epoch": 2.7045496750232125,
      "grad_norm": 1.4654612932426327,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8922121524810791,
      "num_tokens": 506663888.0,
      "step": 14564
    },
    {
      "epoch": 2.7047353760445683,
      "grad_norm": 1.413422079937373,
      "learning_rate": 1e-06,
      "loss": 0.2822,
      "mean_token_accuracy": 0.8968722820281982,
      "num_tokens": 506700344.0,
      "step": 14565
    },
    {
      "epoch": 2.704921077065924,
      "grad_norm": 1.6067320682060908,
      "learning_rate": 1e-06,
      "loss": 0.3035,
      "mean_token_accuracy": 0.8902090191841125,
      "num_tokens": 506730834.0,
      "step": 14566
    },
    {
      "epoch": 2.7051067780872797,
      "grad_norm": 1.4862184969637313,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8822104930877686,
      "num_tokens": 506772164.0,
      "step": 14567
    },
    {
      "epoch": 2.705292479108635,
      "grad_norm": 1.4975880253852805,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8826300501823425,
      "num_tokens": 506813370.0,
      "step": 14568
    },
    {
      "epoch": 2.7054781801299908,
      "grad_norm": 1.5086606719780393,
      "learning_rate": 1e-06,
      "loss": 0.3088,
      "mean_token_accuracy": 0.8923092484474182,
      "num_tokens": 506847209.0,
      "step": 14569
    },
    {
      "epoch": 2.7056638811513465,
      "grad_norm": 1.4813018160891338,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8907424211502075,
      "num_tokens": 506883378.0,
      "step": 14570
    },
    {
      "epoch": 2.7058495821727018,
      "grad_norm": 1.725294796656919,
      "learning_rate": 1e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8657914996147156,
      "num_tokens": 506917334.0,
      "step": 14571
    },
    {
      "epoch": 2.7060352831940575,
      "grad_norm": 1.653624189563657,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8834387063980103,
      "num_tokens": 506950687.0,
      "step": 14572
    },
    {
      "epoch": 2.7062209842154132,
      "grad_norm": 1.5341906540438628,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8908649682998657,
      "num_tokens": 506984277.0,
      "step": 14573
    },
    {
      "epoch": 2.706406685236769,
      "grad_norm": 1.512127066295382,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8826519250869751,
      "num_tokens": 507021729.0,
      "step": 14574
    },
    {
      "epoch": 2.7065923862581243,
      "grad_norm": 1.6814195617996726,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8812050819396973,
      "num_tokens": 507057108.0,
      "step": 14575
    },
    {
      "epoch": 2.70677808727948,
      "grad_norm": 1.619316149749132,
      "learning_rate": 1e-06,
      "loss": 0.2789,
      "mean_token_accuracy": 0.9015202522277832,
      "num_tokens": 507085742.0,
      "step": 14576
    },
    {
      "epoch": 2.7069637883008357,
      "grad_norm": 1.6182800216140516,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.880323052406311,
      "num_tokens": 507119353.0,
      "step": 14577
    },
    {
      "epoch": 2.707149489322191,
      "grad_norm": 1.5828532655784013,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8823615312576294,
      "num_tokens": 507155190.0,
      "step": 14578
    },
    {
      "epoch": 2.7073351903435467,
      "grad_norm": 1.5028979950341397,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8878509998321533,
      "num_tokens": 507191292.0,
      "step": 14579
    },
    {
      "epoch": 2.7075208913649025,
      "grad_norm": 1.5886600432905484,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.884642481803894,
      "num_tokens": 507226932.0,
      "step": 14580
    },
    {
      "epoch": 2.707706592386258,
      "grad_norm": 1.497830337946089,
      "learning_rate": 1e-06,
      "loss": 0.2919,
      "mean_token_accuracy": 0.8987718820571899,
      "num_tokens": 507261960.0,
      "step": 14581
    },
    {
      "epoch": 2.707892293407614,
      "grad_norm": 1.537468711219755,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8867990374565125,
      "num_tokens": 507298597.0,
      "step": 14582
    },
    {
      "epoch": 2.7080779944289692,
      "grad_norm": 1.5972473721915317,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8889200687408447,
      "num_tokens": 507331249.0,
      "step": 14583
    },
    {
      "epoch": 2.708263695450325,
      "grad_norm": 1.4568958209853482,
      "learning_rate": 1e-06,
      "loss": 0.3463,
      "mean_token_accuracy": 0.8792030811309814,
      "num_tokens": 507374473.0,
      "step": 14584
    },
    {
      "epoch": 2.7084493964716807,
      "grad_norm": 1.753595390948066,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8821808695793152,
      "num_tokens": 507401253.0,
      "step": 14585
    },
    {
      "epoch": 2.708635097493036,
      "grad_norm": 1.7195464408022614,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8894094824790955,
      "num_tokens": 507433867.0,
      "step": 14586
    },
    {
      "epoch": 2.7088207985143917,
      "grad_norm": 1.6906252337406324,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8813725709915161,
      "num_tokens": 507464234.0,
      "step": 14587
    },
    {
      "epoch": 2.7090064995357475,
      "grad_norm": 1.516517109325047,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8785333037376404,
      "num_tokens": 507504148.0,
      "step": 14588
    },
    {
      "epoch": 2.709192200557103,
      "grad_norm": 1.690515415023919,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8856747150421143,
      "num_tokens": 507534980.0,
      "step": 14589
    },
    {
      "epoch": 2.709377901578459,
      "grad_norm": 1.5062052827140893,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8903096914291382,
      "num_tokens": 507570851.0,
      "step": 14590
    },
    {
      "epoch": 2.709563602599814,
      "grad_norm": 1.4848030447286282,
      "learning_rate": 1e-06,
      "loss": 0.2485,
      "mean_token_accuracy": 0.9102758169174194,
      "num_tokens": 507604516.0,
      "step": 14591
    },
    {
      "epoch": 2.70974930362117,
      "grad_norm": 1.671194422722456,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8928399682044983,
      "num_tokens": 507633439.0,
      "step": 14592
    },
    {
      "epoch": 2.7099350046425257,
      "grad_norm": 1.5777079952688373,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8792886734008789,
      "num_tokens": 507662891.0,
      "step": 14593
    },
    {
      "epoch": 2.710120705663881,
      "grad_norm": 1.441446929385858,
      "learning_rate": 1e-06,
      "loss": 0.2908,
      "mean_token_accuracy": 0.8973170518875122,
      "num_tokens": 507700215.0,
      "step": 14594
    },
    {
      "epoch": 2.7103064066852367,
      "grad_norm": 1.500596936991836,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8871946334838867,
      "num_tokens": 507734712.0,
      "step": 14595
    },
    {
      "epoch": 2.7104921077065924,
      "grad_norm": 1.439396203714017,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8941745758056641,
      "num_tokens": 507769909.0,
      "step": 14596
    },
    {
      "epoch": 2.710677808727948,
      "grad_norm": 1.5856099237322172,
      "learning_rate": 1e-06,
      "loss": 0.3405,
      "mean_token_accuracy": 0.8817713260650635,
      "num_tokens": 507805942.0,
      "step": 14597
    },
    {
      "epoch": 2.7108635097493035,
      "grad_norm": 1.518942066353088,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.897783637046814,
      "num_tokens": 507845986.0,
      "step": 14598
    },
    {
      "epoch": 2.711049210770659,
      "grad_norm": 1.5222051758328927,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8940045237541199,
      "num_tokens": 507886883.0,
      "step": 14599
    },
    {
      "epoch": 2.711234911792015,
      "grad_norm": 1.650696135235532,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8792775273323059,
      "num_tokens": 507919586.0,
      "step": 14600
    },
    {
      "epoch": 2.71142061281337,
      "grad_norm": 1.545250759327967,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8983041048049927,
      "num_tokens": 507951488.0,
      "step": 14601
    },
    {
      "epoch": 2.711606313834726,
      "grad_norm": 1.4705682151608486,
      "learning_rate": 1e-06,
      "loss": 0.2894,
      "mean_token_accuracy": 0.8996739983558655,
      "num_tokens": 507989249.0,
      "step": 14602
    },
    {
      "epoch": 2.7117920148560817,
      "grad_norm": 1.488993806563401,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8849973678588867,
      "num_tokens": 508030818.0,
      "step": 14603
    },
    {
      "epoch": 2.7119777158774374,
      "grad_norm": 1.6199720841765306,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8929762244224548,
      "num_tokens": 508061198.0,
      "step": 14604
    },
    {
      "epoch": 2.712163416898793,
      "grad_norm": 1.54257807683934,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8794025182723999,
      "num_tokens": 508097273.0,
      "step": 14605
    },
    {
      "epoch": 2.7123491179201484,
      "grad_norm": 1.660802702006212,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8878045082092285,
      "num_tokens": 508130977.0,
      "step": 14606
    },
    {
      "epoch": 2.712534818941504,
      "grad_norm": 1.5846621849394564,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8893071413040161,
      "num_tokens": 508165434.0,
      "step": 14607
    },
    {
      "epoch": 2.71272051996286,
      "grad_norm": 1.5917848272356598,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.890677809715271,
      "num_tokens": 508196275.0,
      "step": 14608
    },
    {
      "epoch": 2.712906220984215,
      "grad_norm": 1.5866327593878864,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8881804943084717,
      "num_tokens": 508232904.0,
      "step": 14609
    },
    {
      "epoch": 2.713091922005571,
      "grad_norm": 1.55392615399106,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8888304233551025,
      "num_tokens": 508268336.0,
      "step": 14610
    },
    {
      "epoch": 2.7132776230269267,
      "grad_norm": 1.5275089938486344,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8882285356521606,
      "num_tokens": 508302960.0,
      "step": 14611
    },
    {
      "epoch": 2.7134633240482824,
      "grad_norm": 1.6813214750883116,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8801469206809998,
      "num_tokens": 508336488.0,
      "step": 14612
    },
    {
      "epoch": 2.713649025069638,
      "grad_norm": 1.4533243729253185,
      "learning_rate": 1e-06,
      "loss": 0.3083,
      "mean_token_accuracy": 0.8898830413818359,
      "num_tokens": 508376691.0,
      "step": 14613
    },
    {
      "epoch": 2.7138347260909934,
      "grad_norm": 1.7861539662882882,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8884367346763611,
      "num_tokens": 508402372.0,
      "step": 14614
    },
    {
      "epoch": 2.714020427112349,
      "grad_norm": 1.5600595115633762,
      "learning_rate": 1e-06,
      "loss": 0.2671,
      "mean_token_accuracy": 0.9036544561386108,
      "num_tokens": 508434550.0,
      "step": 14615
    },
    {
      "epoch": 2.714206128133705,
      "grad_norm": 1.606954155799089,
      "learning_rate": 1e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.8662563562393188,
      "num_tokens": 508470760.0,
      "step": 14616
    },
    {
      "epoch": 2.71439182915506,
      "grad_norm": 1.5751018489183082,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8875090479850769,
      "num_tokens": 508505747.0,
      "step": 14617
    },
    {
      "epoch": 2.714577530176416,
      "grad_norm": 1.6512988468474636,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.880627453327179,
      "num_tokens": 508540006.0,
      "step": 14618
    },
    {
      "epoch": 2.7147632311977716,
      "grad_norm": 1.6878667707660115,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8926540613174438,
      "num_tokens": 508574050.0,
      "step": 14619
    },
    {
      "epoch": 2.7149489322191274,
      "grad_norm": 1.490600684675005,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8856701254844666,
      "num_tokens": 508610483.0,
      "step": 14620
    },
    {
      "epoch": 2.715134633240483,
      "grad_norm": 1.6649383880303674,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.886227011680603,
      "num_tokens": 508640118.0,
      "step": 14621
    },
    {
      "epoch": 2.7153203342618384,
      "grad_norm": 1.643113658751197,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8718448877334595,
      "num_tokens": 508676097.0,
      "step": 14622
    },
    {
      "epoch": 2.715506035283194,
      "grad_norm": 1.4981792961678233,
      "learning_rate": 1e-06,
      "loss": 0.2823,
      "mean_token_accuracy": 0.8965117931365967,
      "num_tokens": 508708700.0,
      "step": 14623
    },
    {
      "epoch": 2.7156917363045494,
      "grad_norm": 1.5039371277962572,
      "learning_rate": 1e-06,
      "loss": 0.3062,
      "mean_token_accuracy": 0.8915938138961792,
      "num_tokens": 508746671.0,
      "step": 14624
    },
    {
      "epoch": 2.715877437325905,
      "grad_norm": 1.5145976214731882,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8892190456390381,
      "num_tokens": 508785664.0,
      "step": 14625
    },
    {
      "epoch": 2.716063138347261,
      "grad_norm": 1.699375190890403,
      "learning_rate": 1e-06,
      "loss": 0.3612,
      "mean_token_accuracy": 0.8743913173675537,
      "num_tokens": 508822101.0,
      "step": 14626
    },
    {
      "epoch": 2.7162488393686166,
      "grad_norm": 1.5091769225413354,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.896944522857666,
      "num_tokens": 508858800.0,
      "step": 14627
    },
    {
      "epoch": 2.7164345403899723,
      "grad_norm": 1.5461824147087488,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8786641359329224,
      "num_tokens": 508896355.0,
      "step": 14628
    },
    {
      "epoch": 2.7166202414113276,
      "grad_norm": 1.6805026475455533,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8802403211593628,
      "num_tokens": 508926695.0,
      "step": 14629
    },
    {
      "epoch": 2.7168059424326834,
      "grad_norm": 1.5919231170199548,
      "learning_rate": 1e-06,
      "loss": 0.2854,
      "mean_token_accuracy": 0.8987374305725098,
      "num_tokens": 508957214.0,
      "step": 14630
    },
    {
      "epoch": 2.716991643454039,
      "grad_norm": 1.5038898236677312,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8926064968109131,
      "num_tokens": 508991982.0,
      "step": 14631
    },
    {
      "epoch": 2.7171773444753944,
      "grad_norm": 1.7728417086132406,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.879930853843689,
      "num_tokens": 509025352.0,
      "step": 14632
    },
    {
      "epoch": 2.71736304549675,
      "grad_norm": 1.627887944080813,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8742799758911133,
      "num_tokens": 509057792.0,
      "step": 14633
    },
    {
      "epoch": 2.717548746518106,
      "grad_norm": 1.6899539869566322,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8800035715103149,
      "num_tokens": 509095525.0,
      "step": 14634
    },
    {
      "epoch": 2.7177344475394616,
      "grad_norm": 1.6741176126009298,
      "learning_rate": 1e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.8781540393829346,
      "num_tokens": 509128234.0,
      "step": 14635
    },
    {
      "epoch": 2.7179201485608173,
      "grad_norm": 1.5180228350593916,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8793197274208069,
      "num_tokens": 509169212.0,
      "step": 14636
    },
    {
      "epoch": 2.7181058495821726,
      "grad_norm": 1.6904785776693552,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8867901563644409,
      "num_tokens": 509203286.0,
      "step": 14637
    },
    {
      "epoch": 2.7182915506035283,
      "grad_norm": 1.6451693013127786,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.8838263154029846,
      "num_tokens": 509236359.0,
      "step": 14638
    },
    {
      "epoch": 2.718477251624884,
      "grad_norm": 1.5321811200592348,
      "learning_rate": 1e-06,
      "loss": 0.2739,
      "mean_token_accuracy": 0.9010366201400757,
      "num_tokens": 509268091.0,
      "step": 14639
    },
    {
      "epoch": 2.7186629526462394,
      "grad_norm": 1.4707345465439368,
      "learning_rate": 1e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.865182638168335,
      "num_tokens": 509310277.0,
      "step": 14640
    },
    {
      "epoch": 2.718848653667595,
      "grad_norm": 1.710708510190449,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8839863538742065,
      "num_tokens": 509341500.0,
      "step": 14641
    },
    {
      "epoch": 2.719034354688951,
      "grad_norm": 1.5493939800798642,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8729591369628906,
      "num_tokens": 509379007.0,
      "step": 14642
    },
    {
      "epoch": 2.7192200557103066,
      "grad_norm": 1.5756063094022905,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.889742374420166,
      "num_tokens": 509414137.0,
      "step": 14643
    },
    {
      "epoch": 2.7194057567316623,
      "grad_norm": 1.6395986158441125,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8989735841751099,
      "num_tokens": 509441935.0,
      "step": 14644
    },
    {
      "epoch": 2.7195914577530176,
      "grad_norm": 1.489105587983748,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8726671934127808,
      "num_tokens": 509481287.0,
      "step": 14645
    },
    {
      "epoch": 2.7197771587743733,
      "grad_norm": 1.4626760960106,
      "learning_rate": 1e-06,
      "loss": 0.3378,
      "mean_token_accuracy": 0.8845824003219604,
      "num_tokens": 509526599.0,
      "step": 14646
    },
    {
      "epoch": 2.7199628597957286,
      "grad_norm": 1.6197063639231297,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8852415084838867,
      "num_tokens": 509559968.0,
      "step": 14647
    },
    {
      "epoch": 2.7201485608170843,
      "grad_norm": 1.7271340150609165,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8839637637138367,
      "num_tokens": 509588419.0,
      "step": 14648
    },
    {
      "epoch": 2.72033426183844,
      "grad_norm": 1.4949147005489503,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8875765800476074,
      "num_tokens": 509624169.0,
      "step": 14649
    },
    {
      "epoch": 2.720519962859796,
      "grad_norm": 1.5606089494107602,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8910735249519348,
      "num_tokens": 509656622.0,
      "step": 14650
    },
    {
      "epoch": 2.7207056638811515,
      "grad_norm": 1.6289250348422843,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.89094078540802,
      "num_tokens": 509691511.0,
      "step": 14651
    },
    {
      "epoch": 2.720891364902507,
      "grad_norm": 1.4492524386916823,
      "learning_rate": 1e-06,
      "loss": 0.2894,
      "mean_token_accuracy": 0.8958992958068848,
      "num_tokens": 509729774.0,
      "step": 14652
    },
    {
      "epoch": 2.7210770659238626,
      "grad_norm": 1.4946931423322711,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8920366168022156,
      "num_tokens": 509768098.0,
      "step": 14653
    },
    {
      "epoch": 2.7212627669452183,
      "grad_norm": 1.4188206364634477,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8943472504615784,
      "num_tokens": 509809130.0,
      "step": 14654
    },
    {
      "epoch": 2.7214484679665736,
      "grad_norm": 1.4605975055751572,
      "learning_rate": 1e-06,
      "loss": 0.2863,
      "mean_token_accuracy": 0.8984719514846802,
      "num_tokens": 509847782.0,
      "step": 14655
    },
    {
      "epoch": 2.7216341689879293,
      "grad_norm": 1.6423090047930886,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.876879096031189,
      "num_tokens": 509882250.0,
      "step": 14656
    },
    {
      "epoch": 2.721819870009285,
      "grad_norm": 1.6157494410464077,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8778802156448364,
      "num_tokens": 509917265.0,
      "step": 14657
    },
    {
      "epoch": 2.722005571030641,
      "grad_norm": 1.67505578305965,
      "learning_rate": 1e-06,
      "loss": 0.3451,
      "mean_token_accuracy": 0.8803244233131409,
      "num_tokens": 509946467.0,
      "step": 14658
    },
    {
      "epoch": 2.7221912720519965,
      "grad_norm": 1.5369065970175098,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8833085894584656,
      "num_tokens": 509984614.0,
      "step": 14659
    },
    {
      "epoch": 2.722376973073352,
      "grad_norm": 1.7682316936760092,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8885544538497925,
      "num_tokens": 510016370.0,
      "step": 14660
    },
    {
      "epoch": 2.7225626740947075,
      "grad_norm": 1.6327000017638882,
      "learning_rate": 1e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8778795599937439,
      "num_tokens": 510048132.0,
      "step": 14661
    },
    {
      "epoch": 2.7227483751160633,
      "grad_norm": 1.743522281471633,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8895456790924072,
      "num_tokens": 510073001.0,
      "step": 14662
    },
    {
      "epoch": 2.7229340761374186,
      "grad_norm": 1.7087728302515623,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8851286172866821,
      "num_tokens": 510104904.0,
      "step": 14663
    },
    {
      "epoch": 2.7231197771587743,
      "grad_norm": 1.4916092519945534,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8903746604919434,
      "num_tokens": 510142545.0,
      "step": 14664
    },
    {
      "epoch": 2.72330547818013,
      "grad_norm": 1.6074131906078784,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8783986568450928,
      "num_tokens": 510176154.0,
      "step": 14665
    },
    {
      "epoch": 2.7234911792014858,
      "grad_norm": 1.5456761038827052,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8823366165161133,
      "num_tokens": 510212479.0,
      "step": 14666
    },
    {
      "epoch": 2.7236768802228415,
      "grad_norm": 1.606179371308962,
      "learning_rate": 1e-06,
      "loss": 0.3576,
      "mean_token_accuracy": 0.8741335868835449,
      "num_tokens": 510250620.0,
      "step": 14667
    },
    {
      "epoch": 2.723862581244197,
      "grad_norm": 1.4959422461753646,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.890479326248169,
      "num_tokens": 510289574.0,
      "step": 14668
    },
    {
      "epoch": 2.7240482822655525,
      "grad_norm": 1.775428153460011,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8966569900512695,
      "num_tokens": 510313618.0,
      "step": 14669
    },
    {
      "epoch": 2.724233983286908,
      "grad_norm": 1.68583819723759,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8977177143096924,
      "num_tokens": 510341472.0,
      "step": 14670
    },
    {
      "epoch": 2.7244196843082635,
      "grad_norm": 1.9465582599174138,
      "learning_rate": 1e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8700145483016968,
      "num_tokens": 510367419.0,
      "step": 14671
    },
    {
      "epoch": 2.7246053853296193,
      "grad_norm": 1.369861521851377,
      "learning_rate": 1e-06,
      "loss": 0.2834,
      "mean_token_accuracy": 0.8995531797409058,
      "num_tokens": 510407903.0,
      "step": 14672
    },
    {
      "epoch": 2.724791086350975,
      "grad_norm": 1.6829756968918423,
      "learning_rate": 1e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8699575662612915,
      "num_tokens": 510442853.0,
      "step": 14673
    },
    {
      "epoch": 2.7249767873723307,
      "grad_norm": 1.5603693128656801,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8886820077896118,
      "num_tokens": 510477356.0,
      "step": 14674
    },
    {
      "epoch": 2.725162488393686,
      "grad_norm": 1.7588041846074876,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.8919584155082703,
      "num_tokens": 510502808.0,
      "step": 14675
    },
    {
      "epoch": 2.7253481894150418,
      "grad_norm": 1.580787744514973,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8922003507614136,
      "num_tokens": 510537963.0,
      "step": 14676
    },
    {
      "epoch": 2.7255338904363975,
      "grad_norm": 1.630702038021641,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.887719988822937,
      "num_tokens": 510570937.0,
      "step": 14677
    },
    {
      "epoch": 2.7257195914577528,
      "grad_norm": 1.58584183604088,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8961770534515381,
      "num_tokens": 510602304.0,
      "step": 14678
    },
    {
      "epoch": 2.7259052924791085,
      "grad_norm": 1.6238648640510385,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8922414183616638,
      "num_tokens": 510634029.0,
      "step": 14679
    },
    {
      "epoch": 2.7260909935004642,
      "grad_norm": 1.5828616409375462,
      "learning_rate": 1e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.8876702785491943,
      "num_tokens": 510670004.0,
      "step": 14680
    },
    {
      "epoch": 2.72627669452182,
      "grad_norm": 1.6270252604340871,
      "learning_rate": 1e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8798041939735413,
      "num_tokens": 510703292.0,
      "step": 14681
    },
    {
      "epoch": 2.7264623955431757,
      "grad_norm": 1.5597022110948588,
      "learning_rate": 1e-06,
      "loss": 0.2729,
      "mean_token_accuracy": 0.904798150062561,
      "num_tokens": 510732923.0,
      "step": 14682
    },
    {
      "epoch": 2.726648096564531,
      "grad_norm": 1.7053983338036527,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8809118270874023,
      "num_tokens": 510764094.0,
      "step": 14683
    },
    {
      "epoch": 2.7268337975858867,
      "grad_norm": 1.7281233987899558,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8934698104858398,
      "num_tokens": 510788652.0,
      "step": 14684
    },
    {
      "epoch": 2.7270194986072425,
      "grad_norm": 1.5864365465915617,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8856644630432129,
      "num_tokens": 510823592.0,
      "step": 14685
    },
    {
      "epoch": 2.7272051996285978,
      "grad_norm": 1.592206569820162,
      "learning_rate": 1e-06,
      "loss": 0.3035,
      "mean_token_accuracy": 0.8914192914962769,
      "num_tokens": 510853858.0,
      "step": 14686
    },
    {
      "epoch": 2.7273909006499535,
      "grad_norm": 1.6905786019188278,
      "learning_rate": 1e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8746265172958374,
      "num_tokens": 510884203.0,
      "step": 14687
    },
    {
      "epoch": 2.727576601671309,
      "grad_norm": 1.529394013395141,
      "learning_rate": 1e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8768065571784973,
      "num_tokens": 510922959.0,
      "step": 14688
    },
    {
      "epoch": 2.727762302692665,
      "grad_norm": 1.5545868671653338,
      "learning_rate": 1e-06,
      "loss": 0.3191,
      "mean_token_accuracy": 0.890238881111145,
      "num_tokens": 510957034.0,
      "step": 14689
    },
    {
      "epoch": 2.7279480037140207,
      "grad_norm": 1.6995649653614648,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8840652108192444,
      "num_tokens": 510991563.0,
      "step": 14690
    },
    {
      "epoch": 2.728133704735376,
      "grad_norm": 1.4751200492442487,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8793085813522339,
      "num_tokens": 511032812.0,
      "step": 14691
    },
    {
      "epoch": 2.7283194057567317,
      "grad_norm": 1.39686377192942,
      "learning_rate": 1e-06,
      "loss": 0.2735,
      "mean_token_accuracy": 0.9030174016952515,
      "num_tokens": 511067500.0,
      "step": 14692
    },
    {
      "epoch": 2.7285051067780874,
      "grad_norm": 1.639422136858283,
      "learning_rate": 1e-06,
      "loss": 0.2935,
      "mean_token_accuracy": 0.8959627747535706,
      "num_tokens": 511096444.0,
      "step": 14693
    },
    {
      "epoch": 2.7286908077994427,
      "grad_norm": 1.4550562608251905,
      "learning_rate": 1e-06,
      "loss": 0.2842,
      "mean_token_accuracy": 0.8982201218605042,
      "num_tokens": 511133668.0,
      "step": 14694
    },
    {
      "epoch": 2.7288765088207985,
      "grad_norm": 1.5102222639934206,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8810403347015381,
      "num_tokens": 511170899.0,
      "step": 14695
    },
    {
      "epoch": 2.729062209842154,
      "grad_norm": 1.6026892520990657,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8790879249572754,
      "num_tokens": 511206588.0,
      "step": 14696
    },
    {
      "epoch": 2.72924791086351,
      "grad_norm": 1.5689432751467647,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8839467763900757,
      "num_tokens": 511240412.0,
      "step": 14697
    },
    {
      "epoch": 2.729433611884865,
      "grad_norm": 1.7106458264855982,
      "learning_rate": 1e-06,
      "loss": 0.284,
      "mean_token_accuracy": 0.8972349166870117,
      "num_tokens": 511269645.0,
      "step": 14698
    },
    {
      "epoch": 2.729619312906221,
      "grad_norm": 1.6015409736140744,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8872824907302856,
      "num_tokens": 511305553.0,
      "step": 14699
    },
    {
      "epoch": 2.7298050139275767,
      "grad_norm": 1.6741280088985055,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8916908502578735,
      "num_tokens": 511334311.0,
      "step": 14700
    },
    {
      "epoch": 2.729990714948932,
      "grad_norm": 1.5808571394103783,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8814325332641602,
      "num_tokens": 511368561.0,
      "step": 14701
    },
    {
      "epoch": 2.7301764159702877,
      "grad_norm": 1.6189986162330532,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8881291151046753,
      "num_tokens": 511399990.0,
      "step": 14702
    },
    {
      "epoch": 2.7303621169916434,
      "grad_norm": 1.5217275090331706,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8893646001815796,
      "num_tokens": 511439927.0,
      "step": 14703
    },
    {
      "epoch": 2.730547818012999,
      "grad_norm": 1.637276532740231,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8883178234100342,
      "num_tokens": 511476191.0,
      "step": 14704
    },
    {
      "epoch": 2.730733519034355,
      "grad_norm": 1.6317480965640418,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8932958245277405,
      "num_tokens": 511508360.0,
      "step": 14705
    },
    {
      "epoch": 2.73091922005571,
      "grad_norm": 1.5765064028819624,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8881842494010925,
      "num_tokens": 511542926.0,
      "step": 14706
    },
    {
      "epoch": 2.731104921077066,
      "grad_norm": 1.4976665009436378,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8849760293960571,
      "num_tokens": 511582513.0,
      "step": 14707
    },
    {
      "epoch": 2.7312906220984217,
      "grad_norm": 1.5278033078160733,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8780752420425415,
      "num_tokens": 511619247.0,
      "step": 14708
    },
    {
      "epoch": 2.731476323119777,
      "grad_norm": 1.6240650618314425,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.880549430847168,
      "num_tokens": 511653718.0,
      "step": 14709
    },
    {
      "epoch": 2.7316620241411327,
      "grad_norm": 1.5649354230884993,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8804545402526855,
      "num_tokens": 511689015.0,
      "step": 14710
    },
    {
      "epoch": 2.7318477251624884,
      "grad_norm": 1.5619702142806755,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8803401589393616,
      "num_tokens": 511727050.0,
      "step": 14711
    },
    {
      "epoch": 2.732033426183844,
      "grad_norm": 1.582830905007838,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8834341764450073,
      "num_tokens": 511768744.0,
      "step": 14712
    },
    {
      "epoch": 2.7322191272052,
      "grad_norm": 1.5105401852594067,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8804028630256653,
      "num_tokens": 511807008.0,
      "step": 14713
    },
    {
      "epoch": 2.732404828226555,
      "grad_norm": 1.4271009269079564,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8989036083221436,
      "num_tokens": 511846343.0,
      "step": 14714
    },
    {
      "epoch": 2.732590529247911,
      "grad_norm": 1.5714652752376954,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8870344758033752,
      "num_tokens": 511882026.0,
      "step": 14715
    },
    {
      "epoch": 2.7327762302692666,
      "grad_norm": 1.4340100589743017,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8947327136993408,
      "num_tokens": 511915707.0,
      "step": 14716
    },
    {
      "epoch": 2.732961931290622,
      "grad_norm": 1.682756424891423,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.8730454444885254,
      "num_tokens": 511947652.0,
      "step": 14717
    },
    {
      "epoch": 2.7331476323119777,
      "grad_norm": 1.6448270342104483,
      "learning_rate": 1e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.858811616897583,
      "num_tokens": 511982384.0,
      "step": 14718
    },
    {
      "epoch": 2.7333333333333334,
      "grad_norm": 1.3670471097823804,
      "learning_rate": 1e-06,
      "loss": 0.2642,
      "mean_token_accuracy": 0.9031592607498169,
      "num_tokens": 512021804.0,
      "step": 14719
    },
    {
      "epoch": 2.733519034354689,
      "grad_norm": 1.5309468064473424,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8895444869995117,
      "num_tokens": 512060284.0,
      "step": 14720
    },
    {
      "epoch": 2.7337047353760444,
      "grad_norm": 1.5587196790796998,
      "learning_rate": 1e-06,
      "loss": 0.325,
      "mean_token_accuracy": 0.8843017816543579,
      "num_tokens": 512097009.0,
      "step": 14721
    },
    {
      "epoch": 2.7338904363974,
      "grad_norm": 1.5638510674808896,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8907690048217773,
      "num_tokens": 512130979.0,
      "step": 14722
    },
    {
      "epoch": 2.734076137418756,
      "grad_norm": 1.3543450266238073,
      "learning_rate": 1e-06,
      "loss": 0.2868,
      "mean_token_accuracy": 0.8963192105293274,
      "num_tokens": 512170994.0,
      "step": 14723
    },
    {
      "epoch": 2.734261838440111,
      "grad_norm": 1.5413282116493328,
      "learning_rate": 1e-06,
      "loss": 0.2756,
      "mean_token_accuracy": 0.9038795232772827,
      "num_tokens": 512203689.0,
      "step": 14724
    },
    {
      "epoch": 2.734447539461467,
      "grad_norm": 1.4589972379927028,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8855710029602051,
      "num_tokens": 512245603.0,
      "step": 14725
    },
    {
      "epoch": 2.7346332404828226,
      "grad_norm": 1.761623065399415,
      "learning_rate": 1e-06,
      "loss": 0.3738,
      "mean_token_accuracy": 0.8692398071289062,
      "num_tokens": 512276279.0,
      "step": 14726
    },
    {
      "epoch": 2.7348189415041784,
      "grad_norm": 1.4870161601575735,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8898880481719971,
      "num_tokens": 512314918.0,
      "step": 14727
    },
    {
      "epoch": 2.735004642525534,
      "grad_norm": 1.5582318220366382,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8900855183601379,
      "num_tokens": 512348766.0,
      "step": 14728
    },
    {
      "epoch": 2.7351903435468894,
      "grad_norm": 1.6812256072466005,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8904415369033813,
      "num_tokens": 512380460.0,
      "step": 14729
    },
    {
      "epoch": 2.735376044568245,
      "grad_norm": 1.5453924011762499,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8873394727706909,
      "num_tokens": 512415346.0,
      "step": 14730
    },
    {
      "epoch": 2.735561745589601,
      "grad_norm": 1.5342773397511487,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.887020468711853,
      "num_tokens": 512452458.0,
      "step": 14731
    },
    {
      "epoch": 2.735747446610956,
      "grad_norm": 1.3789483989631108,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8920947313308716,
      "num_tokens": 512493488.0,
      "step": 14732
    },
    {
      "epoch": 2.735933147632312,
      "grad_norm": 1.4543553751265017,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.892187237739563,
      "num_tokens": 512530493.0,
      "step": 14733
    },
    {
      "epoch": 2.7361188486536676,
      "grad_norm": 1.5956243041216098,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8868327140808105,
      "num_tokens": 512568112.0,
      "step": 14734
    },
    {
      "epoch": 2.7363045496750233,
      "grad_norm": 1.455377921426204,
      "learning_rate": 1e-06,
      "loss": 0.2964,
      "mean_token_accuracy": 0.8933874368667603,
      "num_tokens": 512606860.0,
      "step": 14735
    },
    {
      "epoch": 2.736490250696379,
      "grad_norm": 1.4637347204900057,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8894434571266174,
      "num_tokens": 512645113.0,
      "step": 14736
    },
    {
      "epoch": 2.7366759517177344,
      "grad_norm": 1.5249781318183404,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8874416351318359,
      "num_tokens": 512684611.0,
      "step": 14737
    },
    {
      "epoch": 2.73686165273909,
      "grad_norm": 1.6639374498305377,
      "learning_rate": 1e-06,
      "loss": 0.3648,
      "mean_token_accuracy": 0.8759896159172058,
      "num_tokens": 512716905.0,
      "step": 14738
    },
    {
      "epoch": 2.737047353760446,
      "grad_norm": 1.8298693734630929,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8825212121009827,
      "num_tokens": 512744334.0,
      "step": 14739
    },
    {
      "epoch": 2.737233054781801,
      "grad_norm": 1.8337678868443654,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8778014183044434,
      "num_tokens": 512769925.0,
      "step": 14740
    },
    {
      "epoch": 2.737418755803157,
      "grad_norm": 1.4242175774051058,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8967628479003906,
      "num_tokens": 512808022.0,
      "step": 14741
    },
    {
      "epoch": 2.7376044568245126,
      "grad_norm": 1.4901560649389234,
      "learning_rate": 1e-06,
      "loss": 0.2775,
      "mean_token_accuracy": 0.9018873572349548,
      "num_tokens": 512841434.0,
      "step": 14742
    },
    {
      "epoch": 2.7377901578458683,
      "grad_norm": 1.5487001133199814,
      "learning_rate": 1e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8781061768531799,
      "num_tokens": 512879408.0,
      "step": 14743
    },
    {
      "epoch": 2.7379758588672236,
      "grad_norm": 1.388364493506992,
      "learning_rate": 1e-06,
      "loss": 0.3072,
      "mean_token_accuracy": 0.8922332525253296,
      "num_tokens": 512920992.0,
      "step": 14744
    },
    {
      "epoch": 2.7381615598885793,
      "grad_norm": 1.4764404096807344,
      "learning_rate": 1e-06,
      "loss": 0.2709,
      "mean_token_accuracy": 0.9033129811286926,
      "num_tokens": 512957104.0,
      "step": 14745
    },
    {
      "epoch": 2.738347260909935,
      "grad_norm": 1.523776850578979,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8848705291748047,
      "num_tokens": 512993476.0,
      "step": 14746
    },
    {
      "epoch": 2.7385329619312904,
      "grad_norm": 1.4945315799793746,
      "learning_rate": 1e-06,
      "loss": 0.3012,
      "mean_token_accuracy": 0.8932712078094482,
      "num_tokens": 513027449.0,
      "step": 14747
    },
    {
      "epoch": 2.738718662952646,
      "grad_norm": 1.6875405099374057,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.8755190372467041,
      "num_tokens": 513057607.0,
      "step": 14748
    },
    {
      "epoch": 2.738904363974002,
      "grad_norm": 1.6747660684861616,
      "learning_rate": 1e-06,
      "loss": 0.3685,
      "mean_token_accuracy": 0.8678247332572937,
      "num_tokens": 513089736.0,
      "step": 14749
    },
    {
      "epoch": 2.7390900649953576,
      "grad_norm": 1.5153205113527515,
      "learning_rate": 1e-06,
      "loss": 0.3619,
      "mean_token_accuracy": 0.87309730052948,
      "num_tokens": 513133285.0,
      "step": 14750
    },
    {
      "epoch": 2.7392757660167133,
      "grad_norm": 1.6287736686339223,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8950287699699402,
      "num_tokens": 513162827.0,
      "step": 14751
    },
    {
      "epoch": 2.7394614670380686,
      "grad_norm": 1.6050921560129898,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.893875002861023,
      "num_tokens": 513194356.0,
      "step": 14752
    },
    {
      "epoch": 2.7396471680594243,
      "grad_norm": 1.541924630095298,
      "learning_rate": 1e-06,
      "loss": 0.295,
      "mean_token_accuracy": 0.8935348391532898,
      "num_tokens": 513228306.0,
      "step": 14753
    },
    {
      "epoch": 2.73983286908078,
      "grad_norm": 1.6244099749115153,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.879241943359375,
      "num_tokens": 513264119.0,
      "step": 14754
    },
    {
      "epoch": 2.7400185701021353,
      "grad_norm": 1.6651854516639502,
      "learning_rate": 1e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8780556917190552,
      "num_tokens": 513299532.0,
      "step": 14755
    },
    {
      "epoch": 2.740204271123491,
      "grad_norm": 1.5633230428491718,
      "learning_rate": 1e-06,
      "loss": 0.3469,
      "mean_token_accuracy": 0.8822603225708008,
      "num_tokens": 513336875.0,
      "step": 14756
    },
    {
      "epoch": 2.740389972144847,
      "grad_norm": 1.4306768437344097,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8822188377380371,
      "num_tokens": 513376764.0,
      "step": 14757
    },
    {
      "epoch": 2.7405756731662025,
      "grad_norm": 1.6723239929888793,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8819960355758667,
      "num_tokens": 513408701.0,
      "step": 14758
    },
    {
      "epoch": 2.7407613741875583,
      "grad_norm": 1.6694068281897032,
      "learning_rate": 1e-06,
      "loss": 0.3195,
      "mean_token_accuracy": 0.8877555131912231,
      "num_tokens": 513438971.0,
      "step": 14759
    },
    {
      "epoch": 2.7409470752089136,
      "grad_norm": 1.4812703380436847,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8817488551139832,
      "num_tokens": 513480266.0,
      "step": 14760
    },
    {
      "epoch": 2.7411327762302693,
      "grad_norm": 1.6595304833486537,
      "learning_rate": 1e-06,
      "loss": 0.2778,
      "mean_token_accuracy": 0.8996931314468384,
      "num_tokens": 513507580.0,
      "step": 14761
    },
    {
      "epoch": 2.741318477251625,
      "grad_norm": 1.5409748772129015,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8744125962257385,
      "num_tokens": 513544586.0,
      "step": 14762
    },
    {
      "epoch": 2.7415041782729803,
      "grad_norm": 1.3965414612896623,
      "learning_rate": 1e-06,
      "loss": 0.2867,
      "mean_token_accuracy": 0.9001219272613525,
      "num_tokens": 513583654.0,
      "step": 14763
    },
    {
      "epoch": 2.741689879294336,
      "grad_norm": 1.4662425756321236,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8793350458145142,
      "num_tokens": 513627390.0,
      "step": 14764
    },
    {
      "epoch": 2.741875580315692,
      "grad_norm": 1.49731213953407,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8852897882461548,
      "num_tokens": 513664102.0,
      "step": 14765
    },
    {
      "epoch": 2.7420612813370475,
      "grad_norm": 1.647381106980985,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8791738748550415,
      "num_tokens": 513697973.0,
      "step": 14766
    },
    {
      "epoch": 2.742246982358403,
      "grad_norm": 1.5263486828965238,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8920726180076599,
      "num_tokens": 513733819.0,
      "step": 14767
    },
    {
      "epoch": 2.7424326833797585,
      "grad_norm": 1.6728244055650363,
      "learning_rate": 1e-06,
      "loss": 0.296,
      "mean_token_accuracy": 0.8960697650909424,
      "num_tokens": 513764062.0,
      "step": 14768
    },
    {
      "epoch": 2.7426183844011143,
      "grad_norm": 1.7063373222387765,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8742295503616333,
      "num_tokens": 513798212.0,
      "step": 14769
    },
    {
      "epoch": 2.7428040854224696,
      "grad_norm": 1.4203840586244947,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8889753818511963,
      "num_tokens": 513842145.0,
      "step": 14770
    },
    {
      "epoch": 2.7429897864438253,
      "grad_norm": 1.518309502278358,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8875019550323486,
      "num_tokens": 513879565.0,
      "step": 14771
    },
    {
      "epoch": 2.743175487465181,
      "grad_norm": 1.6224071623575844,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8768744468688965,
      "num_tokens": 513914452.0,
      "step": 14772
    },
    {
      "epoch": 2.7433611884865368,
      "grad_norm": 1.5012285995625583,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8884456753730774,
      "num_tokens": 513950884.0,
      "step": 14773
    },
    {
      "epoch": 2.7435468895078925,
      "grad_norm": 1.692067453062254,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8779178857803345,
      "num_tokens": 513982421.0,
      "step": 14774
    },
    {
      "epoch": 2.743732590529248,
      "grad_norm": 1.404265636241937,
      "learning_rate": 1e-06,
      "loss": 0.2659,
      "mean_token_accuracy": 0.9032657146453857,
      "num_tokens": 514017567.0,
      "step": 14775
    },
    {
      "epoch": 2.7439182915506035,
      "grad_norm": 1.5373064505797394,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8910306692123413,
      "num_tokens": 514052193.0,
      "step": 14776
    },
    {
      "epoch": 2.7441039925719592,
      "grad_norm": 1.6676689890662295,
      "learning_rate": 1e-06,
      "loss": 0.2878,
      "mean_token_accuracy": 0.8986961841583252,
      "num_tokens": 514085167.0,
      "step": 14777
    },
    {
      "epoch": 2.7442896935933145,
      "grad_norm": 1.4917171875143287,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8869912624359131,
      "num_tokens": 514123422.0,
      "step": 14778
    },
    {
      "epoch": 2.7444753946146703,
      "grad_norm": 1.4580706304209698,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8902415037155151,
      "num_tokens": 514160229.0,
      "step": 14779
    },
    {
      "epoch": 2.744661095636026,
      "grad_norm": 1.845098473931634,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.887572169303894,
      "num_tokens": 514183941.0,
      "step": 14780
    },
    {
      "epoch": 2.7448467966573817,
      "grad_norm": 1.5783445476974305,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8867438435554504,
      "num_tokens": 514218485.0,
      "step": 14781
    },
    {
      "epoch": 2.7450324976787375,
      "grad_norm": 1.689330206208329,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8854573965072632,
      "num_tokens": 514247811.0,
      "step": 14782
    },
    {
      "epoch": 2.7452181987000928,
      "grad_norm": 1.467165940321173,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8867380619049072,
      "num_tokens": 514291136.0,
      "step": 14783
    },
    {
      "epoch": 2.7454038997214485,
      "grad_norm": 1.4225582319491321,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8960431218147278,
      "num_tokens": 514329002.0,
      "step": 14784
    },
    {
      "epoch": 2.7455896007428042,
      "grad_norm": 1.550700183884696,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.8832102417945862,
      "num_tokens": 514364416.0,
      "step": 14785
    },
    {
      "epoch": 2.7457753017641595,
      "grad_norm": 1.4805240785922498,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8747823238372803,
      "num_tokens": 514405868.0,
      "step": 14786
    },
    {
      "epoch": 2.7459610027855152,
      "grad_norm": 1.6696785774355196,
      "learning_rate": 1e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.8735223412513733,
      "num_tokens": 514442160.0,
      "step": 14787
    },
    {
      "epoch": 2.746146703806871,
      "grad_norm": 1.5991583686618702,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8840256929397583,
      "num_tokens": 514476640.0,
      "step": 14788
    },
    {
      "epoch": 2.7463324048282267,
      "grad_norm": 1.5210653836770283,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8944479823112488,
      "num_tokens": 514514904.0,
      "step": 14789
    },
    {
      "epoch": 2.7465181058495824,
      "grad_norm": 1.5384126135626148,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8840165138244629,
      "num_tokens": 514553065.0,
      "step": 14790
    },
    {
      "epoch": 2.7467038068709377,
      "grad_norm": 1.552639104416961,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8739667534828186,
      "num_tokens": 514593784.0,
      "step": 14791
    },
    {
      "epoch": 2.7468895078922935,
      "grad_norm": 1.5907695432979623,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8902475833892822,
      "num_tokens": 514627100.0,
      "step": 14792
    },
    {
      "epoch": 2.7470752089136488,
      "grad_norm": 1.5299498531711113,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8926092386245728,
      "num_tokens": 514663035.0,
      "step": 14793
    },
    {
      "epoch": 2.7472609099350045,
      "grad_norm": 1.456731009372436,
      "learning_rate": 1e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8876381516456604,
      "num_tokens": 514702996.0,
      "step": 14794
    },
    {
      "epoch": 2.7474466109563602,
      "grad_norm": 1.6069081995154069,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8775758147239685,
      "num_tokens": 514736278.0,
      "step": 14795
    },
    {
      "epoch": 2.747632311977716,
      "grad_norm": 1.5308284345163135,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8834944367408752,
      "num_tokens": 514771780.0,
      "step": 14796
    },
    {
      "epoch": 2.7478180129990717,
      "grad_norm": 1.600523935659301,
      "learning_rate": 1e-06,
      "loss": 0.3167,
      "mean_token_accuracy": 0.8892472386360168,
      "num_tokens": 514805309.0,
      "step": 14797
    },
    {
      "epoch": 2.748003714020427,
      "grad_norm": 1.5120473172814397,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.88860023021698,
      "num_tokens": 514840666.0,
      "step": 14798
    },
    {
      "epoch": 2.7481894150417827,
      "grad_norm": 1.555951000428957,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8891441226005554,
      "num_tokens": 514873229.0,
      "step": 14799
    },
    {
      "epoch": 2.7483751160631384,
      "grad_norm": 1.4908932349412651,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8943967819213867,
      "num_tokens": 514909518.0,
      "step": 14800
    },
    {
      "epoch": 2.7485608170844937,
      "grad_norm": 1.4847971008510876,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8954206705093384,
      "num_tokens": 514943110.0,
      "step": 14801
    },
    {
      "epoch": 2.7487465181058495,
      "grad_norm": 1.8188555829271693,
      "learning_rate": 1e-06,
      "loss": 0.3589,
      "mean_token_accuracy": 0.8763439059257507,
      "num_tokens": 514973267.0,
      "step": 14802
    },
    {
      "epoch": 2.748932219127205,
      "grad_norm": 1.5834367289362699,
      "learning_rate": 1e-06,
      "loss": 0.3379,
      "mean_token_accuracy": 0.882857620716095,
      "num_tokens": 515008734.0,
      "step": 14803
    },
    {
      "epoch": 2.749117920148561,
      "grad_norm": 1.607440578715436,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8768247961997986,
      "num_tokens": 515044538.0,
      "step": 14804
    },
    {
      "epoch": 2.7493036211699167,
      "grad_norm": 1.5213925600959222,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8936481475830078,
      "num_tokens": 515078693.0,
      "step": 14805
    },
    {
      "epoch": 2.749489322191272,
      "grad_norm": 1.5607232770585246,
      "learning_rate": 1e-06,
      "loss": 0.2706,
      "mean_token_accuracy": 0.90251624584198,
      "num_tokens": 515114118.0,
      "step": 14806
    },
    {
      "epoch": 2.7496750232126277,
      "grad_norm": 1.5951126314718143,
      "learning_rate": 1e-06,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8779033422470093,
      "num_tokens": 515149233.0,
      "step": 14807
    },
    {
      "epoch": 2.7498607242339834,
      "grad_norm": 1.5554503889666125,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8826466202735901,
      "num_tokens": 515186772.0,
      "step": 14808
    },
    {
      "epoch": 2.7500464252553387,
      "grad_norm": 1.563447538786591,
      "learning_rate": 1e-06,
      "loss": 0.2695,
      "mean_token_accuracy": 0.9041385054588318,
      "num_tokens": 515217061.0,
      "step": 14809
    },
    {
      "epoch": 2.7502321262766944,
      "grad_norm": 1.5124767995107065,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.8948167562484741,
      "num_tokens": 515251390.0,
      "step": 14810
    },
    {
      "epoch": 2.75041782729805,
      "grad_norm": 1.5920049809573482,
      "learning_rate": 1e-06,
      "loss": 0.3294,
      "mean_token_accuracy": 0.8862364292144775,
      "num_tokens": 515284377.0,
      "step": 14811
    },
    {
      "epoch": 2.750603528319406,
      "grad_norm": 1.6830897053790452,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.8765824437141418,
      "num_tokens": 515318599.0,
      "step": 14812
    },
    {
      "epoch": 2.7507892293407616,
      "grad_norm": 1.5779176200856173,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8887242674827576,
      "num_tokens": 515353872.0,
      "step": 14813
    },
    {
      "epoch": 2.750974930362117,
      "grad_norm": 1.5610232661165964,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8864471316337585,
      "num_tokens": 515387019.0,
      "step": 14814
    },
    {
      "epoch": 2.7511606313834727,
      "grad_norm": 1.5337122495448459,
      "learning_rate": 1e-06,
      "loss": 0.3464,
      "mean_token_accuracy": 0.8819338083267212,
      "num_tokens": 515427736.0,
      "step": 14815
    },
    {
      "epoch": 2.751346332404828,
      "grad_norm": 1.6296799892514353,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8906183242797852,
      "num_tokens": 515460242.0,
      "step": 14816
    },
    {
      "epoch": 2.7515320334261837,
      "grad_norm": 1.6010927470218501,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8794729709625244,
      "num_tokens": 515497540.0,
      "step": 14817
    },
    {
      "epoch": 2.7517177344475394,
      "grad_norm": 1.4151363814653735,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8911360502243042,
      "num_tokens": 515538938.0,
      "step": 14818
    },
    {
      "epoch": 2.751903435468895,
      "grad_norm": 1.5132671102348798,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8776400685310364,
      "num_tokens": 515576718.0,
      "step": 14819
    },
    {
      "epoch": 2.752089136490251,
      "grad_norm": 1.524380811020172,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8946360349655151,
      "num_tokens": 515613490.0,
      "step": 14820
    },
    {
      "epoch": 2.752274837511606,
      "grad_norm": 1.4507032188445508,
      "learning_rate": 1e-06,
      "loss": 0.2964,
      "mean_token_accuracy": 0.8930210471153259,
      "num_tokens": 515650966.0,
      "step": 14821
    },
    {
      "epoch": 2.752460538532962,
      "grad_norm": 1.4209134499195477,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8957487344741821,
      "num_tokens": 515688446.0,
      "step": 14822
    },
    {
      "epoch": 2.7526462395543176,
      "grad_norm": 1.558873933136668,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8813697695732117,
      "num_tokens": 515726476.0,
      "step": 14823
    },
    {
      "epoch": 2.752831940575673,
      "grad_norm": 1.4341312198547556,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8907703757286072,
      "num_tokens": 515765995.0,
      "step": 14824
    },
    {
      "epoch": 2.7530176415970287,
      "grad_norm": 1.4878820152630625,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.886558473110199,
      "num_tokens": 515806282.0,
      "step": 14825
    },
    {
      "epoch": 2.7532033426183844,
      "grad_norm": 1.6969959077335268,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8818057775497437,
      "num_tokens": 515838019.0,
      "step": 14826
    },
    {
      "epoch": 2.75338904363974,
      "grad_norm": 1.5607490993026059,
      "learning_rate": 1e-06,
      "loss": 0.3351,
      "mean_token_accuracy": 0.8819128274917603,
      "num_tokens": 515871336.0,
      "step": 14827
    },
    {
      "epoch": 2.753574744661096,
      "grad_norm": 1.448917371820308,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8823343515396118,
      "num_tokens": 515914705.0,
      "step": 14828
    },
    {
      "epoch": 2.753760445682451,
      "grad_norm": 1.6255722249543443,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8953945636749268,
      "num_tokens": 515946603.0,
      "step": 14829
    },
    {
      "epoch": 2.753946146703807,
      "grad_norm": 1.428256478223656,
      "learning_rate": 1e-06,
      "loss": 0.2761,
      "mean_token_accuracy": 0.9040594100952148,
      "num_tokens": 515982259.0,
      "step": 14830
    },
    {
      "epoch": 2.7541318477251626,
      "grad_norm": 1.5019112289754268,
      "learning_rate": 1e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8730570077896118,
      "num_tokens": 516024007.0,
      "step": 14831
    },
    {
      "epoch": 2.754317548746518,
      "grad_norm": 1.522020791218324,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8852986097335815,
      "num_tokens": 516059881.0,
      "step": 14832
    },
    {
      "epoch": 2.7545032497678736,
      "grad_norm": 1.5915952863208407,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8895000219345093,
      "num_tokens": 516094467.0,
      "step": 14833
    },
    {
      "epoch": 2.7546889507892294,
      "grad_norm": 1.708975936026952,
      "learning_rate": 1e-06,
      "loss": 0.3481,
      "mean_token_accuracy": 0.8766568899154663,
      "num_tokens": 516125543.0,
      "step": 14834
    },
    {
      "epoch": 2.754874651810585,
      "grad_norm": 1.4541669300436622,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8969743847846985,
      "num_tokens": 516165207.0,
      "step": 14835
    },
    {
      "epoch": 2.755060352831941,
      "grad_norm": 1.6018210720207646,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8670741319656372,
      "num_tokens": 516201770.0,
      "step": 14836
    },
    {
      "epoch": 2.755246053853296,
      "grad_norm": 1.5610063958402574,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8854849338531494,
      "num_tokens": 516236026.0,
      "step": 14837
    },
    {
      "epoch": 2.755431754874652,
      "grad_norm": 1.8245074183332997,
      "learning_rate": 1e-06,
      "loss": 0.3045,
      "mean_token_accuracy": 0.892900824546814,
      "num_tokens": 516260673.0,
      "step": 14838
    },
    {
      "epoch": 2.755617455896007,
      "grad_norm": 1.476716509065503,
      "learning_rate": 1e-06,
      "loss": 0.343,
      "mean_token_accuracy": 0.8801398277282715,
      "num_tokens": 516304174.0,
      "step": 14839
    },
    {
      "epoch": 2.755803156917363,
      "grad_norm": 1.5700559922010253,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.892865777015686,
      "num_tokens": 516337501.0,
      "step": 14840
    },
    {
      "epoch": 2.7559888579387186,
      "grad_norm": 1.5342065657288146,
      "learning_rate": 1e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8763629794120789,
      "num_tokens": 516382226.0,
      "step": 14841
    },
    {
      "epoch": 2.7561745589600744,
      "grad_norm": 1.68181770780861,
      "learning_rate": 1e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.8702715635299683,
      "num_tokens": 516418331.0,
      "step": 14842
    },
    {
      "epoch": 2.75636025998143,
      "grad_norm": 1.6273346182894628,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8891613483428955,
      "num_tokens": 516450169.0,
      "step": 14843
    },
    {
      "epoch": 2.7565459610027854,
      "grad_norm": 1.479900543217537,
      "learning_rate": 1e-06,
      "loss": 0.2945,
      "mean_token_accuracy": 0.898287832736969,
      "num_tokens": 516483456.0,
      "step": 14844
    },
    {
      "epoch": 2.756731662024141,
      "grad_norm": 1.5280252658464417,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.873080849647522,
      "num_tokens": 516524354.0,
      "step": 14845
    },
    {
      "epoch": 2.756917363045497,
      "grad_norm": 1.5037692908902447,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8851515054702759,
      "num_tokens": 516563230.0,
      "step": 14846
    },
    {
      "epoch": 2.757103064066852,
      "grad_norm": 1.4163301957514114,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8871535062789917,
      "num_tokens": 516603498.0,
      "step": 14847
    },
    {
      "epoch": 2.757288765088208,
      "grad_norm": 1.6311927999111804,
      "learning_rate": 1e-06,
      "loss": 0.3503,
      "mean_token_accuracy": 0.8801305294036865,
      "num_tokens": 516636502.0,
      "step": 14848
    },
    {
      "epoch": 2.7574744661095636,
      "grad_norm": 1.6078091088659041,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.8949528932571411,
      "num_tokens": 516669202.0,
      "step": 14849
    },
    {
      "epoch": 2.7576601671309193,
      "grad_norm": 1.6814297054826126,
      "learning_rate": 1e-06,
      "loss": 0.2504,
      "mean_token_accuracy": 0.9081913232803345,
      "num_tokens": 516695556.0,
      "step": 14850
    },
    {
      "epoch": 2.757845868152275,
      "grad_norm": 1.5875820058218246,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8837164044380188,
      "num_tokens": 516730518.0,
      "step": 14851
    },
    {
      "epoch": 2.7580315691736303,
      "grad_norm": 1.633017303774976,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8909584283828735,
      "num_tokens": 516761955.0,
      "step": 14852
    },
    {
      "epoch": 2.758217270194986,
      "grad_norm": 1.636443430365031,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8890514969825745,
      "num_tokens": 516794530.0,
      "step": 14853
    },
    {
      "epoch": 2.758402971216342,
      "grad_norm": 1.5577512600975603,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8966601490974426,
      "num_tokens": 516829289.0,
      "step": 14854
    },
    {
      "epoch": 2.758588672237697,
      "grad_norm": 1.7430647775918444,
      "learning_rate": 1e-06,
      "loss": 0.3703,
      "mean_token_accuracy": 0.8706806898117065,
      "num_tokens": 516858399.0,
      "step": 14855
    },
    {
      "epoch": 2.758774373259053,
      "grad_norm": 1.49679574303386,
      "learning_rate": 1e-06,
      "loss": 0.2826,
      "mean_token_accuracy": 0.8978147506713867,
      "num_tokens": 516896108.0,
      "step": 14856
    },
    {
      "epoch": 2.7589600742804086,
      "grad_norm": 1.5768821469245995,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8838956952095032,
      "num_tokens": 516933005.0,
      "step": 14857
    },
    {
      "epoch": 2.7591457753017643,
      "grad_norm": 1.5115098895689856,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8950467109680176,
      "num_tokens": 516967285.0,
      "step": 14858
    },
    {
      "epoch": 2.75933147632312,
      "grad_norm": 1.4745026401543313,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8799476623535156,
      "num_tokens": 517009264.0,
      "step": 14859
    },
    {
      "epoch": 2.7595171773444753,
      "grad_norm": 1.5777451542042076,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8775902390480042,
      "num_tokens": 517045313.0,
      "step": 14860
    },
    {
      "epoch": 2.759702878365831,
      "grad_norm": 1.5676530067969978,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8896999359130859,
      "num_tokens": 517081127.0,
      "step": 14861
    },
    {
      "epoch": 2.759888579387187,
      "grad_norm": 1.766562707255871,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8849007487297058,
      "num_tokens": 517107873.0,
      "step": 14862
    },
    {
      "epoch": 2.760074280408542,
      "grad_norm": 1.5257456921251218,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8877983093261719,
      "num_tokens": 517145472.0,
      "step": 14863
    },
    {
      "epoch": 2.760259981429898,
      "grad_norm": 1.6223304033488404,
      "learning_rate": 1e-06,
      "loss": 0.2869,
      "mean_token_accuracy": 0.8977110385894775,
      "num_tokens": 517177918.0,
      "step": 14864
    },
    {
      "epoch": 2.7604456824512535,
      "grad_norm": 1.8139161652670206,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8729408979415894,
      "num_tokens": 517209067.0,
      "step": 14865
    },
    {
      "epoch": 2.7606313834726093,
      "grad_norm": 1.6479973095672737,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8915393352508545,
      "num_tokens": 517241620.0,
      "step": 14866
    },
    {
      "epoch": 2.7608170844939646,
      "grad_norm": 1.6137014145251358,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8868705630302429,
      "num_tokens": 517274236.0,
      "step": 14867
    },
    {
      "epoch": 2.7610027855153203,
      "grad_norm": 1.6125431747231878,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8787716627120972,
      "num_tokens": 517306291.0,
      "step": 14868
    },
    {
      "epoch": 2.761188486536676,
      "grad_norm": 1.5315440428476934,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8928835391998291,
      "num_tokens": 517340626.0,
      "step": 14869
    },
    {
      "epoch": 2.7613741875580313,
      "grad_norm": 1.6944419526016294,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8846153020858765,
      "num_tokens": 517373469.0,
      "step": 14870
    },
    {
      "epoch": 2.761559888579387,
      "grad_norm": 1.6320601821456517,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8845244646072388,
      "num_tokens": 517404862.0,
      "step": 14871
    },
    {
      "epoch": 2.761745589600743,
      "grad_norm": 1.502062979101656,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8893711566925049,
      "num_tokens": 517441859.0,
      "step": 14872
    },
    {
      "epoch": 2.7619312906220985,
      "grad_norm": 1.6569474700033533,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.879253089427948,
      "num_tokens": 517475119.0,
      "step": 14873
    },
    {
      "epoch": 2.7621169916434543,
      "grad_norm": 1.516062383878239,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8916712999343872,
      "num_tokens": 517509570.0,
      "step": 14874
    },
    {
      "epoch": 2.7623026926648095,
      "grad_norm": 1.772916125438945,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8902939558029175,
      "num_tokens": 517541579.0,
      "step": 14875
    },
    {
      "epoch": 2.7624883936861653,
      "grad_norm": 1.5498967664281098,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8898782730102539,
      "num_tokens": 517573748.0,
      "step": 14876
    },
    {
      "epoch": 2.762674094707521,
      "grad_norm": 1.613441293771448,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8886688947677612,
      "num_tokens": 517608904.0,
      "step": 14877
    },
    {
      "epoch": 2.7628597957288763,
      "grad_norm": 1.5188469937227853,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8849401473999023,
      "num_tokens": 517646354.0,
      "step": 14878
    },
    {
      "epoch": 2.763045496750232,
      "grad_norm": 1.6678552214468738,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8785011172294617,
      "num_tokens": 517679275.0,
      "step": 14879
    },
    {
      "epoch": 2.7632311977715878,
      "grad_norm": 1.5478117585197124,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.884577751159668,
      "num_tokens": 517712399.0,
      "step": 14880
    },
    {
      "epoch": 2.7634168987929435,
      "grad_norm": 1.3958855421457403,
      "learning_rate": 1e-06,
      "loss": 0.2899,
      "mean_token_accuracy": 0.896175742149353,
      "num_tokens": 517751664.0,
      "step": 14881
    },
    {
      "epoch": 2.7636025998142992,
      "grad_norm": 1.651859087263892,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8836319446563721,
      "num_tokens": 517782476.0,
      "step": 14882
    },
    {
      "epoch": 2.7637883008356545,
      "grad_norm": 1.5707644449660614,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8856824636459351,
      "num_tokens": 517817188.0,
      "step": 14883
    },
    {
      "epoch": 2.7639740018570103,
      "grad_norm": 1.5098319173441173,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8967458009719849,
      "num_tokens": 517853515.0,
      "step": 14884
    },
    {
      "epoch": 2.764159702878366,
      "grad_norm": 1.5240569304231384,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8733961582183838,
      "num_tokens": 517890875.0,
      "step": 14885
    },
    {
      "epoch": 2.7643454038997213,
      "grad_norm": 1.5015353271977894,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8864168524742126,
      "num_tokens": 517928256.0,
      "step": 14886
    },
    {
      "epoch": 2.764531104921077,
      "grad_norm": 1.4878068113642169,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8817538022994995,
      "num_tokens": 517966601.0,
      "step": 14887
    },
    {
      "epoch": 2.7647168059424327,
      "grad_norm": 1.5510350550920375,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.8869379162788391,
      "num_tokens": 518001603.0,
      "step": 14888
    },
    {
      "epoch": 2.7649025069637885,
      "grad_norm": 1.5686204399377917,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.8930729031562805,
      "num_tokens": 518037030.0,
      "step": 14889
    },
    {
      "epoch": 2.7650882079851438,
      "grad_norm": 1.6430445655508925,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8951869010925293,
      "num_tokens": 518065508.0,
      "step": 14890
    },
    {
      "epoch": 2.7652739090064995,
      "grad_norm": 1.480355960728674,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8905255794525146,
      "num_tokens": 518103196.0,
      "step": 14891
    },
    {
      "epoch": 2.7654596100278552,
      "grad_norm": 1.7167648273573801,
      "learning_rate": 1e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8753623366355896,
      "num_tokens": 518138636.0,
      "step": 14892
    },
    {
      "epoch": 2.7656453110492105,
      "grad_norm": 1.666294048721503,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8808772563934326,
      "num_tokens": 518172126.0,
      "step": 14893
    },
    {
      "epoch": 2.7658310120705663,
      "grad_norm": 1.4516582247925254,
      "learning_rate": 1e-06,
      "loss": 0.3017,
      "mean_token_accuracy": 0.8912860751152039,
      "num_tokens": 518208957.0,
      "step": 14894
    },
    {
      "epoch": 2.766016713091922,
      "grad_norm": 1.553163694684053,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8875961303710938,
      "num_tokens": 518242922.0,
      "step": 14895
    },
    {
      "epoch": 2.7662024141132777,
      "grad_norm": 1.4622519504590663,
      "learning_rate": 1e-06,
      "loss": 0.3086,
      "mean_token_accuracy": 0.8913134336471558,
      "num_tokens": 518282507.0,
      "step": 14896
    },
    {
      "epoch": 2.7663881151346335,
      "grad_norm": 1.5792489585870777,
      "learning_rate": 1e-06,
      "loss": 0.3457,
      "mean_token_accuracy": 0.8772714138031006,
      "num_tokens": 518317839.0,
      "step": 14897
    },
    {
      "epoch": 2.7665738161559887,
      "grad_norm": 1.451708613219532,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8746132850646973,
      "num_tokens": 518361462.0,
      "step": 14898
    },
    {
      "epoch": 2.7667595171773445,
      "grad_norm": 1.52310467977009,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8822922706604004,
      "num_tokens": 518403856.0,
      "step": 14899
    },
    {
      "epoch": 2.7669452181987,
      "grad_norm": 1.5454271324805953,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8964064717292786,
      "num_tokens": 518439350.0,
      "step": 14900
    },
    {
      "epoch": 2.7671309192200555,
      "grad_norm": 1.4515630949119966,
      "learning_rate": 1e-06,
      "loss": 0.2751,
      "mean_token_accuracy": 0.9011268615722656,
      "num_tokens": 518475939.0,
      "step": 14901
    },
    {
      "epoch": 2.7673166202414112,
      "grad_norm": 1.555318530185866,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8885369896888733,
      "num_tokens": 518509265.0,
      "step": 14902
    },
    {
      "epoch": 2.767502321262767,
      "grad_norm": 1.60973135269603,
      "learning_rate": 1e-06,
      "loss": 0.2792,
      "mean_token_accuracy": 0.9007269144058228,
      "num_tokens": 518540809.0,
      "step": 14903
    },
    {
      "epoch": 2.7676880222841227,
      "grad_norm": 1.6308459660994725,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8888217210769653,
      "num_tokens": 518570391.0,
      "step": 14904
    },
    {
      "epoch": 2.7678737233054784,
      "grad_norm": 1.5342370758389234,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8911442756652832,
      "num_tokens": 518605373.0,
      "step": 14905
    },
    {
      "epoch": 2.7680594243268337,
      "grad_norm": 1.5485929613306686,
      "learning_rate": 1e-06,
      "loss": 0.3154,
      "mean_token_accuracy": 0.8896143436431885,
      "num_tokens": 518641488.0,
      "step": 14906
    },
    {
      "epoch": 2.7682451253481895,
      "grad_norm": 1.6896899660941138,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8865312337875366,
      "num_tokens": 518671715.0,
      "step": 14907
    },
    {
      "epoch": 2.768430826369545,
      "grad_norm": 1.465733634335551,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8882735967636108,
      "num_tokens": 518708726.0,
      "step": 14908
    },
    {
      "epoch": 2.7686165273909005,
      "grad_norm": 1.4922879405309915,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8819670081138611,
      "num_tokens": 518750049.0,
      "step": 14909
    },
    {
      "epoch": 2.768802228412256,
      "grad_norm": 1.7626775806052826,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8881257772445679,
      "num_tokens": 518778114.0,
      "step": 14910
    },
    {
      "epoch": 2.768987929433612,
      "grad_norm": 1.518149828396313,
      "learning_rate": 1e-06,
      "loss": 0.3083,
      "mean_token_accuracy": 0.8908810615539551,
      "num_tokens": 518811995.0,
      "step": 14911
    },
    {
      "epoch": 2.7691736304549677,
      "grad_norm": 1.6521524106855978,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.89228755235672,
      "num_tokens": 518842601.0,
      "step": 14912
    },
    {
      "epoch": 2.769359331476323,
      "grad_norm": 1.5140125673405278,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8880643844604492,
      "num_tokens": 518882135.0,
      "step": 14913
    },
    {
      "epoch": 2.7695450324976787,
      "grad_norm": 1.495405114760463,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.8946856260299683,
      "num_tokens": 518920438.0,
      "step": 14914
    },
    {
      "epoch": 2.7697307335190344,
      "grad_norm": 1.4961651962731588,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8910163640975952,
      "num_tokens": 518957082.0,
      "step": 14915
    },
    {
      "epoch": 2.7699164345403897,
      "grad_norm": 1.5324555626570628,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8836052417755127,
      "num_tokens": 518991404.0,
      "step": 14916
    },
    {
      "epoch": 2.7701021355617454,
      "grad_norm": 1.4945456160891473,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8966337442398071,
      "num_tokens": 519025334.0,
      "step": 14917
    },
    {
      "epoch": 2.770287836583101,
      "grad_norm": 1.3541257639933144,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8880765438079834,
      "num_tokens": 519069676.0,
      "step": 14918
    },
    {
      "epoch": 2.770473537604457,
      "grad_norm": 1.6281337250382708,
      "learning_rate": 1e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8595519661903381,
      "num_tokens": 519105471.0,
      "step": 14919
    },
    {
      "epoch": 2.7706592386258126,
      "grad_norm": 1.548716452961175,
      "learning_rate": 1e-06,
      "loss": 0.2928,
      "mean_token_accuracy": 0.8987653255462646,
      "num_tokens": 519139694.0,
      "step": 14920
    },
    {
      "epoch": 2.770844939647168,
      "grad_norm": 1.5964970431327359,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.892061710357666,
      "num_tokens": 519170708.0,
      "step": 14921
    },
    {
      "epoch": 2.7710306406685237,
      "grad_norm": 1.575178691573876,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8812738656997681,
      "num_tokens": 519205530.0,
      "step": 14922
    },
    {
      "epoch": 2.7712163416898794,
      "grad_norm": 1.5567870976065235,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8782966136932373,
      "num_tokens": 519240996.0,
      "step": 14923
    },
    {
      "epoch": 2.7714020427112347,
      "grad_norm": 1.646885858711145,
      "learning_rate": 1e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8714696168899536,
      "num_tokens": 519278065.0,
      "step": 14924
    },
    {
      "epoch": 2.7715877437325904,
      "grad_norm": 1.6510841494729194,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8865621089935303,
      "num_tokens": 519312863.0,
      "step": 14925
    },
    {
      "epoch": 2.771773444753946,
      "grad_norm": 1.6833433664998898,
      "learning_rate": 1e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8637838959693909,
      "num_tokens": 519348100.0,
      "step": 14926
    },
    {
      "epoch": 2.771959145775302,
      "grad_norm": 1.5482891226702697,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8938654661178589,
      "num_tokens": 519380783.0,
      "step": 14927
    },
    {
      "epoch": 2.7721448467966576,
      "grad_norm": 1.4943427047474442,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8961452841758728,
      "num_tokens": 519417956.0,
      "step": 14928
    },
    {
      "epoch": 2.772330547818013,
      "grad_norm": 1.609039988678518,
      "learning_rate": 1e-06,
      "loss": 0.2827,
      "mean_token_accuracy": 0.9017779231071472,
      "num_tokens": 519449316.0,
      "step": 14929
    },
    {
      "epoch": 2.7725162488393686,
      "grad_norm": 1.4501907682203639,
      "learning_rate": 1e-06,
      "loss": 0.2769,
      "mean_token_accuracy": 0.9004273414611816,
      "num_tokens": 519487276.0,
      "step": 14930
    },
    {
      "epoch": 2.7727019498607244,
      "grad_norm": 1.7714824105473364,
      "learning_rate": 1e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8737096190452576,
      "num_tokens": 519515918.0,
      "step": 14931
    },
    {
      "epoch": 2.7728876508820797,
      "grad_norm": 1.6668853058741113,
      "learning_rate": 1e-06,
      "loss": 0.3399,
      "mean_token_accuracy": 0.884116530418396,
      "num_tokens": 519548561.0,
      "step": 14932
    },
    {
      "epoch": 2.7730733519034354,
      "grad_norm": 1.4946479735182308,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.8952026963233948,
      "num_tokens": 519587265.0,
      "step": 14933
    },
    {
      "epoch": 2.773259052924791,
      "grad_norm": 1.4586779741757938,
      "learning_rate": 1e-06,
      "loss": 0.2758,
      "mean_token_accuracy": 0.8996851444244385,
      "num_tokens": 519622629.0,
      "step": 14934
    },
    {
      "epoch": 2.773444753946147,
      "grad_norm": 1.661214479179207,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8877907395362854,
      "num_tokens": 519654311.0,
      "step": 14935
    },
    {
      "epoch": 2.773630454967502,
      "grad_norm": 1.5831580182534997,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8939143419265747,
      "num_tokens": 519685698.0,
      "step": 14936
    },
    {
      "epoch": 2.773816155988858,
      "grad_norm": 1.7360937845418836,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8862824440002441,
      "num_tokens": 519716711.0,
      "step": 14937
    },
    {
      "epoch": 2.7740018570102136,
      "grad_norm": 1.6048185607837309,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8854029178619385,
      "num_tokens": 519752543.0,
      "step": 14938
    },
    {
      "epoch": 2.774187558031569,
      "grad_norm": 1.539691378719105,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.879306435585022,
      "num_tokens": 519790699.0,
      "step": 14939
    },
    {
      "epoch": 2.7743732590529246,
      "grad_norm": 1.45048677883166,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8756356239318848,
      "num_tokens": 519834723.0,
      "step": 14940
    },
    {
      "epoch": 2.7745589600742804,
      "grad_norm": 1.655977429594286,
      "learning_rate": 1e-06,
      "loss": 0.3518,
      "mean_token_accuracy": 0.8784801959991455,
      "num_tokens": 519870068.0,
      "step": 14941
    },
    {
      "epoch": 2.774744661095636,
      "grad_norm": 1.4919240120234896,
      "learning_rate": 1e-06,
      "loss": 0.3239,
      "mean_token_accuracy": 0.8870483636856079,
      "num_tokens": 519907106.0,
      "step": 14942
    },
    {
      "epoch": 2.774930362116992,
      "grad_norm": 1.4531146100319399,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8950399160385132,
      "num_tokens": 519944527.0,
      "step": 14943
    },
    {
      "epoch": 2.775116063138347,
      "grad_norm": 1.352168730705983,
      "learning_rate": 1e-06,
      "loss": 0.274,
      "mean_token_accuracy": 0.9009730219841003,
      "num_tokens": 519982935.0,
      "step": 14944
    },
    {
      "epoch": 2.775301764159703,
      "grad_norm": 1.4555082582617456,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8880491256713867,
      "num_tokens": 520022882.0,
      "step": 14945
    },
    {
      "epoch": 2.7754874651810586,
      "grad_norm": 1.5184517491986829,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8844205141067505,
      "num_tokens": 520057950.0,
      "step": 14946
    },
    {
      "epoch": 2.775673166202414,
      "grad_norm": 1.4379265809437265,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.8941047191619873,
      "num_tokens": 520097377.0,
      "step": 14947
    },
    {
      "epoch": 2.7758588672237696,
      "grad_norm": 1.6565801238554259,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8688704967498779,
      "num_tokens": 520131722.0,
      "step": 14948
    },
    {
      "epoch": 2.7760445682451254,
      "grad_norm": 1.6172284921140045,
      "learning_rate": 1e-06,
      "loss": 0.3292,
      "mean_token_accuracy": 0.8843422532081604,
      "num_tokens": 520163203.0,
      "step": 14949
    },
    {
      "epoch": 2.776230269266481,
      "grad_norm": 1.6064833786335695,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8837457895278931,
      "num_tokens": 520198221.0,
      "step": 14950
    },
    {
      "epoch": 2.776415970287837,
      "grad_norm": 1.6849543856778153,
      "learning_rate": 1e-06,
      "loss": 0.373,
      "mean_token_accuracy": 0.8711038827896118,
      "num_tokens": 520232390.0,
      "step": 14951
    },
    {
      "epoch": 2.776601671309192,
      "grad_norm": 1.4523983403432705,
      "learning_rate": 1e-06,
      "loss": 0.28,
      "mean_token_accuracy": 0.8969488143920898,
      "num_tokens": 520270513.0,
      "step": 14952
    },
    {
      "epoch": 2.776787372330548,
      "grad_norm": 1.540635080400389,
      "learning_rate": 1e-06,
      "loss": 0.3352,
      "mean_token_accuracy": 0.8822516202926636,
      "num_tokens": 520308354.0,
      "step": 14953
    },
    {
      "epoch": 2.7769730733519036,
      "grad_norm": 1.5800294582182128,
      "learning_rate": 1e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8717877864837646,
      "num_tokens": 520349416.0,
      "step": 14954
    },
    {
      "epoch": 2.777158774373259,
      "grad_norm": 1.6435328079603644,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8818187117576599,
      "num_tokens": 520381818.0,
      "step": 14955
    },
    {
      "epoch": 2.7773444753946146,
      "grad_norm": 1.5290224358265345,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8993582725524902,
      "num_tokens": 520417225.0,
      "step": 14956
    },
    {
      "epoch": 2.7775301764159703,
      "grad_norm": 1.4694163980819512,
      "learning_rate": 1e-06,
      "loss": 0.3483,
      "mean_token_accuracy": 0.8765798211097717,
      "num_tokens": 520457991.0,
      "step": 14957
    },
    {
      "epoch": 2.777715877437326,
      "grad_norm": 1.6711061660625313,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8876771926879883,
      "num_tokens": 520490339.0,
      "step": 14958
    },
    {
      "epoch": 2.777901578458682,
      "grad_norm": 1.5018987064748297,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8954559564590454,
      "num_tokens": 520524792.0,
      "step": 14959
    },
    {
      "epoch": 2.778087279480037,
      "grad_norm": 1.887039112816968,
      "learning_rate": 1e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.875401496887207,
      "num_tokens": 520550666.0,
      "step": 14960
    },
    {
      "epoch": 2.778272980501393,
      "grad_norm": 1.6649466873572034,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8727206587791443,
      "num_tokens": 520584132.0,
      "step": 14961
    },
    {
      "epoch": 2.778458681522748,
      "grad_norm": 1.6386002009539151,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8804867267608643,
      "num_tokens": 520618475.0,
      "step": 14962
    },
    {
      "epoch": 2.778644382544104,
      "grad_norm": 1.6842775135546892,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8838477730751038,
      "num_tokens": 520648297.0,
      "step": 14963
    },
    {
      "epoch": 2.7788300835654596,
      "grad_norm": 1.7292675565261442,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8774886727333069,
      "num_tokens": 520679519.0,
      "step": 14964
    },
    {
      "epoch": 2.7790157845868153,
      "grad_norm": 1.6909324716186402,
      "learning_rate": 1e-06,
      "loss": 0.3229,
      "mean_token_accuracy": 0.8871131539344788,
      "num_tokens": 520708771.0,
      "step": 14965
    },
    {
      "epoch": 2.779201485608171,
      "grad_norm": 1.576716786629959,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.887047290802002,
      "num_tokens": 520750018.0,
      "step": 14966
    },
    {
      "epoch": 2.7793871866295263,
      "grad_norm": 1.7690596504988325,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8797343969345093,
      "num_tokens": 520780166.0,
      "step": 14967
    },
    {
      "epoch": 2.779572887650882,
      "grad_norm": 1.5303840386177026,
      "learning_rate": 1e-06,
      "loss": 0.2982,
      "mean_token_accuracy": 0.8926503658294678,
      "num_tokens": 520815494.0,
      "step": 14968
    },
    {
      "epoch": 2.779758588672238,
      "grad_norm": 1.5456473201922376,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8900848031044006,
      "num_tokens": 520850134.0,
      "step": 14969
    },
    {
      "epoch": 2.779944289693593,
      "grad_norm": 1.4315881152028227,
      "learning_rate": 1e-06,
      "loss": 0.3166,
      "mean_token_accuracy": 0.8884505033493042,
      "num_tokens": 520890506.0,
      "step": 14970
    },
    {
      "epoch": 2.780129990714949,
      "grad_norm": 1.556595478576853,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8881020545959473,
      "num_tokens": 520927691.0,
      "step": 14971
    },
    {
      "epoch": 2.7803156917363046,
      "grad_norm": 1.7629892618418612,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8751254081726074,
      "num_tokens": 520956583.0,
      "step": 14972
    },
    {
      "epoch": 2.7805013927576603,
      "grad_norm": 1.556743793017416,
      "learning_rate": 1e-06,
      "loss": 0.3581,
      "mean_token_accuracy": 0.8774942755699158,
      "num_tokens": 520994415.0,
      "step": 14973
    },
    {
      "epoch": 2.780687093779016,
      "grad_norm": 1.6537394793759475,
      "learning_rate": 1e-06,
      "loss": 0.3354,
      "mean_token_accuracy": 0.8827053904533386,
      "num_tokens": 521026473.0,
      "step": 14974
    },
    {
      "epoch": 2.7808727948003713,
      "grad_norm": 1.701622496650778,
      "learning_rate": 1e-06,
      "loss": 0.3333,
      "mean_token_accuracy": 0.8845924139022827,
      "num_tokens": 521058698.0,
      "step": 14975
    },
    {
      "epoch": 2.781058495821727,
      "grad_norm": 1.6360005656861845,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8858217000961304,
      "num_tokens": 521091878.0,
      "step": 14976
    },
    {
      "epoch": 2.7812441968430828,
      "grad_norm": 1.6262668409615264,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8805307149887085,
      "num_tokens": 521130848.0,
      "step": 14977
    },
    {
      "epoch": 2.781429897864438,
      "grad_norm": 1.6865714861896137,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.8762246966362,
      "num_tokens": 521161298.0,
      "step": 14978
    },
    {
      "epoch": 2.781615598885794,
      "grad_norm": 1.6673281325551914,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.894782543182373,
      "num_tokens": 521190407.0,
      "step": 14979
    },
    {
      "epoch": 2.7818012999071495,
      "grad_norm": 1.5745161574215578,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8842198252677917,
      "num_tokens": 521226090.0,
      "step": 14980
    },
    {
      "epoch": 2.7819870009285053,
      "grad_norm": 1.5742384517267833,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8843283653259277,
      "num_tokens": 521260946.0,
      "step": 14981
    },
    {
      "epoch": 2.782172701949861,
      "grad_norm": 1.6197289527354928,
      "learning_rate": 1e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8710213899612427,
      "num_tokens": 521294641.0,
      "step": 14982
    },
    {
      "epoch": 2.7823584029712163,
      "grad_norm": 1.5773269229980753,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8825116157531738,
      "num_tokens": 521328371.0,
      "step": 14983
    },
    {
      "epoch": 2.782544103992572,
      "grad_norm": 1.4792456717658429,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8888883590698242,
      "num_tokens": 521365053.0,
      "step": 14984
    },
    {
      "epoch": 2.7827298050139273,
      "grad_norm": 1.6266634286566897,
      "learning_rate": 1e-06,
      "loss": 0.2891,
      "mean_token_accuracy": 0.8939565420150757,
      "num_tokens": 521399939.0,
      "step": 14985
    },
    {
      "epoch": 2.782915506035283,
      "grad_norm": 1.5756076924776612,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8839467167854309,
      "num_tokens": 521435145.0,
      "step": 14986
    },
    {
      "epoch": 2.7831012070566388,
      "grad_norm": 1.470372243500065,
      "learning_rate": 1e-06,
      "loss": 0.3159,
      "mean_token_accuracy": 0.8893473744392395,
      "num_tokens": 521472670.0,
      "step": 14987
    },
    {
      "epoch": 2.7832869080779945,
      "grad_norm": 1.590463663457844,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8807387351989746,
      "num_tokens": 521510604.0,
      "step": 14988
    },
    {
      "epoch": 2.7834726090993502,
      "grad_norm": 1.4914349914494083,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.8917701244354248,
      "num_tokens": 521545183.0,
      "step": 14989
    },
    {
      "epoch": 2.7836583101207055,
      "grad_norm": 1.7011524292106759,
      "learning_rate": 1e-06,
      "loss": 0.3893,
      "mean_token_accuracy": 0.8628852367401123,
      "num_tokens": 521579627.0,
      "step": 14990
    },
    {
      "epoch": 2.7838440111420613,
      "grad_norm": 1.7081080661309653,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8951750993728638,
      "num_tokens": 521612999.0,
      "step": 14991
    },
    {
      "epoch": 2.784029712163417,
      "grad_norm": 1.544826389590652,
      "learning_rate": 1e-06,
      "loss": 0.2617,
      "mean_token_accuracy": 0.9021709561347961,
      "num_tokens": 521644592.0,
      "step": 14992
    },
    {
      "epoch": 2.7842154131847723,
      "grad_norm": 1.5371913888062596,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8941649198532104,
      "num_tokens": 521675997.0,
      "step": 14993
    },
    {
      "epoch": 2.784401114206128,
      "grad_norm": 1.6106507125268739,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8837822079658508,
      "num_tokens": 521708209.0,
      "step": 14994
    },
    {
      "epoch": 2.7845868152274837,
      "grad_norm": 1.5764621854733505,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.8956524729728699,
      "num_tokens": 521741697.0,
      "step": 14995
    },
    {
      "epoch": 2.7847725162488395,
      "grad_norm": 1.4508710381435175,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8857669830322266,
      "num_tokens": 521782053.0,
      "step": 14996
    },
    {
      "epoch": 2.784958217270195,
      "grad_norm": 1.4600136465167823,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8932539820671082,
      "num_tokens": 521817120.0,
      "step": 14997
    },
    {
      "epoch": 2.7851439182915505,
      "grad_norm": 1.5245071983372525,
      "learning_rate": 1e-06,
      "loss": 0.3014,
      "mean_token_accuracy": 0.893605649471283,
      "num_tokens": 521851059.0,
      "step": 14998
    },
    {
      "epoch": 2.7853296193129062,
      "grad_norm": 1.4893915732266425,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8841663599014282,
      "num_tokens": 521892770.0,
      "step": 14999
    },
    {
      "epoch": 2.785515320334262,
      "grad_norm": 1.7914350350335,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8946069478988647,
      "num_tokens": 521917795.0,
      "step": 15000
    },
    {
      "epoch": 2.7857010213556173,
      "grad_norm": 1.775838108174093,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8879812359809875,
      "num_tokens": 521945940.0,
      "step": 15001
    },
    {
      "epoch": 2.785886722376973,
      "grad_norm": 1.4925887721840816,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8873811960220337,
      "num_tokens": 521981430.0,
      "step": 15002
    },
    {
      "epoch": 2.7860724233983287,
      "grad_norm": 1.5255522814409432,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8935137391090393,
      "num_tokens": 522015341.0,
      "step": 15003
    },
    {
      "epoch": 2.7862581244196845,
      "grad_norm": 1.7186881395234728,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.870392918586731,
      "num_tokens": 522050414.0,
      "step": 15004
    },
    {
      "epoch": 2.78644382544104,
      "grad_norm": 1.700244071817999,
      "learning_rate": 1e-06,
      "loss": 0.3251,
      "mean_token_accuracy": 0.8876737952232361,
      "num_tokens": 522077846.0,
      "step": 15005
    },
    {
      "epoch": 2.7866295264623955,
      "grad_norm": 1.5552877211909533,
      "learning_rate": 1e-06,
      "loss": 0.3628,
      "mean_token_accuracy": 0.8745354413986206,
      "num_tokens": 522116928.0,
      "step": 15006
    },
    {
      "epoch": 2.786815227483751,
      "grad_norm": 1.4170532286546313,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8754525184631348,
      "num_tokens": 522157244.0,
      "step": 15007
    },
    {
      "epoch": 2.7870009285051065,
      "grad_norm": 1.4804503123953738,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.881557822227478,
      "num_tokens": 522194539.0,
      "step": 15008
    },
    {
      "epoch": 2.7871866295264622,
      "grad_norm": 1.6172670947391923,
      "learning_rate": 1e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.8685654401779175,
      "num_tokens": 522230447.0,
      "step": 15009
    },
    {
      "epoch": 2.787372330547818,
      "grad_norm": 1.495545701289728,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.894018828868866,
      "num_tokens": 522267618.0,
      "step": 15010
    },
    {
      "epoch": 2.7875580315691737,
      "grad_norm": 1.4950070125498407,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.887910008430481,
      "num_tokens": 522303999.0,
      "step": 15011
    },
    {
      "epoch": 2.7877437325905294,
      "grad_norm": 1.5840440500885364,
      "learning_rate": 1e-06,
      "loss": 0.3115,
      "mean_token_accuracy": 0.8923305869102478,
      "num_tokens": 522336744.0,
      "step": 15012
    },
    {
      "epoch": 2.7879294336118847,
      "grad_norm": 1.4347248470469454,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.8990403413772583,
      "num_tokens": 522375972.0,
      "step": 15013
    },
    {
      "epoch": 2.7881151346332405,
      "grad_norm": 1.5019255393913031,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8937811255455017,
      "num_tokens": 522410125.0,
      "step": 15014
    },
    {
      "epoch": 2.788300835654596,
      "grad_norm": 1.6795608824607873,
      "learning_rate": 1e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8598929643630981,
      "num_tokens": 522443542.0,
      "step": 15015
    },
    {
      "epoch": 2.7884865366759515,
      "grad_norm": 1.5519838109086517,
      "learning_rate": 1e-06,
      "loss": 0.2947,
      "mean_token_accuracy": 0.8991268873214722,
      "num_tokens": 522477490.0,
      "step": 15016
    },
    {
      "epoch": 2.788672237697307,
      "grad_norm": 1.5999114407569874,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8820675015449524,
      "num_tokens": 522514140.0,
      "step": 15017
    },
    {
      "epoch": 2.788857938718663,
      "grad_norm": 1.6536144627280345,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8725200891494751,
      "num_tokens": 522546267.0,
      "step": 15018
    },
    {
      "epoch": 2.7890436397400187,
      "grad_norm": 1.5955158637794822,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8904714584350586,
      "num_tokens": 522579598.0,
      "step": 15019
    },
    {
      "epoch": 2.7892293407613744,
      "grad_norm": 1.416261908060885,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8892005085945129,
      "num_tokens": 522619537.0,
      "step": 15020
    },
    {
      "epoch": 2.7894150417827297,
      "grad_norm": 1.5488377813535585,
      "learning_rate": 1e-06,
      "loss": 0.2933,
      "mean_token_accuracy": 0.8965543508529663,
      "num_tokens": 522654354.0,
      "step": 15021
    },
    {
      "epoch": 2.7896007428040854,
      "grad_norm": 1.484790406699447,
      "learning_rate": 1e-06,
      "loss": 0.2705,
      "mean_token_accuracy": 0.9013435244560242,
      "num_tokens": 522690692.0,
      "step": 15022
    },
    {
      "epoch": 2.789786443825441,
      "grad_norm": 1.5366471376233855,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8806871175765991,
      "num_tokens": 522727967.0,
      "step": 15023
    },
    {
      "epoch": 2.7899721448467965,
      "grad_norm": 1.6991615483168676,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8807819485664368,
      "num_tokens": 522759425.0,
      "step": 15024
    },
    {
      "epoch": 2.790157845868152,
      "grad_norm": 1.6679255964716635,
      "learning_rate": 1e-06,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8826053738594055,
      "num_tokens": 522800698.0,
      "step": 15025
    },
    {
      "epoch": 2.790343546889508,
      "grad_norm": 1.6124118786895765,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8930203914642334,
      "num_tokens": 522833489.0,
      "step": 15026
    },
    {
      "epoch": 2.7905292479108637,
      "grad_norm": 1.3990813220744462,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8946517705917358,
      "num_tokens": 522874467.0,
      "step": 15027
    },
    {
      "epoch": 2.7907149489322194,
      "grad_norm": 1.413886480680085,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8953121900558472,
      "num_tokens": 522911550.0,
      "step": 15028
    },
    {
      "epoch": 2.7909006499535747,
      "grad_norm": 1.6991909689948028,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8860254883766174,
      "num_tokens": 522941788.0,
      "step": 15029
    },
    {
      "epoch": 2.7910863509749304,
      "grad_norm": 1.641434785770896,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8863497972488403,
      "num_tokens": 522976922.0,
      "step": 15030
    },
    {
      "epoch": 2.791272051996286,
      "grad_norm": 1.4868734113647835,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.87626713514328,
      "num_tokens": 523018723.0,
      "step": 15031
    },
    {
      "epoch": 2.7914577530176414,
      "grad_norm": 1.5931551724803281,
      "learning_rate": 1e-06,
      "loss": 0.2675,
      "mean_token_accuracy": 0.9044402241706848,
      "num_tokens": 523048074.0,
      "step": 15032
    },
    {
      "epoch": 2.791643454038997,
      "grad_norm": 1.4318926715622904,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8800119161605835,
      "num_tokens": 523090960.0,
      "step": 15033
    },
    {
      "epoch": 2.791829155060353,
      "grad_norm": 1.502147808783947,
      "learning_rate": 1e-06,
      "loss": 0.3308,
      "mean_token_accuracy": 0.8821732997894287,
      "num_tokens": 523127274.0,
      "step": 15034
    },
    {
      "epoch": 2.7920148560817086,
      "grad_norm": 1.3782986433470477,
      "learning_rate": 1e-06,
      "loss": 0.2787,
      "mean_token_accuracy": 0.90170818567276,
      "num_tokens": 523168742.0,
      "step": 15035
    },
    {
      "epoch": 2.792200557103064,
      "grad_norm": 1.4553183697480752,
      "learning_rate": 1e-06,
      "loss": 0.2664,
      "mean_token_accuracy": 0.9043579697608948,
      "num_tokens": 523203888.0,
      "step": 15036
    },
    {
      "epoch": 2.7923862581244197,
      "grad_norm": 1.5890104172027475,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8808203935623169,
      "num_tokens": 523236539.0,
      "step": 15037
    },
    {
      "epoch": 2.7925719591457754,
      "grad_norm": 1.4996938743830774,
      "learning_rate": 1e-06,
      "loss": 0.2987,
      "mean_token_accuracy": 0.8928486108779907,
      "num_tokens": 523274305.0,
      "step": 15038
    },
    {
      "epoch": 2.7927576601671307,
      "grad_norm": 1.68011014345459,
      "learning_rate": 1e-06,
      "loss": 0.3305,
      "mean_token_accuracy": 0.884568989276886,
      "num_tokens": 523305901.0,
      "step": 15039
    },
    {
      "epoch": 2.7929433611884864,
      "grad_norm": 1.5606295669165005,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8823657035827637,
      "num_tokens": 523339378.0,
      "step": 15040
    },
    {
      "epoch": 2.793129062209842,
      "grad_norm": 1.658953413291525,
      "learning_rate": 1e-06,
      "loss": 0.3557,
      "mean_token_accuracy": 0.8784015774726868,
      "num_tokens": 523374865.0,
      "step": 15041
    },
    {
      "epoch": 2.793314763231198,
      "grad_norm": 1.6641859310806075,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8772501945495605,
      "num_tokens": 523404994.0,
      "step": 15042
    },
    {
      "epoch": 2.7935004642525536,
      "grad_norm": 1.4936212324782268,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8873441219329834,
      "num_tokens": 523441384.0,
      "step": 15043
    },
    {
      "epoch": 2.793686165273909,
      "grad_norm": 1.46629292736501,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8931166529655457,
      "num_tokens": 523478020.0,
      "step": 15044
    },
    {
      "epoch": 2.7938718662952646,
      "grad_norm": 1.4623794176580787,
      "learning_rate": 1e-06,
      "loss": 0.2691,
      "mean_token_accuracy": 0.9037858247756958,
      "num_tokens": 523512422.0,
      "step": 15045
    },
    {
      "epoch": 2.7940575673166204,
      "grad_norm": 1.7123265447270128,
      "learning_rate": 1e-06,
      "loss": 0.3744,
      "mean_token_accuracy": 0.8691728115081787,
      "num_tokens": 523543298.0,
      "step": 15046
    },
    {
      "epoch": 2.7942432683379756,
      "grad_norm": 1.5936054111536464,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.8939988613128662,
      "num_tokens": 523573590.0,
      "step": 15047
    },
    {
      "epoch": 2.7944289693593314,
      "grad_norm": 1.6597832685951928,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8903345465660095,
      "num_tokens": 523602612.0,
      "step": 15048
    },
    {
      "epoch": 2.794614670380687,
      "grad_norm": 1.4853958957868445,
      "learning_rate": 1e-06,
      "loss": 0.3275,
      "mean_token_accuracy": 0.8884301781654358,
      "num_tokens": 523638527.0,
      "step": 15049
    },
    {
      "epoch": 2.794800371402043,
      "grad_norm": 1.5827923678875755,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8801161050796509,
      "num_tokens": 523673060.0,
      "step": 15050
    },
    {
      "epoch": 2.7949860724233986,
      "grad_norm": 1.5425717981049116,
      "learning_rate": 1e-06,
      "loss": 0.3214,
      "mean_token_accuracy": 0.8859414458274841,
      "num_tokens": 523709258.0,
      "step": 15051
    },
    {
      "epoch": 2.795171773444754,
      "grad_norm": 1.6061822524884684,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8800397515296936,
      "num_tokens": 523744764.0,
      "step": 15052
    },
    {
      "epoch": 2.7953574744661096,
      "grad_norm": 1.6411050759221253,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8833834528923035,
      "num_tokens": 523774083.0,
      "step": 15053
    },
    {
      "epoch": 2.7955431754874653,
      "grad_norm": 1.6612341222267804,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8822381496429443,
      "num_tokens": 523804996.0,
      "step": 15054
    },
    {
      "epoch": 2.7957288765088206,
      "grad_norm": 1.559397933217124,
      "learning_rate": 1e-06,
      "loss": 0.3495,
      "mean_token_accuracy": 0.8768616318702698,
      "num_tokens": 523840570.0,
      "step": 15055
    },
    {
      "epoch": 2.7959145775301764,
      "grad_norm": 1.568333044587099,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8923467993736267,
      "num_tokens": 523875605.0,
      "step": 15056
    },
    {
      "epoch": 2.796100278551532,
      "grad_norm": 1.610710308327321,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8799418210983276,
      "num_tokens": 523910460.0,
      "step": 15057
    },
    {
      "epoch": 2.796285979572888,
      "grad_norm": 1.4420520704767654,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8939144015312195,
      "num_tokens": 523945893.0,
      "step": 15058
    },
    {
      "epoch": 2.796471680594243,
      "grad_norm": 1.6830536985637572,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8892059326171875,
      "num_tokens": 523976151.0,
      "step": 15059
    },
    {
      "epoch": 2.796657381615599,
      "grad_norm": 1.6613249768280236,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8968634009361267,
      "num_tokens": 524005969.0,
      "step": 15060
    },
    {
      "epoch": 2.7968430826369546,
      "grad_norm": 1.597687276732592,
      "learning_rate": 1e-06,
      "loss": 0.3583,
      "mean_token_accuracy": 0.8797590732574463,
      "num_tokens": 524041257.0,
      "step": 15061
    },
    {
      "epoch": 2.79702878365831,
      "grad_norm": 1.795182384947898,
      "learning_rate": 1e-06,
      "loss": 0.337,
      "mean_token_accuracy": 0.8871592879295349,
      "num_tokens": 524066151.0,
      "step": 15062
    },
    {
      "epoch": 2.7972144846796656,
      "grad_norm": 1.60022343745571,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.901679515838623,
      "num_tokens": 524095311.0,
      "step": 15063
    },
    {
      "epoch": 2.7974001857010213,
      "grad_norm": 1.5583745975649916,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8916398286819458,
      "num_tokens": 524130404.0,
      "step": 15064
    },
    {
      "epoch": 2.797585886722377,
      "grad_norm": 1.474878825006911,
      "learning_rate": 1e-06,
      "loss": 0.2872,
      "mean_token_accuracy": 0.8964473009109497,
      "num_tokens": 524168035.0,
      "step": 15065
    },
    {
      "epoch": 2.797771587743733,
      "grad_norm": 1.399346622811908,
      "learning_rate": 1e-06,
      "loss": 0.2819,
      "mean_token_accuracy": 0.9022817015647888,
      "num_tokens": 524207383.0,
      "step": 15066
    },
    {
      "epoch": 2.797957288765088,
      "grad_norm": 1.4674398646748337,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.889029324054718,
      "num_tokens": 524248079.0,
      "step": 15067
    },
    {
      "epoch": 2.798142989786444,
      "grad_norm": 1.6813343030406496,
      "learning_rate": 1e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8973743915557861,
      "num_tokens": 524273767.0,
      "step": 15068
    },
    {
      "epoch": 2.7983286908077996,
      "grad_norm": 1.3940615537000052,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8894814848899841,
      "num_tokens": 524314651.0,
      "step": 15069
    },
    {
      "epoch": 2.798514391829155,
      "grad_norm": 1.5944936397881917,
      "learning_rate": 1e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8773790597915649,
      "num_tokens": 524350653.0,
      "step": 15070
    },
    {
      "epoch": 2.7987000928505106,
      "grad_norm": 1.440550499845607,
      "learning_rate": 1e-06,
      "loss": 0.2831,
      "mean_token_accuracy": 0.8985036015510559,
      "num_tokens": 524389493.0,
      "step": 15071
    },
    {
      "epoch": 2.7988857938718663,
      "grad_norm": 1.3924959919829836,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.8884557485580444,
      "num_tokens": 524430592.0,
      "step": 15072
    },
    {
      "epoch": 2.799071494893222,
      "grad_norm": 1.585415115076567,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8883043527603149,
      "num_tokens": 524463395.0,
      "step": 15073
    },
    {
      "epoch": 2.799257195914578,
      "grad_norm": 1.570192555620356,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8845476508140564,
      "num_tokens": 524500327.0,
      "step": 15074
    },
    {
      "epoch": 2.799442896935933,
      "grad_norm": 1.548058391240832,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.886698842048645,
      "num_tokens": 524534711.0,
      "step": 15075
    },
    {
      "epoch": 2.799628597957289,
      "grad_norm": 1.5485597079754623,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8893685340881348,
      "num_tokens": 524568678.0,
      "step": 15076
    },
    {
      "epoch": 2.7998142989786445,
      "grad_norm": 1.6169138448186018,
      "learning_rate": 1e-06,
      "loss": 0.3133,
      "mean_token_accuracy": 0.8888505697250366,
      "num_tokens": 524605052.0,
      "step": 15077
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4839663791607152,
      "learning_rate": 1e-06,
      "loss": 0.2936,
      "mean_token_accuracy": 0.8933970928192139,
      "num_tokens": 524642261.0,
      "step": 15078
    },
    {
      "epoch": 2.8001857010213556,
      "grad_norm": 1.5973232796683667,
      "learning_rate": 1e-06,
      "loss": 0.2857,
      "mean_token_accuracy": 0.9023818969726562,
      "num_tokens": 524670150.0,
      "step": 15079
    },
    {
      "epoch": 2.8003714020427113,
      "grad_norm": 1.4669388413939601,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8855303525924683,
      "num_tokens": 524707654.0,
      "step": 15080
    },
    {
      "epoch": 2.800557103064067,
      "grad_norm": 1.6399493920767514,
      "learning_rate": 1e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8696147203445435,
      "num_tokens": 524743969.0,
      "step": 15081
    },
    {
      "epoch": 2.8007428040854223,
      "grad_norm": 1.4422916626419582,
      "learning_rate": 1e-06,
      "loss": 0.3158,
      "mean_token_accuracy": 0.8869800567626953,
      "num_tokens": 524783352.0,
      "step": 15082
    },
    {
      "epoch": 2.800928505106778,
      "grad_norm": 1.5239786164499445,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8864414691925049,
      "num_tokens": 524821535.0,
      "step": 15083
    },
    {
      "epoch": 2.8011142061281338,
      "grad_norm": 1.5529349443458116,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8718234300613403,
      "num_tokens": 524858627.0,
      "step": 15084
    },
    {
      "epoch": 2.801299907149489,
      "grad_norm": 1.4911963453738593,
      "learning_rate": 1e-06,
      "loss": 0.2872,
      "mean_token_accuracy": 0.8984279632568359,
      "num_tokens": 524895778.0,
      "step": 15085
    },
    {
      "epoch": 2.801485608170845,
      "grad_norm": 1.5503840117096566,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8727714419364929,
      "num_tokens": 524933615.0,
      "step": 15086
    },
    {
      "epoch": 2.8016713091922005,
      "grad_norm": 1.7154991835989304,
      "learning_rate": 1e-06,
      "loss": 0.3216,
      "mean_token_accuracy": 0.8908543586730957,
      "num_tokens": 524967895.0,
      "step": 15087
    },
    {
      "epoch": 2.8018570102135563,
      "grad_norm": 1.669386735800381,
      "learning_rate": 1e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.8846781849861145,
      "num_tokens": 525001083.0,
      "step": 15088
    },
    {
      "epoch": 2.802042711234912,
      "grad_norm": 1.654702247640482,
      "learning_rate": 1e-06,
      "loss": 0.3265,
      "mean_token_accuracy": 0.8869311809539795,
      "num_tokens": 525032915.0,
      "step": 15089
    },
    {
      "epoch": 2.8022284122562673,
      "grad_norm": 1.7900820318026813,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8869472742080688,
      "num_tokens": 525058912.0,
      "step": 15090
    },
    {
      "epoch": 2.802414113277623,
      "grad_norm": 1.5959020266361108,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8982311487197876,
      "num_tokens": 525088076.0,
      "step": 15091
    },
    {
      "epoch": 2.8025998142989788,
      "grad_norm": 1.5893075365461253,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8824830055236816,
      "num_tokens": 525124719.0,
      "step": 15092
    },
    {
      "epoch": 2.802785515320334,
      "grad_norm": 1.4545920568951016,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.8992027044296265,
      "num_tokens": 525161131.0,
      "step": 15093
    },
    {
      "epoch": 2.8029712163416898,
      "grad_norm": 1.5687649836485151,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8879299163818359,
      "num_tokens": 525195350.0,
      "step": 15094
    },
    {
      "epoch": 2.8031569173630455,
      "grad_norm": 1.7755971902190704,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.891862690448761,
      "num_tokens": 525220959.0,
      "step": 15095
    },
    {
      "epoch": 2.8033426183844012,
      "grad_norm": 1.6575535251262312,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8852225542068481,
      "num_tokens": 525254450.0,
      "step": 15096
    },
    {
      "epoch": 2.803528319405757,
      "grad_norm": 1.6301412185728168,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8861700892448425,
      "num_tokens": 525284902.0,
      "step": 15097
    },
    {
      "epoch": 2.8037140204271123,
      "grad_norm": 1.564673798325073,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8800238370895386,
      "num_tokens": 525323682.0,
      "step": 15098
    },
    {
      "epoch": 2.803899721448468,
      "grad_norm": 1.5465000872663002,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8926113843917847,
      "num_tokens": 525361598.0,
      "step": 15099
    },
    {
      "epoch": 2.8040854224698237,
      "grad_norm": 1.6017238759829213,
      "learning_rate": 1e-06,
      "loss": 0.2945,
      "mean_token_accuracy": 0.8980944752693176,
      "num_tokens": 525395339.0,
      "step": 15100
    },
    {
      "epoch": 2.804271123491179,
      "grad_norm": 1.66973026235008,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8920469284057617,
      "num_tokens": 525426709.0,
      "step": 15101
    },
    {
      "epoch": 2.8044568245125348,
      "grad_norm": 1.4746000672654176,
      "learning_rate": 1e-06,
      "loss": 0.3136,
      "mean_token_accuracy": 0.8915163278579712,
      "num_tokens": 525464570.0,
      "step": 15102
    },
    {
      "epoch": 2.8046425255338905,
      "grad_norm": 1.5398457811674244,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8908382654190063,
      "num_tokens": 525500145.0,
      "step": 15103
    },
    {
      "epoch": 2.804828226555246,
      "grad_norm": 1.6255505474988463,
      "learning_rate": 1e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8629542589187622,
      "num_tokens": 525538501.0,
      "step": 15104
    },
    {
      "epoch": 2.8050139275766015,
      "grad_norm": 1.646537766844159,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8801595568656921,
      "num_tokens": 525569746.0,
      "step": 15105
    },
    {
      "epoch": 2.8051996285979572,
      "grad_norm": 1.5899155716612092,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8930546045303345,
      "num_tokens": 525605010.0,
      "step": 15106
    },
    {
      "epoch": 2.805385329619313,
      "grad_norm": 1.6447824814782728,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8896236419677734,
      "num_tokens": 525639751.0,
      "step": 15107
    },
    {
      "epoch": 2.8055710306406683,
      "grad_norm": 1.7959400696342926,
      "learning_rate": 1e-06,
      "loss": 0.3694,
      "mean_token_accuracy": 0.8715993762016296,
      "num_tokens": 525670059.0,
      "step": 15108
    },
    {
      "epoch": 2.805756731662024,
      "grad_norm": 1.664255296898059,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8880748152732849,
      "num_tokens": 525698015.0,
      "step": 15109
    },
    {
      "epoch": 2.8059424326833797,
      "grad_norm": 1.579181008462787,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8941227793693542,
      "num_tokens": 525729303.0,
      "step": 15110
    },
    {
      "epoch": 2.8061281337047355,
      "grad_norm": 1.4760832528239654,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.8956538438796997,
      "num_tokens": 525770725.0,
      "step": 15111
    },
    {
      "epoch": 2.806313834726091,
      "grad_norm": 1.4870081628786764,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8883268237113953,
      "num_tokens": 525808549.0,
      "step": 15112
    },
    {
      "epoch": 2.8064995357474465,
      "grad_norm": 1.6220310473545472,
      "learning_rate": 1e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8667783141136169,
      "num_tokens": 525847192.0,
      "step": 15113
    },
    {
      "epoch": 2.806685236768802,
      "grad_norm": 1.5115341796500483,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8916033506393433,
      "num_tokens": 525881453.0,
      "step": 15114
    },
    {
      "epoch": 2.806870937790158,
      "grad_norm": 1.527043405612597,
      "learning_rate": 1e-06,
      "loss": 0.3338,
      "mean_token_accuracy": 0.8877457976341248,
      "num_tokens": 525918106.0,
      "step": 15115
    },
    {
      "epoch": 2.8070566388115132,
      "grad_norm": 1.4359060970929516,
      "learning_rate": 1e-06,
      "loss": 0.3019,
      "mean_token_accuracy": 0.8912099599838257,
      "num_tokens": 525954881.0,
      "step": 15116
    },
    {
      "epoch": 2.807242339832869,
      "grad_norm": 1.6801847135214418,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8832659721374512,
      "num_tokens": 525989282.0,
      "step": 15117
    },
    {
      "epoch": 2.8074280408542247,
      "grad_norm": 1.5298467899212074,
      "learning_rate": 1e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8734613060951233,
      "num_tokens": 526030284.0,
      "step": 15118
    },
    {
      "epoch": 2.8076137418755804,
      "grad_norm": 1.7746798287750516,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8861914873123169,
      "num_tokens": 526055997.0,
      "step": 15119
    },
    {
      "epoch": 2.807799442896936,
      "grad_norm": 1.5607475569462383,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8781977891921997,
      "num_tokens": 526094491.0,
      "step": 15120
    },
    {
      "epoch": 2.8079851439182915,
      "grad_norm": 1.507621191315854,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8877511024475098,
      "num_tokens": 526132754.0,
      "step": 15121
    },
    {
      "epoch": 2.808170844939647,
      "grad_norm": 1.403693983933002,
      "learning_rate": 1e-06,
      "loss": 0.2792,
      "mean_token_accuracy": 0.8979097008705139,
      "num_tokens": 526172507.0,
      "step": 15122
    },
    {
      "epoch": 2.808356545961003,
      "grad_norm": 1.464989846760078,
      "learning_rate": 1e-06,
      "loss": 0.2996,
      "mean_token_accuracy": 0.8945263624191284,
      "num_tokens": 526211693.0,
      "step": 15123
    },
    {
      "epoch": 2.808542246982358,
      "grad_norm": 1.609643647732923,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.8887654542922974,
      "num_tokens": 526241770.0,
      "step": 15124
    },
    {
      "epoch": 2.808727948003714,
      "grad_norm": 1.6052959343418851,
      "learning_rate": 1e-06,
      "loss": 0.289,
      "mean_token_accuracy": 0.8999176025390625,
      "num_tokens": 526272286.0,
      "step": 15125
    },
    {
      "epoch": 2.8089136490250697,
      "grad_norm": 1.4680513140754836,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8787050843238831,
      "num_tokens": 526312831.0,
      "step": 15126
    },
    {
      "epoch": 2.8090993500464254,
      "grad_norm": 1.4982400471470012,
      "learning_rate": 1e-06,
      "loss": 0.2639,
      "mean_token_accuracy": 0.9027231931686401,
      "num_tokens": 526347713.0,
      "step": 15127
    },
    {
      "epoch": 2.809285051067781,
      "grad_norm": 1.55169235218359,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.882230818271637,
      "num_tokens": 526382494.0,
      "step": 15128
    },
    {
      "epoch": 2.8094707520891364,
      "grad_norm": 1.6382770814188086,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8954728245735168,
      "num_tokens": 526413290.0,
      "step": 15129
    },
    {
      "epoch": 2.809656453110492,
      "grad_norm": 1.625051819494351,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8817211389541626,
      "num_tokens": 526449421.0,
      "step": 15130
    },
    {
      "epoch": 2.8098421541318475,
      "grad_norm": 1.4842164293083846,
      "learning_rate": 1e-06,
      "loss": 0.2775,
      "mean_token_accuracy": 0.9031341075897217,
      "num_tokens": 526483718.0,
      "step": 15131
    },
    {
      "epoch": 2.810027855153203,
      "grad_norm": 1.5895988149752012,
      "learning_rate": 1e-06,
      "loss": 0.3603,
      "mean_token_accuracy": 0.8755806088447571,
      "num_tokens": 526522456.0,
      "step": 15132
    },
    {
      "epoch": 2.810213556174559,
      "grad_norm": 1.614542923771722,
      "learning_rate": 1e-06,
      "loss": 0.331,
      "mean_token_accuracy": 0.882151186466217,
      "num_tokens": 526559136.0,
      "step": 15133
    },
    {
      "epoch": 2.8103992571959147,
      "grad_norm": 1.5439929223120263,
      "learning_rate": 1e-06,
      "loss": 0.2978,
      "mean_token_accuracy": 0.8961358070373535,
      "num_tokens": 526590607.0,
      "step": 15134
    },
    {
      "epoch": 2.8105849582172704,
      "grad_norm": 1.7195518917592838,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.8942581415176392,
      "num_tokens": 526617621.0,
      "step": 15135
    },
    {
      "epoch": 2.8107706592386257,
      "grad_norm": 1.5696823721815831,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8804651498794556,
      "num_tokens": 526653519.0,
      "step": 15136
    },
    {
      "epoch": 2.8109563602599814,
      "grad_norm": 1.4161029140996697,
      "learning_rate": 1e-06,
      "loss": 0.3186,
      "mean_token_accuracy": 0.8900790214538574,
      "num_tokens": 526693997.0,
      "step": 15137
    },
    {
      "epoch": 2.811142061281337,
      "grad_norm": 1.3611418442440284,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.894366979598999,
      "num_tokens": 526735102.0,
      "step": 15138
    },
    {
      "epoch": 2.8113277623026924,
      "grad_norm": 1.4959215434873085,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8879050612449646,
      "num_tokens": 526771294.0,
      "step": 15139
    },
    {
      "epoch": 2.811513463324048,
      "grad_norm": 1.4958465057289063,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8896312117576599,
      "num_tokens": 526808578.0,
      "step": 15140
    },
    {
      "epoch": 2.811699164345404,
      "grad_norm": 1.6211259799261797,
      "learning_rate": 1e-06,
      "loss": 0.3825,
      "mean_token_accuracy": 0.8659391403198242,
      "num_tokens": 526847651.0,
      "step": 15141
    },
    {
      "epoch": 2.8118848653667596,
      "grad_norm": 1.562090036292451,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8962321281433105,
      "num_tokens": 526877032.0,
      "step": 15142
    },
    {
      "epoch": 2.8120705663881154,
      "grad_norm": 1.522811312987056,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8867783546447754,
      "num_tokens": 526910099.0,
      "step": 15143
    },
    {
      "epoch": 2.8122562674094707,
      "grad_norm": 1.4694505193377196,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.8964641094207764,
      "num_tokens": 526944658.0,
      "step": 15144
    },
    {
      "epoch": 2.8124419684308264,
      "grad_norm": 1.398232308436811,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8906158208847046,
      "num_tokens": 526984668.0,
      "step": 15145
    },
    {
      "epoch": 2.812627669452182,
      "grad_norm": 1.5802193172243517,
      "learning_rate": 1e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.8873972296714783,
      "num_tokens": 527018914.0,
      "step": 15146
    },
    {
      "epoch": 2.8128133704735374,
      "grad_norm": 1.4512587120509197,
      "learning_rate": 1e-06,
      "loss": 0.3515,
      "mean_token_accuracy": 0.8797286748886108,
      "num_tokens": 527062857.0,
      "step": 15147
    },
    {
      "epoch": 2.812999071494893,
      "grad_norm": 1.6244306189365234,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8775432109832764,
      "num_tokens": 527098201.0,
      "step": 15148
    },
    {
      "epoch": 2.813184772516249,
      "grad_norm": 1.5294397281065324,
      "learning_rate": 1e-06,
      "loss": 0.2794,
      "mean_token_accuracy": 0.9017083644866943,
      "num_tokens": 527133096.0,
      "step": 15149
    },
    {
      "epoch": 2.8133704735376046,
      "grad_norm": 1.7094955705105015,
      "learning_rate": 1e-06,
      "loss": 0.2655,
      "mean_token_accuracy": 0.9021730422973633,
      "num_tokens": 527159724.0,
      "step": 15150
    },
    {
      "epoch": 2.8135561745589603,
      "grad_norm": 1.4986175099384094,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8804664611816406,
      "num_tokens": 527201204.0,
      "step": 15151
    },
    {
      "epoch": 2.8137418755803156,
      "grad_norm": 1.5194398141928556,
      "learning_rate": 1e-06,
      "loss": 0.3571,
      "mean_token_accuracy": 0.8766651153564453,
      "num_tokens": 527238371.0,
      "step": 15152
    },
    {
      "epoch": 2.8139275766016714,
      "grad_norm": 1.4377880137104813,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.8904765248298645,
      "num_tokens": 527276009.0,
      "step": 15153
    },
    {
      "epoch": 2.8141132776230267,
      "grad_norm": 1.5107958953891323,
      "learning_rate": 1e-06,
      "loss": 0.2998,
      "mean_token_accuracy": 0.8930317759513855,
      "num_tokens": 527309389.0,
      "step": 15154
    },
    {
      "epoch": 2.8142989786443824,
      "grad_norm": 1.4027769612288021,
      "learning_rate": 1e-06,
      "loss": 0.2668,
      "mean_token_accuracy": 0.9054535031318665,
      "num_tokens": 527348823.0,
      "step": 15155
    },
    {
      "epoch": 2.814484679665738,
      "grad_norm": 1.7476608009185122,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8852178454399109,
      "num_tokens": 527377112.0,
      "step": 15156
    },
    {
      "epoch": 2.814670380687094,
      "grad_norm": 1.7110369710334499,
      "learning_rate": 1e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8747386932373047,
      "num_tokens": 527414155.0,
      "step": 15157
    },
    {
      "epoch": 2.8148560817084496,
      "grad_norm": 1.4395602099261324,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8918317556381226,
      "num_tokens": 527451982.0,
      "step": 15158
    },
    {
      "epoch": 2.815041782729805,
      "grad_norm": 1.6467557414423617,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8751183748245239,
      "num_tokens": 527485563.0,
      "step": 15159
    },
    {
      "epoch": 2.8152274837511606,
      "grad_norm": 1.4964627394065677,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.889424204826355,
      "num_tokens": 527524000.0,
      "step": 15160
    },
    {
      "epoch": 2.8154131847725163,
      "grad_norm": 1.457789986728475,
      "learning_rate": 1e-06,
      "loss": 0.2848,
      "mean_token_accuracy": 0.8992800116539001,
      "num_tokens": 527559821.0,
      "step": 15161
    },
    {
      "epoch": 2.8155988857938716,
      "grad_norm": 1.5858639272668662,
      "learning_rate": 1e-06,
      "loss": 0.28,
      "mean_token_accuracy": 0.9003963470458984,
      "num_tokens": 527593467.0,
      "step": 15162
    },
    {
      "epoch": 2.8157845868152274,
      "grad_norm": 1.596405761025996,
      "learning_rate": 1e-06,
      "loss": 0.3321,
      "mean_token_accuracy": 0.8838746547698975,
      "num_tokens": 527626395.0,
      "step": 15163
    },
    {
      "epoch": 2.815970287836583,
      "grad_norm": 1.8049659527093151,
      "learning_rate": 1e-06,
      "loss": 0.3588,
      "mean_token_accuracy": 0.8745640516281128,
      "num_tokens": 527655098.0,
      "step": 15164
    },
    {
      "epoch": 2.816155988857939,
      "grad_norm": 1.5742671884920412,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8907502889633179,
      "num_tokens": 527687240.0,
      "step": 15165
    },
    {
      "epoch": 2.8163416898792946,
      "grad_norm": 1.6055715097820005,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8962721824645996,
      "num_tokens": 527719651.0,
      "step": 15166
    },
    {
      "epoch": 2.81652739090065,
      "grad_norm": 1.5995496850806852,
      "learning_rate": 1e-06,
      "loss": 0.3034,
      "mean_token_accuracy": 0.8964947462081909,
      "num_tokens": 527752489.0,
      "step": 15167
    },
    {
      "epoch": 2.8167130919220056,
      "grad_norm": 1.5244834030575372,
      "learning_rate": 1e-06,
      "loss": 0.2899,
      "mean_token_accuracy": 0.898349940776825,
      "num_tokens": 527787508.0,
      "step": 15168
    },
    {
      "epoch": 2.8168987929433613,
      "grad_norm": 1.6026430410768386,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8847751021385193,
      "num_tokens": 527818127.0,
      "step": 15169
    },
    {
      "epoch": 2.8170844939647166,
      "grad_norm": 1.4363277854167285,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8853834867477417,
      "num_tokens": 527860279.0,
      "step": 15170
    },
    {
      "epoch": 2.8172701949860723,
      "grad_norm": 1.635902422401135,
      "learning_rate": 1e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8706309795379639,
      "num_tokens": 527897944.0,
      "step": 15171
    },
    {
      "epoch": 2.817455896007428,
      "grad_norm": 1.4649264628939676,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.8913137316703796,
      "num_tokens": 527935207.0,
      "step": 15172
    },
    {
      "epoch": 2.817641597028784,
      "grad_norm": 1.6906023617881154,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.893122673034668,
      "num_tokens": 527966524.0,
      "step": 15173
    },
    {
      "epoch": 2.8178272980501395,
      "grad_norm": 1.4642141783865754,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8817018866539001,
      "num_tokens": 528006986.0,
      "step": 15174
    },
    {
      "epoch": 2.818012999071495,
      "grad_norm": 1.5304736544945075,
      "learning_rate": 1e-06,
      "loss": 0.3065,
      "mean_token_accuracy": 0.8913750648498535,
      "num_tokens": 528041708.0,
      "step": 15175
    },
    {
      "epoch": 2.8181987000928506,
      "grad_norm": 1.516182155715537,
      "learning_rate": 1e-06,
      "loss": 0.3521,
      "mean_token_accuracy": 0.8753804564476013,
      "num_tokens": 528079801.0,
      "step": 15176
    },
    {
      "epoch": 2.8183844011142063,
      "grad_norm": 1.6778035786273735,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.891075849533081,
      "num_tokens": 528110383.0,
      "step": 15177
    },
    {
      "epoch": 2.8185701021355616,
      "grad_norm": 1.5287783388012373,
      "learning_rate": 1e-06,
      "loss": 0.2857,
      "mean_token_accuracy": 0.8984195590019226,
      "num_tokens": 528144217.0,
      "step": 15178
    },
    {
      "epoch": 2.8187558031569173,
      "grad_norm": 1.6467767312311128,
      "learning_rate": 1e-06,
      "loss": 0.3771,
      "mean_token_accuracy": 0.869519829750061,
      "num_tokens": 528178919.0,
      "step": 15179
    },
    {
      "epoch": 2.818941504178273,
      "grad_norm": 1.4267476406932473,
      "learning_rate": 1e-06,
      "loss": 0.2793,
      "mean_token_accuracy": 0.9011942148208618,
      "num_tokens": 528218309.0,
      "step": 15180
    },
    {
      "epoch": 2.819127205199629,
      "grad_norm": 1.630261952219976,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8737895488739014,
      "num_tokens": 528251435.0,
      "step": 15181
    },
    {
      "epoch": 2.819312906220984,
      "grad_norm": 1.5051625363238401,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.8832852840423584,
      "num_tokens": 528288982.0,
      "step": 15182
    },
    {
      "epoch": 2.81949860724234,
      "grad_norm": 1.520633155854492,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8884080648422241,
      "num_tokens": 528327496.0,
      "step": 15183
    },
    {
      "epoch": 2.8196843082636955,
      "grad_norm": 1.5596199644868287,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8917315006256104,
      "num_tokens": 528360569.0,
      "step": 15184
    },
    {
      "epoch": 2.819870009285051,
      "grad_norm": 1.5051352773060314,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.891640305519104,
      "num_tokens": 528396429.0,
      "step": 15185
    },
    {
      "epoch": 2.8200557103064066,
      "grad_norm": 1.6128043886871777,
      "learning_rate": 1e-06,
      "loss": 0.2992,
      "mean_token_accuracy": 0.893316388130188,
      "num_tokens": 528427922.0,
      "step": 15186
    },
    {
      "epoch": 2.8202414113277623,
      "grad_norm": 1.5064813171146272,
      "learning_rate": 1e-06,
      "loss": 0.332,
      "mean_token_accuracy": 0.8835934400558472,
      "num_tokens": 528465896.0,
      "step": 15187
    },
    {
      "epoch": 2.820427112349118,
      "grad_norm": 1.5374632896201335,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8771777153015137,
      "num_tokens": 528502588.0,
      "step": 15188
    },
    {
      "epoch": 2.8206128133704738,
      "grad_norm": 1.562907680711306,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.8786041736602783,
      "num_tokens": 528537169.0,
      "step": 15189
    },
    {
      "epoch": 2.820798514391829,
      "grad_norm": 1.7080726566207374,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8805258274078369,
      "num_tokens": 528570741.0,
      "step": 15190
    },
    {
      "epoch": 2.820984215413185,
      "grad_norm": 1.5556366546147837,
      "learning_rate": 1e-06,
      "loss": 0.3456,
      "mean_token_accuracy": 0.8816076517105103,
      "num_tokens": 528609433.0,
      "step": 15191
    },
    {
      "epoch": 2.8211699164345405,
      "grad_norm": 1.6056247098247642,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8878272771835327,
      "num_tokens": 528641779.0,
      "step": 15192
    },
    {
      "epoch": 2.821355617455896,
      "grad_norm": 1.5046587119431338,
      "learning_rate": 1e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8954805135726929,
      "num_tokens": 528678451.0,
      "step": 15193
    },
    {
      "epoch": 2.8215413184772515,
      "grad_norm": 1.57555551657525,
      "learning_rate": 1e-06,
      "loss": 0.3339,
      "mean_token_accuracy": 0.8844472765922546,
      "num_tokens": 528713185.0,
      "step": 15194
    },
    {
      "epoch": 2.8217270194986073,
      "grad_norm": 1.582931236035226,
      "learning_rate": 1e-06,
      "loss": 0.2875,
      "mean_token_accuracy": 0.8981627821922302,
      "num_tokens": 528742088.0,
      "step": 15195
    },
    {
      "epoch": 2.821912720519963,
      "grad_norm": 1.6029674680095576,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8890595436096191,
      "num_tokens": 528775406.0,
      "step": 15196
    },
    {
      "epoch": 2.8220984215413187,
      "grad_norm": 1.6170853604532986,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8787489533424377,
      "num_tokens": 528809228.0,
      "step": 15197
    },
    {
      "epoch": 2.822284122562674,
      "grad_norm": 1.5029234674017922,
      "learning_rate": 1e-06,
      "loss": 0.3257,
      "mean_token_accuracy": 0.8880131244659424,
      "num_tokens": 528845249.0,
      "step": 15198
    },
    {
      "epoch": 2.8224698235840298,
      "grad_norm": 1.699165658991486,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8906217813491821,
      "num_tokens": 528872187.0,
      "step": 15199
    },
    {
      "epoch": 2.8226555246053855,
      "grad_norm": 1.5516298906679806,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8806310892105103,
      "num_tokens": 528915527.0,
      "step": 15200
    },
    {
      "epoch": 2.822841225626741,
      "grad_norm": 1.5062036606614033,
      "learning_rate": 1e-06,
      "loss": 0.2952,
      "mean_token_accuracy": 0.8936827182769775,
      "num_tokens": 528952021.0,
      "step": 15201
    },
    {
      "epoch": 2.8230269266480965,
      "grad_norm": 1.5930983885758758,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8838822841644287,
      "num_tokens": 528988447.0,
      "step": 15202
    },
    {
      "epoch": 2.8232126276694522,
      "grad_norm": 1.5486458429236583,
      "learning_rate": 1e-06,
      "loss": 0.346,
      "mean_token_accuracy": 0.8799304962158203,
      "num_tokens": 529024712.0,
      "step": 15203
    },
    {
      "epoch": 2.823398328690808,
      "grad_norm": 1.5887051012605935,
      "learning_rate": 1e-06,
      "loss": 0.314,
      "mean_token_accuracy": 0.8903133869171143,
      "num_tokens": 529055953.0,
      "step": 15204
    },
    {
      "epoch": 2.8235840297121633,
      "grad_norm": 1.5021575877138043,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.884503960609436,
      "num_tokens": 529094730.0,
      "step": 15205
    },
    {
      "epoch": 2.823769730733519,
      "grad_norm": 1.5173713554805415,
      "learning_rate": 1e-06,
      "loss": 0.2877,
      "mean_token_accuracy": 0.8976054191589355,
      "num_tokens": 529127169.0,
      "step": 15206
    },
    {
      "epoch": 2.8239554317548747,
      "grad_norm": 1.4620032597221426,
      "learning_rate": 1e-06,
      "loss": 0.2765,
      "mean_token_accuracy": 0.9009078741073608,
      "num_tokens": 529162648.0,
      "step": 15207
    },
    {
      "epoch": 2.82414113277623,
      "grad_norm": 1.5985866835547005,
      "learning_rate": 1e-06,
      "loss": 0.3137,
      "mean_token_accuracy": 0.8922463059425354,
      "num_tokens": 529192530.0,
      "step": 15208
    },
    {
      "epoch": 2.8243268337975858,
      "grad_norm": 1.5266855950950942,
      "learning_rate": 1e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8721361756324768,
      "num_tokens": 529235785.0,
      "step": 15209
    },
    {
      "epoch": 2.8245125348189415,
      "grad_norm": 1.3795111046620037,
      "learning_rate": 1e-06,
      "loss": 0.3049,
      "mean_token_accuracy": 0.8947048187255859,
      "num_tokens": 529280261.0,
      "step": 15210
    },
    {
      "epoch": 2.8246982358402972,
      "grad_norm": 1.660192407486735,
      "learning_rate": 1e-06,
      "loss": 0.2962,
      "mean_token_accuracy": 0.8964796662330627,
      "num_tokens": 529308086.0,
      "step": 15211
    },
    {
      "epoch": 2.824883936861653,
      "grad_norm": 1.5408817628828622,
      "learning_rate": 1e-06,
      "loss": 0.3259,
      "mean_token_accuracy": 0.8863074779510498,
      "num_tokens": 529346781.0,
      "step": 15212
    },
    {
      "epoch": 2.8250696378830082,
      "grad_norm": 1.6165081968392827,
      "learning_rate": 1e-06,
      "loss": 0.309,
      "mean_token_accuracy": 0.8912680149078369,
      "num_tokens": 529378908.0,
      "step": 15213
    },
    {
      "epoch": 2.825255338904364,
      "grad_norm": 1.6029017737410292,
      "learning_rate": 1e-06,
      "loss": 0.3212,
      "mean_token_accuracy": 0.885793924331665,
      "num_tokens": 529413279.0,
      "step": 15214
    },
    {
      "epoch": 2.8254410399257197,
      "grad_norm": 1.5194221669588353,
      "learning_rate": 1e-06,
      "loss": 0.291,
      "mean_token_accuracy": 0.8970764875411987,
      "num_tokens": 529446757.0,
      "step": 15215
    },
    {
      "epoch": 2.825626740947075,
      "grad_norm": 1.5234800962285417,
      "learning_rate": 1e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8670000433921814,
      "num_tokens": 529488348.0,
      "step": 15216
    },
    {
      "epoch": 2.8258124419684307,
      "grad_norm": 1.7116991752964865,
      "learning_rate": 1e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8750172257423401,
      "num_tokens": 529522441.0,
      "step": 15217
    },
    {
      "epoch": 2.8259981429897865,
      "grad_norm": 1.8064687391011272,
      "learning_rate": 1e-06,
      "loss": 0.3421,
      "mean_token_accuracy": 0.8828723430633545,
      "num_tokens": 529550020.0,
      "step": 15218
    },
    {
      "epoch": 2.826183844011142,
      "grad_norm": 1.5503066276843558,
      "learning_rate": 1e-06,
      "loss": 0.3417,
      "mean_token_accuracy": 0.8786102533340454,
      "num_tokens": 529583811.0,
      "step": 15219
    },
    {
      "epoch": 2.826369545032498,
      "grad_norm": 1.5484113406744668,
      "learning_rate": 1e-06,
      "loss": 0.3156,
      "mean_token_accuracy": 0.8879923820495605,
      "num_tokens": 529619056.0,
      "step": 15220
    },
    {
      "epoch": 2.826555246053853,
      "grad_norm": 1.5975597627934077,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.876214325428009,
      "num_tokens": 529652296.0,
      "step": 15221
    },
    {
      "epoch": 2.826740947075209,
      "grad_norm": 1.65121530808787,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8947150111198425,
      "num_tokens": 529687692.0,
      "step": 15222
    },
    {
      "epoch": 2.8269266480965647,
      "grad_norm": 1.502462378675784,
      "learning_rate": 1e-06,
      "loss": 0.3254,
      "mean_token_accuracy": 0.8876902461051941,
      "num_tokens": 529727324.0,
      "step": 15223
    },
    {
      "epoch": 2.82711234911792,
      "grad_norm": 1.5679765195671684,
      "learning_rate": 1e-06,
      "loss": 0.329,
      "mean_token_accuracy": 0.8861537575721741,
      "num_tokens": 529764710.0,
      "step": 15224
    },
    {
      "epoch": 2.8272980501392757,
      "grad_norm": 1.551687704064046,
      "learning_rate": 1e-06,
      "loss": 0.3256,
      "mean_token_accuracy": 0.886284351348877,
      "num_tokens": 529800727.0,
      "step": 15225
    },
    {
      "epoch": 2.8274837511606314,
      "grad_norm": 1.628758554210782,
      "learning_rate": 1e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8736200928688049,
      "num_tokens": 529837472.0,
      "step": 15226
    },
    {
      "epoch": 2.827669452181987,
      "grad_norm": 1.8431591584998286,
      "learning_rate": 1e-06,
      "loss": 0.3173,
      "mean_token_accuracy": 0.8892894983291626,
      "num_tokens": 529861769.0,
      "step": 15227
    },
    {
      "epoch": 2.8278551532033425,
      "grad_norm": 1.5522139414760827,
      "learning_rate": 1e-06,
      "loss": 0.3508,
      "mean_token_accuracy": 0.8783000707626343,
      "num_tokens": 529903611.0,
      "step": 15228
    },
    {
      "epoch": 2.828040854224698,
      "grad_norm": 1.6698774662906626,
      "learning_rate": 1e-06,
      "loss": 0.2797,
      "mean_token_accuracy": 0.8994351625442505,
      "num_tokens": 529931273.0,
      "step": 15229
    },
    {
      "epoch": 2.828226555246054,
      "grad_norm": 1.549571343620335,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8877265453338623,
      "num_tokens": 529964156.0,
      "step": 15230
    },
    {
      "epoch": 2.828412256267409,
      "grad_norm": 1.7482769820181652,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8835586905479431,
      "num_tokens": 529993559.0,
      "step": 15231
    },
    {
      "epoch": 2.828597957288765,
      "grad_norm": 1.5986941220657707,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8822475671768188,
      "num_tokens": 530027815.0,
      "step": 15232
    },
    {
      "epoch": 2.8287836583101207,
      "grad_norm": 1.384856151036573,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8993117213249207,
      "num_tokens": 530067771.0,
      "step": 15233
    },
    {
      "epoch": 2.8289693593314764,
      "grad_norm": 1.7437748294171356,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8793411254882812,
      "num_tokens": 530099351.0,
      "step": 15234
    },
    {
      "epoch": 2.829155060352832,
      "grad_norm": 1.4541357433667799,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.8995533585548401,
      "num_tokens": 530137376.0,
      "step": 15235
    },
    {
      "epoch": 2.8293407613741874,
      "grad_norm": 1.4335994825037122,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8918032646179199,
      "num_tokens": 530178497.0,
      "step": 15236
    },
    {
      "epoch": 2.829526462395543,
      "grad_norm": 1.493515235377546,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8875960111618042,
      "num_tokens": 530215159.0,
      "step": 15237
    },
    {
      "epoch": 2.829712163416899,
      "grad_norm": 1.5843665581453585,
      "learning_rate": 1e-06,
      "loss": 0.321,
      "mean_token_accuracy": 0.8862614035606384,
      "num_tokens": 530249074.0,
      "step": 15238
    },
    {
      "epoch": 2.829897864438254,
      "grad_norm": 1.5244016795530884,
      "learning_rate": 1e-06,
      "loss": 0.3526,
      "mean_token_accuracy": 0.8744670152664185,
      "num_tokens": 530287971.0,
      "step": 15239
    },
    {
      "epoch": 2.83008356545961,
      "grad_norm": 1.6416492976572734,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.882601261138916,
      "num_tokens": 530328524.0,
      "step": 15240
    },
    {
      "epoch": 2.8302692664809657,
      "grad_norm": 1.6380815027861548,
      "learning_rate": 1e-06,
      "loss": 0.3125,
      "mean_token_accuracy": 0.8883362412452698,
      "num_tokens": 530359345.0,
      "step": 15241
    },
    {
      "epoch": 2.8304549675023214,
      "grad_norm": 1.5450874592281378,
      "learning_rate": 1e-06,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8822968006134033,
      "num_tokens": 530397043.0,
      "step": 15242
    },
    {
      "epoch": 2.830640668523677,
      "grad_norm": 1.7911324633856054,
      "learning_rate": 1e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.871023416519165,
      "num_tokens": 530427127.0,
      "step": 15243
    },
    {
      "epoch": 2.8308263695450324,
      "grad_norm": 1.5607429945494504,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8895165324211121,
      "num_tokens": 530459486.0,
      "step": 15244
    },
    {
      "epoch": 2.831012070566388,
      "grad_norm": 1.7745615904746557,
      "learning_rate": 1e-06,
      "loss": 0.3485,
      "mean_token_accuracy": 0.8788712024688721,
      "num_tokens": 530490331.0,
      "step": 15245
    },
    {
      "epoch": 2.831197771587744,
      "grad_norm": 1.398350020768748,
      "learning_rate": 1e-06,
      "loss": 0.2589,
      "mean_token_accuracy": 0.9056875705718994,
      "num_tokens": 530528313.0,
      "step": 15246
    },
    {
      "epoch": 2.831383472609099,
      "grad_norm": 1.5195729944296148,
      "learning_rate": 1e-06,
      "loss": 0.2894,
      "mean_token_accuracy": 0.8987463712692261,
      "num_tokens": 530560979.0,
      "step": 15247
    },
    {
      "epoch": 2.831569173630455,
      "grad_norm": 1.4875921545227746,
      "learning_rate": 1e-06,
      "loss": 0.299,
      "mean_token_accuracy": 0.8968973159790039,
      "num_tokens": 530596484.0,
      "step": 15248
    },
    {
      "epoch": 2.8317548746518106,
      "grad_norm": 1.4387121931682116,
      "learning_rate": 1e-06,
      "loss": 0.2791,
      "mean_token_accuracy": 0.8997941017150879,
      "num_tokens": 530631758.0,
      "step": 15249
    },
    {
      "epoch": 2.8319405756731664,
      "grad_norm": 1.6637176289786948,
      "learning_rate": 1e-06,
      "loss": 0.3289,
      "mean_token_accuracy": 0.8850650787353516,
      "num_tokens": 530660976.0,
      "step": 15250
    },
    {
      "epoch": 2.8321262766945217,
      "grad_norm": 1.6411444811665046,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8937311172485352,
      "num_tokens": 530691044.0,
      "step": 15251
    },
    {
      "epoch": 2.8323119777158774,
      "grad_norm": 1.7055579955681726,
      "learning_rate": 1e-06,
      "loss": 0.354,
      "mean_token_accuracy": 0.8756978511810303,
      "num_tokens": 530722537.0,
      "step": 15252
    },
    {
      "epoch": 2.832497678737233,
      "grad_norm": 1.6129800837587023,
      "learning_rate": 1e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8661973476409912,
      "num_tokens": 530759845.0,
      "step": 15253
    },
    {
      "epoch": 2.8326833797585884,
      "grad_norm": 1.4915598533700831,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8929941058158875,
      "num_tokens": 530794876.0,
      "step": 15254
    },
    {
      "epoch": 2.832869080779944,
      "grad_norm": 1.4527759500318804,
      "learning_rate": 1e-06,
      "loss": 0.3068,
      "mean_token_accuracy": 0.8930601477622986,
      "num_tokens": 530833016.0,
      "step": 15255
    },
    {
      "epoch": 2.8330547818013,
      "grad_norm": 1.6348285168254069,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.8908786773681641,
      "num_tokens": 530867337.0,
      "step": 15256
    },
    {
      "epoch": 2.8332404828226556,
      "grad_norm": 1.4281277100442429,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.8838490843772888,
      "num_tokens": 530907035.0,
      "step": 15257
    },
    {
      "epoch": 2.8334261838440113,
      "grad_norm": 1.5510952614314404,
      "learning_rate": 1e-06,
      "loss": 0.3535,
      "mean_token_accuracy": 0.8770540952682495,
      "num_tokens": 530947508.0,
      "step": 15258
    },
    {
      "epoch": 2.8336118848653666,
      "grad_norm": 1.6571923456524142,
      "learning_rate": 1e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8790930509567261,
      "num_tokens": 530976790.0,
      "step": 15259
    },
    {
      "epoch": 2.8337975858867224,
      "grad_norm": 1.4756333514857962,
      "learning_rate": 1e-06,
      "loss": 0.2473,
      "mean_token_accuracy": 0.912001371383667,
      "num_tokens": 531008242.0,
      "step": 15260
    },
    {
      "epoch": 2.833983286908078,
      "grad_norm": 1.5981334193844297,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8920527696609497,
      "num_tokens": 531037745.0,
      "step": 15261
    },
    {
      "epoch": 2.8341689879294334,
      "grad_norm": 1.5197756817609394,
      "learning_rate": 1e-06,
      "loss": 0.3182,
      "mean_token_accuracy": 0.8894640207290649,
      "num_tokens": 531077177.0,
      "step": 15262
    },
    {
      "epoch": 2.834354688950789,
      "grad_norm": 1.4006891232484195,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8922419548034668,
      "num_tokens": 531115740.0,
      "step": 15263
    },
    {
      "epoch": 2.834540389972145,
      "grad_norm": 1.5101464027200966,
      "learning_rate": 1e-06,
      "loss": 0.3281,
      "mean_token_accuracy": 0.8857619762420654,
      "num_tokens": 531152065.0,
      "step": 15264
    },
    {
      "epoch": 2.8347260909935006,
      "grad_norm": 1.443892705724351,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.885291576385498,
      "num_tokens": 531194078.0,
      "step": 15265
    },
    {
      "epoch": 2.8349117920148563,
      "grad_norm": 1.4792567526365195,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.89593505859375,
      "num_tokens": 531227360.0,
      "step": 15266
    },
    {
      "epoch": 2.8350974930362116,
      "grad_norm": 1.6928495593782842,
      "learning_rate": 1e-06,
      "loss": 0.3377,
      "mean_token_accuracy": 0.8841192126274109,
      "num_tokens": 531259396.0,
      "step": 15267
    },
    {
      "epoch": 2.8352831940575673,
      "grad_norm": 1.654609656447545,
      "learning_rate": 1e-06,
      "loss": 0.3425,
      "mean_token_accuracy": 0.8797587156295776,
      "num_tokens": 531290355.0,
      "step": 15268
    },
    {
      "epoch": 2.835468895078923,
      "grad_norm": 1.5391999799776743,
      "learning_rate": 1e-06,
      "loss": 0.3498,
      "mean_token_accuracy": 0.8769062757492065,
      "num_tokens": 531329163.0,
      "step": 15269
    },
    {
      "epoch": 2.8356545961002784,
      "grad_norm": 1.4987090526388414,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8918846845626831,
      "num_tokens": 531370017.0,
      "step": 15270
    },
    {
      "epoch": 2.835840297121634,
      "grad_norm": 1.5948125925993935,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8841215372085571,
      "num_tokens": 531404284.0,
      "step": 15271
    },
    {
      "epoch": 2.83602599814299,
      "grad_norm": 1.5402961404780928,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8817350268363953,
      "num_tokens": 531439001.0,
      "step": 15272
    },
    {
      "epoch": 2.8362116991643456,
      "grad_norm": 1.8647006385313782,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.8794125914573669,
      "num_tokens": 531462182.0,
      "step": 15273
    },
    {
      "epoch": 2.836397400185701,
      "grad_norm": 1.4190030137176468,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8906130790710449,
      "num_tokens": 531501908.0,
      "step": 15274
    },
    {
      "epoch": 2.8365831012070566,
      "grad_norm": 1.4464665070724056,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8787762522697449,
      "num_tokens": 531544382.0,
      "step": 15275
    },
    {
      "epoch": 2.8367688022284123,
      "grad_norm": 1.5607425005220987,
      "learning_rate": 1e-06,
      "loss": 0.3393,
      "mean_token_accuracy": 0.8821413516998291,
      "num_tokens": 531579500.0,
      "step": 15276
    },
    {
      "epoch": 2.8369545032497676,
      "grad_norm": 1.5943645238533484,
      "learning_rate": 1e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8740944862365723,
      "num_tokens": 531615948.0,
      "step": 15277
    },
    {
      "epoch": 2.8371402042711233,
      "grad_norm": 1.6897538059811072,
      "learning_rate": 1e-06,
      "loss": 0.2743,
      "mean_token_accuracy": 0.8991426229476929,
      "num_tokens": 531642605.0,
      "step": 15278
    },
    {
      "epoch": 2.837325905292479,
      "grad_norm": 1.5956397215339506,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8939816951751709,
      "num_tokens": 531676478.0,
      "step": 15279
    },
    {
      "epoch": 2.837511606313835,
      "grad_norm": 1.4124317542399765,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8908964395523071,
      "num_tokens": 531715848.0,
      "step": 15280
    },
    {
      "epoch": 2.8376973073351905,
      "grad_norm": 1.4506391785807407,
      "learning_rate": 1e-06,
      "loss": 0.3383,
      "mean_token_accuracy": 0.8833131790161133,
      "num_tokens": 531756416.0,
      "step": 15281
    },
    {
      "epoch": 2.837883008356546,
      "grad_norm": 1.5217021582896262,
      "learning_rate": 1e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8749974966049194,
      "num_tokens": 531795914.0,
      "step": 15282
    },
    {
      "epoch": 2.8380687093779016,
      "grad_norm": 1.6025175348487775,
      "learning_rate": 1e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.886039137840271,
      "num_tokens": 531830159.0,
      "step": 15283
    },
    {
      "epoch": 2.8382544103992573,
      "grad_norm": 1.5234674248790705,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8872368335723877,
      "num_tokens": 531865333.0,
      "step": 15284
    },
    {
      "epoch": 2.8384401114206126,
      "grad_norm": 1.7721931820894143,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.888069748878479,
      "num_tokens": 531892429.0,
      "step": 15285
    },
    {
      "epoch": 2.8386258124419683,
      "grad_norm": 1.491478272554144,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8809758424758911,
      "num_tokens": 531931774.0,
      "step": 15286
    },
    {
      "epoch": 2.838811513463324,
      "grad_norm": 1.5619206570728366,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8949165344238281,
      "num_tokens": 531963538.0,
      "step": 15287
    },
    {
      "epoch": 2.83899721448468,
      "grad_norm": 1.4549653587281484,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8852250576019287,
      "num_tokens": 532004043.0,
      "step": 15288
    },
    {
      "epoch": 2.8391829155060355,
      "grad_norm": 1.5954889990839956,
      "learning_rate": 1e-06,
      "loss": 0.2961,
      "mean_token_accuracy": 0.8940446376800537,
      "num_tokens": 532038319.0,
      "step": 15289
    },
    {
      "epoch": 2.839368616527391,
      "grad_norm": 1.6941070307974804,
      "learning_rate": 1e-06,
      "loss": 0.3032,
      "mean_token_accuracy": 0.8973232507705688,
      "num_tokens": 532066742.0,
      "step": 15290
    },
    {
      "epoch": 2.8395543175487465,
      "grad_norm": 1.5776294357794718,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8743077516555786,
      "num_tokens": 532105498.0,
      "step": 15291
    },
    {
      "epoch": 2.8397400185701023,
      "grad_norm": 1.5661050159020347,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.8767728805541992,
      "num_tokens": 532143530.0,
      "step": 15292
    },
    {
      "epoch": 2.8399257195914576,
      "grad_norm": 1.5608757572026535,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8817417621612549,
      "num_tokens": 532178713.0,
      "step": 15293
    },
    {
      "epoch": 2.8401114206128133,
      "grad_norm": 1.581626780906839,
      "learning_rate": 1e-06,
      "loss": 0.2985,
      "mean_token_accuracy": 0.8943331241607666,
      "num_tokens": 532213657.0,
      "step": 15294
    },
    {
      "epoch": 2.840297121634169,
      "grad_norm": 1.5651445705482645,
      "learning_rate": 1e-06,
      "loss": 0.3165,
      "mean_token_accuracy": 0.8889344930648804,
      "num_tokens": 532250086.0,
      "step": 15295
    },
    {
      "epoch": 2.8404828226555248,
      "grad_norm": 1.5679005930485639,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8846982717514038,
      "num_tokens": 532286144.0,
      "step": 15296
    },
    {
      "epoch": 2.8406685236768805,
      "grad_norm": 1.5392154428214901,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8796588778495789,
      "num_tokens": 532323454.0,
      "step": 15297
    },
    {
      "epoch": 2.840854224698236,
      "grad_norm": 1.6151450324020427,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8826813101768494,
      "num_tokens": 532361229.0,
      "step": 15298
    },
    {
      "epoch": 2.8410399257195915,
      "grad_norm": 1.6742890741140903,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8704915046691895,
      "num_tokens": 532395323.0,
      "step": 15299
    },
    {
      "epoch": 2.841225626740947,
      "grad_norm": 1.4731103930318572,
      "learning_rate": 1e-06,
      "loss": 0.3201,
      "mean_token_accuracy": 0.8876592516899109,
      "num_tokens": 532432836.0,
      "step": 15300
    },
    {
      "epoch": 2.8414113277623025,
      "grad_norm": 1.6378793779291654,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8773798942565918,
      "num_tokens": 532465154.0,
      "step": 15301
    },
    {
      "epoch": 2.8415970287836583,
      "grad_norm": 1.554496220127855,
      "learning_rate": 1e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.868171751499176,
      "num_tokens": 532505624.0,
      "step": 15302
    },
    {
      "epoch": 2.841782729805014,
      "grad_norm": 1.448579121809187,
      "learning_rate": 1e-06,
      "loss": 0.316,
      "mean_token_accuracy": 0.8909118175506592,
      "num_tokens": 532544868.0,
      "step": 15303
    },
    {
      "epoch": 2.8419684308263697,
      "grad_norm": 1.5470530029152112,
      "learning_rate": 1e-06,
      "loss": 0.3288,
      "mean_token_accuracy": 0.88514643907547,
      "num_tokens": 532580372.0,
      "step": 15304
    },
    {
      "epoch": 2.842154131847725,
      "grad_norm": 1.5838066930720407,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.8974896669387817,
      "num_tokens": 532611767.0,
      "step": 15305
    },
    {
      "epoch": 2.8423398328690808,
      "grad_norm": 1.4093489216694328,
      "learning_rate": 1e-06,
      "loss": 0.2891,
      "mean_token_accuracy": 0.896242618560791,
      "num_tokens": 532655674.0,
      "step": 15306
    },
    {
      "epoch": 2.8425255338904365,
      "grad_norm": 1.7022434986401107,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8702645301818848,
      "num_tokens": 532686538.0,
      "step": 15307
    },
    {
      "epoch": 2.842711234911792,
      "grad_norm": 1.4678857517875032,
      "learning_rate": 1e-06,
      "loss": 0.2859,
      "mean_token_accuracy": 0.8973156213760376,
      "num_tokens": 532721911.0,
      "step": 15308
    },
    {
      "epoch": 2.8428969359331475,
      "grad_norm": 1.5590324167187721,
      "learning_rate": 1e-06,
      "loss": 0.3505,
      "mean_token_accuracy": 0.8758734464645386,
      "num_tokens": 532758829.0,
      "step": 15309
    },
    {
      "epoch": 2.8430826369545033,
      "grad_norm": 1.502162262782164,
      "learning_rate": 1e-06,
      "loss": 0.3076,
      "mean_token_accuracy": 0.8922240734100342,
      "num_tokens": 532792535.0,
      "step": 15310
    },
    {
      "epoch": 2.843268337975859,
      "grad_norm": 1.6213949020466538,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8921931982040405,
      "num_tokens": 532823801.0,
      "step": 15311
    },
    {
      "epoch": 2.8434540389972147,
      "grad_norm": 1.8047404046344118,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8755265474319458,
      "num_tokens": 532854612.0,
      "step": 15312
    },
    {
      "epoch": 2.84363974001857,
      "grad_norm": 1.458339981650249,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8830081224441528,
      "num_tokens": 532893970.0,
      "step": 15313
    },
    {
      "epoch": 2.8438254410399257,
      "grad_norm": 1.6085775782907021,
      "learning_rate": 1e-06,
      "loss": 0.3033,
      "mean_token_accuracy": 0.8916577100753784,
      "num_tokens": 532926559.0,
      "step": 15314
    },
    {
      "epoch": 2.8440111420612815,
      "grad_norm": 1.6755369115005325,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8933005928993225,
      "num_tokens": 532956815.0,
      "step": 15315
    },
    {
      "epoch": 2.8441968430826368,
      "grad_norm": 1.482824022352518,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.8961002230644226,
      "num_tokens": 532993604.0,
      "step": 15316
    },
    {
      "epoch": 2.8443825441039925,
      "grad_norm": 1.5129209552989447,
      "learning_rate": 1e-06,
      "loss": 0.3218,
      "mean_token_accuracy": 0.8874597549438477,
      "num_tokens": 533031151.0,
      "step": 15317
    },
    {
      "epoch": 2.8445682451253482,
      "grad_norm": 1.7905497543944986,
      "learning_rate": 1e-06,
      "loss": 0.3465,
      "mean_token_accuracy": 0.877159833908081,
      "num_tokens": 533059194.0,
      "step": 15318
    },
    {
      "epoch": 2.844753946146704,
      "grad_norm": 1.7874734686625724,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8753348588943481,
      "num_tokens": 533088954.0,
      "step": 15319
    },
    {
      "epoch": 2.8449396471680597,
      "grad_norm": 1.5441185362290246,
      "learning_rate": 1e-06,
      "loss": 0.2942,
      "mean_token_accuracy": 0.893717885017395,
      "num_tokens": 533121830.0,
      "step": 15320
    },
    {
      "epoch": 2.845125348189415,
      "grad_norm": 1.5045973211372008,
      "learning_rate": 1e-06,
      "loss": 0.2687,
      "mean_token_accuracy": 0.902158260345459,
      "num_tokens": 533152522.0,
      "step": 15321
    },
    {
      "epoch": 2.8453110492107707,
      "grad_norm": 1.610505847773691,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.892507791519165,
      "num_tokens": 533187728.0,
      "step": 15322
    },
    {
      "epoch": 2.845496750232126,
      "grad_norm": 1.5409460208334518,
      "learning_rate": 1e-06,
      "loss": 0.3192,
      "mean_token_accuracy": 0.8882881999015808,
      "num_tokens": 533225957.0,
      "step": 15323
    },
    {
      "epoch": 2.8456824512534817,
      "grad_norm": 1.5544807536333705,
      "learning_rate": 1e-06,
      "loss": 0.3064,
      "mean_token_accuracy": 0.8926447629928589,
      "num_tokens": 533259389.0,
      "step": 15324
    },
    {
      "epoch": 2.8458681522748375,
      "grad_norm": 1.4912391338481394,
      "learning_rate": 1e-06,
      "loss": 0.2916,
      "mean_token_accuracy": 0.8983875513076782,
      "num_tokens": 533294895.0,
      "step": 15325
    },
    {
      "epoch": 2.846053853296193,
      "grad_norm": 1.5895758357062848,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8928245306015015,
      "num_tokens": 533327563.0,
      "step": 15326
    },
    {
      "epoch": 2.846239554317549,
      "grad_norm": 1.5967572844951652,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.8922137022018433,
      "num_tokens": 533359025.0,
      "step": 15327
    },
    {
      "epoch": 2.8464252553389042,
      "grad_norm": 1.6901685805552227,
      "learning_rate": 1e-06,
      "loss": 0.2843,
      "mean_token_accuracy": 0.8995590209960938,
      "num_tokens": 533388497.0,
      "step": 15328
    },
    {
      "epoch": 2.84661095636026,
      "grad_norm": 1.4630881282395665,
      "learning_rate": 1e-06,
      "loss": 0.2795,
      "mean_token_accuracy": 0.9028276205062866,
      "num_tokens": 533424474.0,
      "step": 15329
    },
    {
      "epoch": 2.8467966573816157,
      "grad_norm": 1.3875910972793977,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8890337944030762,
      "num_tokens": 533466884.0,
      "step": 15330
    },
    {
      "epoch": 2.846982358402971,
      "grad_norm": 1.5939371615323836,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8846002221107483,
      "num_tokens": 533501009.0,
      "step": 15331
    },
    {
      "epoch": 2.8471680594243267,
      "grad_norm": 1.6589535259667765,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8863955140113831,
      "num_tokens": 533533594.0,
      "step": 15332
    },
    {
      "epoch": 2.8473537604456824,
      "grad_norm": 1.519703840168509,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.889793872833252,
      "num_tokens": 533569960.0,
      "step": 15333
    },
    {
      "epoch": 2.847539461467038,
      "grad_norm": 1.6120044643171383,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8769296407699585,
      "num_tokens": 533608587.0,
      "step": 15334
    },
    {
      "epoch": 2.847725162488394,
      "grad_norm": 1.4662294980087864,
      "learning_rate": 1e-06,
      "loss": 0.3092,
      "mean_token_accuracy": 0.8910319209098816,
      "num_tokens": 533645623.0,
      "step": 15335
    },
    {
      "epoch": 2.847910863509749,
      "grad_norm": 1.7884057005980345,
      "learning_rate": 1e-06,
      "loss": 0.3242,
      "mean_token_accuracy": 0.8836774826049805,
      "num_tokens": 533671745.0,
      "step": 15336
    },
    {
      "epoch": 2.848096564531105,
      "grad_norm": 1.4108793305275262,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.8977682590484619,
      "num_tokens": 533709014.0,
      "step": 15337
    },
    {
      "epoch": 2.8482822655524607,
      "grad_norm": 1.5754172969312599,
      "learning_rate": 1e-06,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8788250088691711,
      "num_tokens": 533744219.0,
      "step": 15338
    },
    {
      "epoch": 2.848467966573816,
      "grad_norm": 1.5612355952238954,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8971319794654846,
      "num_tokens": 533778311.0,
      "step": 15339
    },
    {
      "epoch": 2.8486536675951717,
      "grad_norm": 1.7156158712511376,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8905045986175537,
      "num_tokens": 533808764.0,
      "step": 15340
    },
    {
      "epoch": 2.8488393686165274,
      "grad_norm": 1.6083570031382322,
      "learning_rate": 1e-06,
      "loss": 0.3442,
      "mean_token_accuracy": 0.8809818029403687,
      "num_tokens": 533841625.0,
      "step": 15341
    },
    {
      "epoch": 2.849025069637883,
      "grad_norm": 1.5759238930734114,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8736572265625,
      "num_tokens": 533880539.0,
      "step": 15342
    },
    {
      "epoch": 2.849210770659239,
      "grad_norm": 1.6290319897154215,
      "learning_rate": 1e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8744513988494873,
      "num_tokens": 533914276.0,
      "step": 15343
    },
    {
      "epoch": 2.849396471680594,
      "grad_norm": 1.604792515946915,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.883222222328186,
      "num_tokens": 533948233.0,
      "step": 15344
    },
    {
      "epoch": 2.84958217270195,
      "grad_norm": 1.687050838927038,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8857744932174683,
      "num_tokens": 533981693.0,
      "step": 15345
    },
    {
      "epoch": 2.8497678737233056,
      "grad_norm": 1.455563826094962,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8897144198417664,
      "num_tokens": 534019859.0,
      "step": 15346
    },
    {
      "epoch": 2.849953574744661,
      "grad_norm": 1.77530196433413,
      "learning_rate": 1e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.874962568283081,
      "num_tokens": 534050855.0,
      "step": 15347
    },
    {
      "epoch": 2.8501392757660167,
      "grad_norm": 1.6211842211751417,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8891788721084595,
      "num_tokens": 534081414.0,
      "step": 15348
    },
    {
      "epoch": 2.8503249767873724,
      "grad_norm": 1.560138692241177,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8864959478378296,
      "num_tokens": 534115299.0,
      "step": 15349
    },
    {
      "epoch": 2.850510677808728,
      "grad_norm": 1.5356011320090923,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8887057304382324,
      "num_tokens": 534149042.0,
      "step": 15350
    },
    {
      "epoch": 2.8506963788300834,
      "grad_norm": 1.7379119578737212,
      "learning_rate": 1e-06,
      "loss": 0.2892,
      "mean_token_accuracy": 0.8978856801986694,
      "num_tokens": 534176746.0,
      "step": 15351
    },
    {
      "epoch": 2.850882079851439,
      "grad_norm": 1.6198029419765525,
      "learning_rate": 1e-06,
      "loss": 0.2937,
      "mean_token_accuracy": 0.8922411799430847,
      "num_tokens": 534206640.0,
      "step": 15352
    },
    {
      "epoch": 2.851067780872795,
      "grad_norm": 1.6966124534796851,
      "learning_rate": 1e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8629084825515747,
      "num_tokens": 534241101.0,
      "step": 15353
    },
    {
      "epoch": 2.85125348189415,
      "grad_norm": 1.5603322077830901,
      "learning_rate": 1e-06,
      "loss": 0.3003,
      "mean_token_accuracy": 0.8929459452629089,
      "num_tokens": 534277218.0,
      "step": 15354
    },
    {
      "epoch": 2.851439182915506,
      "grad_norm": 1.6561023341799306,
      "learning_rate": 1e-06,
      "loss": 0.3085,
      "mean_token_accuracy": 0.8892250061035156,
      "num_tokens": 534312877.0,
      "step": 15355
    },
    {
      "epoch": 2.8516248839368616,
      "grad_norm": 1.5961599842031333,
      "learning_rate": 1e-06,
      "loss": 0.3702,
      "mean_token_accuracy": 0.873141884803772,
      "num_tokens": 534350110.0,
      "step": 15356
    },
    {
      "epoch": 2.8518105849582174,
      "grad_norm": 1.559684003082245,
      "learning_rate": 1e-06,
      "loss": 0.2828,
      "mean_token_accuracy": 0.897502064704895,
      "num_tokens": 534383352.0,
      "step": 15357
    },
    {
      "epoch": 2.851996285979573,
      "grad_norm": 1.4624462219696646,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.8855195045471191,
      "num_tokens": 534423711.0,
      "step": 15358
    },
    {
      "epoch": 2.8521819870009284,
      "grad_norm": 1.5819187654099074,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8877444863319397,
      "num_tokens": 534462049.0,
      "step": 15359
    },
    {
      "epoch": 2.852367688022284,
      "grad_norm": 1.4078903214799314,
      "learning_rate": 1e-06,
      "loss": 0.3433,
      "mean_token_accuracy": 0.8807679414749146,
      "num_tokens": 534505718.0,
      "step": 15360
    },
    {
      "epoch": 2.85255338904364,
      "grad_norm": 1.71415469921696,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8809930086135864,
      "num_tokens": 534538156.0,
      "step": 15361
    },
    {
      "epoch": 2.852739090064995,
      "grad_norm": 1.697143236378635,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8936227560043335,
      "num_tokens": 534565135.0,
      "step": 15362
    },
    {
      "epoch": 2.852924791086351,
      "grad_norm": 1.485418141003352,
      "learning_rate": 1e-06,
      "loss": 0.3439,
      "mean_token_accuracy": 0.8817765712738037,
      "num_tokens": 534604320.0,
      "step": 15363
    },
    {
      "epoch": 2.8531104921077066,
      "grad_norm": 1.5875086880534206,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8802262544631958,
      "num_tokens": 534641090.0,
      "step": 15364
    },
    {
      "epoch": 2.8532961931290624,
      "grad_norm": 1.499145397092797,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8798013925552368,
      "num_tokens": 534681137.0,
      "step": 15365
    },
    {
      "epoch": 2.853481894150418,
      "grad_norm": 1.659468823241102,
      "learning_rate": 1e-06,
      "loss": 0.2949,
      "mean_token_accuracy": 0.8993421792984009,
      "num_tokens": 534710763.0,
      "step": 15366
    },
    {
      "epoch": 2.8536675951717734,
      "grad_norm": 1.597219620256435,
      "learning_rate": 1e-06,
      "loss": 0.3297,
      "mean_token_accuracy": 0.8829219341278076,
      "num_tokens": 534748112.0,
      "step": 15367
    },
    {
      "epoch": 2.853853296193129,
      "grad_norm": 1.6625826389826102,
      "learning_rate": 1e-06,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8824093341827393,
      "num_tokens": 534783349.0,
      "step": 15368
    },
    {
      "epoch": 2.854038997214485,
      "grad_norm": 1.7058735979323083,
      "learning_rate": 1e-06,
      "loss": 0.3349,
      "mean_token_accuracy": 0.8842595815658569,
      "num_tokens": 534814586.0,
      "step": 15369
    },
    {
      "epoch": 2.85422469823584,
      "grad_norm": 1.5834132007628539,
      "learning_rate": 1e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8963696360588074,
      "num_tokens": 534845999.0,
      "step": 15370
    },
    {
      "epoch": 2.854410399257196,
      "grad_norm": 1.5455026238160066,
      "learning_rate": 1e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8826040029525757,
      "num_tokens": 534884194.0,
      "step": 15371
    },
    {
      "epoch": 2.8545961002785516,
      "grad_norm": 1.8621638262847935,
      "learning_rate": 1e-06,
      "loss": 0.351,
      "mean_token_accuracy": 0.8780108690261841,
      "num_tokens": 534913742.0,
      "step": 15372
    },
    {
      "epoch": 2.8547818012999073,
      "grad_norm": 1.648033995362873,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.8948098421096802,
      "num_tokens": 534945362.0,
      "step": 15373
    },
    {
      "epoch": 2.8549675023212626,
      "grad_norm": 1.5847985825641713,
      "learning_rate": 1e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8750749826431274,
      "num_tokens": 534984486.0,
      "step": 15374
    },
    {
      "epoch": 2.8551532033426184,
      "grad_norm": 1.5295998388646372,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8738099336624146,
      "num_tokens": 535024720.0,
      "step": 15375
    },
    {
      "epoch": 2.855338904363974,
      "grad_norm": 1.4942032666143823,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8763337731361389,
      "num_tokens": 535064571.0,
      "step": 15376
    },
    {
      "epoch": 2.8555246053853294,
      "grad_norm": 1.5110014489379804,
      "learning_rate": 1e-06,
      "loss": 0.3479,
      "mean_token_accuracy": 0.877196729183197,
      "num_tokens": 535103859.0,
      "step": 15377
    },
    {
      "epoch": 2.855710306406685,
      "grad_norm": 1.5498194641933083,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8909159898757935,
      "num_tokens": 535141868.0,
      "step": 15378
    },
    {
      "epoch": 2.855896007428041,
      "grad_norm": 1.6635401758991435,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8730562329292297,
      "num_tokens": 535176383.0,
      "step": 15379
    },
    {
      "epoch": 2.8560817084493966,
      "grad_norm": 1.4903116731873656,
      "learning_rate": 1e-06,
      "loss": 0.327,
      "mean_token_accuracy": 0.8861150741577148,
      "num_tokens": 535215964.0,
      "step": 15380
    },
    {
      "epoch": 2.8562674094707523,
      "grad_norm": 1.4540046328909988,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8846855163574219,
      "num_tokens": 535257281.0,
      "step": 15381
    },
    {
      "epoch": 2.8564531104921076,
      "grad_norm": 1.4739006598265312,
      "learning_rate": 1e-06,
      "loss": 0.3071,
      "mean_token_accuracy": 0.8906422853469849,
      "num_tokens": 535294867.0,
      "step": 15382
    },
    {
      "epoch": 2.8566388115134633,
      "grad_norm": 1.6755051616444987,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8870398998260498,
      "num_tokens": 535326038.0,
      "step": 15383
    },
    {
      "epoch": 2.856824512534819,
      "grad_norm": 1.644542666726892,
      "learning_rate": 1e-06,
      "loss": 0.3236,
      "mean_token_accuracy": 0.885600209236145,
      "num_tokens": 535357162.0,
      "step": 15384
    },
    {
      "epoch": 2.8570102135561743,
      "grad_norm": 1.7024128438248625,
      "learning_rate": 1e-06,
      "loss": 0.3221,
      "mean_token_accuracy": 0.8863668441772461,
      "num_tokens": 535384478.0,
      "step": 15385
    },
    {
      "epoch": 2.85719591457753,
      "grad_norm": 1.644462335853995,
      "learning_rate": 1e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.8860087990760803,
      "num_tokens": 535422021.0,
      "step": 15386
    },
    {
      "epoch": 2.857381615598886,
      "grad_norm": 1.5085118392030779,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.8879042863845825,
      "num_tokens": 535461129.0,
      "step": 15387
    },
    {
      "epoch": 2.8575673166202415,
      "grad_norm": 1.6048087965389102,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8854327201843262,
      "num_tokens": 535494751.0,
      "step": 15388
    },
    {
      "epoch": 2.8577530176415973,
      "grad_norm": 1.5446662075547732,
      "learning_rate": 1e-06,
      "loss": 0.3008,
      "mean_token_accuracy": 0.8931774497032166,
      "num_tokens": 535528689.0,
      "step": 15389
    },
    {
      "epoch": 2.8579387186629526,
      "grad_norm": 1.6660670506562094,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.8775098323822021,
      "num_tokens": 535560562.0,
      "step": 15390
    },
    {
      "epoch": 2.8581244196843083,
      "grad_norm": 1.7966736377576131,
      "learning_rate": 1e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.883186936378479,
      "num_tokens": 535592348.0,
      "step": 15391
    },
    {
      "epoch": 2.858310120705664,
      "grad_norm": 1.6451265578019123,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8968995809555054,
      "num_tokens": 535621750.0,
      "step": 15392
    },
    {
      "epoch": 2.8584958217270193,
      "grad_norm": 1.6069565342841647,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8942884206771851,
      "num_tokens": 535655263.0,
      "step": 15393
    },
    {
      "epoch": 2.858681522748375,
      "grad_norm": 1.5603468652764458,
      "learning_rate": 1e-06,
      "loss": 0.2759,
      "mean_token_accuracy": 0.9034842848777771,
      "num_tokens": 535686623.0,
      "step": 15394
    },
    {
      "epoch": 2.858867223769731,
      "grad_norm": 1.5586779857481807,
      "learning_rate": 1e-06,
      "loss": 0.2772,
      "mean_token_accuracy": 0.9012488722801208,
      "num_tokens": 535718786.0,
      "step": 15395
    },
    {
      "epoch": 2.8590529247910865,
      "grad_norm": 1.6481841138349955,
      "learning_rate": 1e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.8802611827850342,
      "num_tokens": 535752691.0,
      "step": 15396
    },
    {
      "epoch": 2.859238625812442,
      "grad_norm": 1.6381021814305428,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8757429718971252,
      "num_tokens": 535787312.0,
      "step": 15397
    },
    {
      "epoch": 2.8594243268337975,
      "grad_norm": 1.4481106209266847,
      "learning_rate": 1e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.875758707523346,
      "num_tokens": 535827692.0,
      "step": 15398
    },
    {
      "epoch": 2.8596100278551533,
      "grad_norm": 1.4831032344328994,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8923208117485046,
      "num_tokens": 535864694.0,
      "step": 15399
    },
    {
      "epoch": 2.8597957288765086,
      "grad_norm": 1.6216582798291064,
      "learning_rate": 1e-06,
      "loss": 0.3639,
      "mean_token_accuracy": 0.8749391436576843,
      "num_tokens": 535900568.0,
      "step": 15400
    },
    {
      "epoch": 2.8599814298978643,
      "grad_norm": 1.6232273374360926,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8845082521438599,
      "num_tokens": 535931370.0,
      "step": 15401
    },
    {
      "epoch": 2.86016713091922,
      "grad_norm": 1.4252043378430117,
      "learning_rate": 1e-06,
      "loss": 0.3241,
      "mean_token_accuracy": 0.884477972984314,
      "num_tokens": 535970700.0,
      "step": 15402
    },
    {
      "epoch": 2.8603528319405758,
      "grad_norm": 1.6243005059332862,
      "learning_rate": 1e-06,
      "loss": 0.3086,
      "mean_token_accuracy": 0.8912470936775208,
      "num_tokens": 536005955.0,
      "step": 15403
    },
    {
      "epoch": 2.8605385329619315,
      "grad_norm": 1.5326684285893653,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8843947649002075,
      "num_tokens": 536043890.0,
      "step": 15404
    },
    {
      "epoch": 2.860724233983287,
      "grad_norm": 1.4462472886810929,
      "learning_rate": 1e-06,
      "loss": 0.2887,
      "mean_token_accuracy": 0.8981794118881226,
      "num_tokens": 536081439.0,
      "step": 15405
    },
    {
      "epoch": 2.8609099350046425,
      "grad_norm": 1.644871836624712,
      "learning_rate": 1e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.872848391532898,
      "num_tokens": 536117014.0,
      "step": 15406
    },
    {
      "epoch": 2.8610956360259983,
      "grad_norm": 1.5903445535565843,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8759410381317139,
      "num_tokens": 536156130.0,
      "step": 15407
    },
    {
      "epoch": 2.8612813370473535,
      "grad_norm": 1.4086008022976206,
      "learning_rate": 1e-06,
      "loss": 0.3142,
      "mean_token_accuracy": 0.8902607560157776,
      "num_tokens": 536199687.0,
      "step": 15408
    },
    {
      "epoch": 2.8614670380687093,
      "grad_norm": 1.5712506793700056,
      "learning_rate": 1e-06,
      "loss": 0.2984,
      "mean_token_accuracy": 0.8961621522903442,
      "num_tokens": 536232750.0,
      "step": 15409
    },
    {
      "epoch": 2.861652739090065,
      "grad_norm": 1.4846301044224532,
      "learning_rate": 1e-06,
      "loss": 0.3255,
      "mean_token_accuracy": 0.8862647414207458,
      "num_tokens": 536271743.0,
      "step": 15410
    },
    {
      "epoch": 2.8618384401114207,
      "grad_norm": 1.6186077536632657,
      "learning_rate": 1e-06,
      "loss": 0.3096,
      "mean_token_accuracy": 0.8907254934310913,
      "num_tokens": 536301577.0,
      "step": 15411
    },
    {
      "epoch": 2.8620241411327765,
      "grad_norm": 1.5421615166009603,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8840394020080566,
      "num_tokens": 536334381.0,
      "step": 15412
    },
    {
      "epoch": 2.8622098421541318,
      "grad_norm": 1.4760534989368645,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8842118978500366,
      "num_tokens": 536368362.0,
      "step": 15413
    },
    {
      "epoch": 2.8623955431754875,
      "grad_norm": 1.8445527726044193,
      "learning_rate": 1e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8695328235626221,
      "num_tokens": 536396707.0,
      "step": 15414
    },
    {
      "epoch": 2.8625812441968432,
      "grad_norm": 1.5695159744855725,
      "learning_rate": 1e-06,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8912034630775452,
      "num_tokens": 536430321.0,
      "step": 15415
    },
    {
      "epoch": 2.8627669452181985,
      "grad_norm": 1.604851085848725,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8842900395393372,
      "num_tokens": 536463766.0,
      "step": 15416
    },
    {
      "epoch": 2.8629526462395543,
      "grad_norm": 1.5308391071106513,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8931199908256531,
      "num_tokens": 536500331.0,
      "step": 15417
    },
    {
      "epoch": 2.86313834726091,
      "grad_norm": 1.4813238197586085,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8935112953186035,
      "num_tokens": 536535116.0,
      "step": 15418
    },
    {
      "epoch": 2.8633240482822657,
      "grad_norm": 1.7847892173474706,
      "learning_rate": 1e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.881034791469574,
      "num_tokens": 536562655.0,
      "step": 15419
    },
    {
      "epoch": 2.863509749303621,
      "grad_norm": 1.550468818616257,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8825333714485168,
      "num_tokens": 536597113.0,
      "step": 15420
    },
    {
      "epoch": 2.8636954503249767,
      "grad_norm": 1.6614642745592012,
      "learning_rate": 1e-06,
      "loss": 0.3682,
      "mean_token_accuracy": 0.8745949864387512,
      "num_tokens": 536630831.0,
      "step": 15421
    },
    {
      "epoch": 2.8638811513463325,
      "grad_norm": 1.655681358183808,
      "learning_rate": 1e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.8744790554046631,
      "num_tokens": 536662960.0,
      "step": 15422
    },
    {
      "epoch": 2.8640668523676878,
      "grad_norm": 1.5156827496345737,
      "learning_rate": 1e-06,
      "loss": 0.2763,
      "mean_token_accuracy": 0.9020767211914062,
      "num_tokens": 536698383.0,
      "step": 15423
    },
    {
      "epoch": 2.8642525533890435,
      "grad_norm": 1.5959870558743616,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.8845627307891846,
      "num_tokens": 536731181.0,
      "step": 15424
    },
    {
      "epoch": 2.8644382544103992,
      "grad_norm": 1.8154322053228786,
      "learning_rate": 1e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8814252614974976,
      "num_tokens": 536759180.0,
      "step": 15425
    },
    {
      "epoch": 2.864623955431755,
      "grad_norm": 1.5869038445094983,
      "learning_rate": 1e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.881470799446106,
      "num_tokens": 536793238.0,
      "step": 15426
    },
    {
      "epoch": 2.8648096564531107,
      "grad_norm": 1.5007856801078705,
      "learning_rate": 1e-06,
      "loss": 0.277,
      "mean_token_accuracy": 0.899603009223938,
      "num_tokens": 536828778.0,
      "step": 15427
    },
    {
      "epoch": 2.864995357474466,
      "grad_norm": 1.5449251319005783,
      "learning_rate": 1e-06,
      "loss": 0.2912,
      "mean_token_accuracy": 0.8959699869155884,
      "num_tokens": 536861017.0,
      "step": 15428
    },
    {
      "epoch": 2.8651810584958217,
      "grad_norm": 1.4778039135005192,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8867849111557007,
      "num_tokens": 536899647.0,
      "step": 15429
    },
    {
      "epoch": 2.8653667595171775,
      "grad_norm": 1.4739491918926646,
      "learning_rate": 1e-06,
      "loss": 0.2647,
      "mean_token_accuracy": 0.9020471572875977,
      "num_tokens": 536931735.0,
      "step": 15430
    },
    {
      "epoch": 2.8655524605385327,
      "grad_norm": 1.6090240378595662,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8849839568138123,
      "num_tokens": 536967994.0,
      "step": 15431
    },
    {
      "epoch": 2.8657381615598885,
      "grad_norm": 1.4597776301864853,
      "learning_rate": 1e-06,
      "loss": 0.2899,
      "mean_token_accuracy": 0.8966662883758545,
      "num_tokens": 537006230.0,
      "step": 15432
    },
    {
      "epoch": 2.865923862581244,
      "grad_norm": 1.666945080095619,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.8819388747215271,
      "num_tokens": 537037977.0,
      "step": 15433
    },
    {
      "epoch": 2.8661095636026,
      "grad_norm": 1.6587409469096421,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.8852812647819519,
      "num_tokens": 537070246.0,
      "step": 15434
    },
    {
      "epoch": 2.8662952646239557,
      "grad_norm": 1.5812886608612178,
      "learning_rate": 1e-06,
      "loss": 0.2701,
      "mean_token_accuracy": 0.9017773270606995,
      "num_tokens": 537102386.0,
      "step": 15435
    },
    {
      "epoch": 2.866480965645311,
      "grad_norm": 1.629985909848331,
      "learning_rate": 1e-06,
      "loss": 0.3522,
      "mean_token_accuracy": 0.8766890168190002,
      "num_tokens": 537134250.0,
      "step": 15436
    },
    {
      "epoch": 2.8666666666666667,
      "grad_norm": 1.5430161109083118,
      "learning_rate": 1e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8848978281021118,
      "num_tokens": 537168483.0,
      "step": 15437
    },
    {
      "epoch": 2.8668523676880224,
      "grad_norm": 1.5365800279537987,
      "learning_rate": 1e-06,
      "loss": 0.2932,
      "mean_token_accuracy": 0.8960156440734863,
      "num_tokens": 537204307.0,
      "step": 15438
    },
    {
      "epoch": 2.8670380687093777,
      "grad_norm": 1.6288558917023988,
      "learning_rate": 1e-06,
      "loss": 0.2843,
      "mean_token_accuracy": 0.8971073627471924,
      "num_tokens": 537238299.0,
      "step": 15439
    },
    {
      "epoch": 2.8672237697307335,
      "grad_norm": 1.6789090167622467,
      "learning_rate": 1e-06,
      "loss": 0.3418,
      "mean_token_accuracy": 0.882456362247467,
      "num_tokens": 537270101.0,
      "step": 15440
    },
    {
      "epoch": 2.867409470752089,
      "grad_norm": 1.4232173767737053,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8886334300041199,
      "num_tokens": 537312803.0,
      "step": 15441
    },
    {
      "epoch": 2.867595171773445,
      "grad_norm": 1.616191310874147,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8839324712753296,
      "num_tokens": 537345014.0,
      "step": 15442
    },
    {
      "epoch": 2.8677808727948,
      "grad_norm": 1.6088914944855321,
      "learning_rate": 1e-06,
      "loss": 0.3093,
      "mean_token_accuracy": 0.8946640491485596,
      "num_tokens": 537376866.0,
      "step": 15443
    },
    {
      "epoch": 2.867966573816156,
      "grad_norm": 1.602124936667246,
      "learning_rate": 1e-06,
      "loss": 0.2866,
      "mean_token_accuracy": 0.8978039026260376,
      "num_tokens": 537406739.0,
      "step": 15444
    },
    {
      "epoch": 2.8681522748375117,
      "grad_norm": 1.603650804316175,
      "learning_rate": 1e-06,
      "loss": 0.317,
      "mean_token_accuracy": 0.8890164494514465,
      "num_tokens": 537437174.0,
      "step": 15445
    },
    {
      "epoch": 2.868337975858867,
      "grad_norm": 1.7098354109288167,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8820068836212158,
      "num_tokens": 537469056.0,
      "step": 15446
    },
    {
      "epoch": 2.8685236768802227,
      "grad_norm": 1.4714985644285299,
      "learning_rate": 1e-06,
      "loss": 0.3348,
      "mean_token_accuracy": 0.8821180462837219,
      "num_tokens": 537509017.0,
      "step": 15447
    },
    {
      "epoch": 2.8687093779015784,
      "grad_norm": 1.628648022433255,
      "learning_rate": 1e-06,
      "loss": 0.2827,
      "mean_token_accuracy": 0.9025214910507202,
      "num_tokens": 537538268.0,
      "step": 15448
    },
    {
      "epoch": 2.868895078922934,
      "grad_norm": 1.4625722292105374,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8766964673995972,
      "num_tokens": 537576716.0,
      "step": 15449
    },
    {
      "epoch": 2.86908077994429,
      "grad_norm": 1.7217483567652108,
      "learning_rate": 1e-06,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8757431507110596,
      "num_tokens": 537607062.0,
      "step": 15450
    },
    {
      "epoch": 2.869266480965645,
      "grad_norm": 1.8545014639600064,
      "learning_rate": 1e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8670345544815063,
      "num_tokens": 537638535.0,
      "step": 15451
    },
    {
      "epoch": 2.869452181987001,
      "grad_norm": 1.37365274536282,
      "learning_rate": 1e-06,
      "loss": 0.3164,
      "mean_token_accuracy": 0.8881776928901672,
      "num_tokens": 537678037.0,
      "step": 15452
    },
    {
      "epoch": 2.8696378830083566,
      "grad_norm": 1.4934352515101623,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8945872783660889,
      "num_tokens": 537709298.0,
      "step": 15453
    },
    {
      "epoch": 2.869823584029712,
      "grad_norm": 1.5435133416241056,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8915365934371948,
      "num_tokens": 537740496.0,
      "step": 15454
    },
    {
      "epoch": 2.8700092850510677,
      "grad_norm": 1.569748363606898,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8889104723930359,
      "num_tokens": 537772528.0,
      "step": 15455
    },
    {
      "epoch": 2.8701949860724234,
      "grad_norm": 1.4119353482829173,
      "learning_rate": 1e-06,
      "loss": 0.2953,
      "mean_token_accuracy": 0.8967112898826599,
      "num_tokens": 537811630.0,
      "step": 15456
    },
    {
      "epoch": 2.870380687093779,
      "grad_norm": 1.417034172124189,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.8860634565353394,
      "num_tokens": 537849226.0,
      "step": 15457
    },
    {
      "epoch": 2.870566388115135,
      "grad_norm": 1.5916901053933021,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8805400133132935,
      "num_tokens": 537888955.0,
      "step": 15458
    },
    {
      "epoch": 2.87075208913649,
      "grad_norm": 1.3938041032275563,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.888870358467102,
      "num_tokens": 537934857.0,
      "step": 15459
    },
    {
      "epoch": 2.870937790157846,
      "grad_norm": 1.5996135191967518,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.879728376865387,
      "num_tokens": 537970593.0,
      "step": 15460
    },
    {
      "epoch": 2.8711234911792016,
      "grad_norm": 1.7024970383717537,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8779982924461365,
      "num_tokens": 538003601.0,
      "step": 15461
    },
    {
      "epoch": 2.871309192200557,
      "grad_norm": 1.7209598028238817,
      "learning_rate": 1e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8796606063842773,
      "num_tokens": 538034935.0,
      "step": 15462
    },
    {
      "epoch": 2.8714948932219126,
      "grad_norm": 1.4112058136545087,
      "learning_rate": 1e-06,
      "loss": 0.282,
      "mean_token_accuracy": 0.8962039947509766,
      "num_tokens": 538073988.0,
      "step": 15463
    },
    {
      "epoch": 2.8716805942432684,
      "grad_norm": 1.5535246645511487,
      "learning_rate": 1e-06,
      "loss": 0.3254,
      "mean_token_accuracy": 0.887052059173584,
      "num_tokens": 538107620.0,
      "step": 15464
    },
    {
      "epoch": 2.871866295264624,
      "grad_norm": 1.6052396140182335,
      "learning_rate": 1e-06,
      "loss": 0.3222,
      "mean_token_accuracy": 0.8828689455986023,
      "num_tokens": 538140525.0,
      "step": 15465
    },
    {
      "epoch": 2.87205199628598,
      "grad_norm": 1.4896397777308854,
      "learning_rate": 1e-06,
      "loss": 0.3429,
      "mean_token_accuracy": 0.8819453716278076,
      "num_tokens": 538181688.0,
      "step": 15466
    },
    {
      "epoch": 2.872237697307335,
      "grad_norm": 1.548729096471719,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8863990306854248,
      "num_tokens": 538215007.0,
      "step": 15467
    },
    {
      "epoch": 2.872423398328691,
      "grad_norm": 1.5724990366601577,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8919762969017029,
      "num_tokens": 538251543.0,
      "step": 15468
    },
    {
      "epoch": 2.872609099350046,
      "grad_norm": 1.566723321197348,
      "learning_rate": 1e-06,
      "loss": 0.2724,
      "mean_token_accuracy": 0.9023895263671875,
      "num_tokens": 538283925.0,
      "step": 15469
    },
    {
      "epoch": 2.872794800371402,
      "grad_norm": 1.7271519695541833,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8821858167648315,
      "num_tokens": 538313112.0,
      "step": 15470
    },
    {
      "epoch": 2.8729805013927576,
      "grad_norm": 1.534894214076051,
      "learning_rate": 1e-06,
      "loss": 0.3028,
      "mean_token_accuracy": 0.893959641456604,
      "num_tokens": 538346495.0,
      "step": 15471
    },
    {
      "epoch": 2.8731662024141134,
      "grad_norm": 1.540829944059299,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8886725902557373,
      "num_tokens": 538382533.0,
      "step": 15472
    },
    {
      "epoch": 2.873351903435469,
      "grad_norm": 1.7031358656402231,
      "learning_rate": 1e-06,
      "loss": 0.3398,
      "mean_token_accuracy": 0.8820726871490479,
      "num_tokens": 538413140.0,
      "step": 15473
    },
    {
      "epoch": 2.8735376044568244,
      "grad_norm": 1.5147184422149953,
      "learning_rate": 1e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8738008737564087,
      "num_tokens": 538455179.0,
      "step": 15474
    },
    {
      "epoch": 2.87372330547818,
      "grad_norm": 1.6545839294512874,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8887014389038086,
      "num_tokens": 538487831.0,
      "step": 15475
    },
    {
      "epoch": 2.873909006499536,
      "grad_norm": 1.527334304069602,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.8873305320739746,
      "num_tokens": 538523644.0,
      "step": 15476
    },
    {
      "epoch": 2.874094707520891,
      "grad_norm": 1.44436524272896,
      "learning_rate": 1e-06,
      "loss": 0.3204,
      "mean_token_accuracy": 0.886405348777771,
      "num_tokens": 538560898.0,
      "step": 15477
    },
    {
      "epoch": 2.874280408542247,
      "grad_norm": 1.7359260350937944,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8780202865600586,
      "num_tokens": 538591994.0,
      "step": 15478
    },
    {
      "epoch": 2.8744661095636026,
      "grad_norm": 1.6293910258546922,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8897168636322021,
      "num_tokens": 538622432.0,
      "step": 15479
    },
    {
      "epoch": 2.8746518105849583,
      "grad_norm": 1.4926732456277625,
      "learning_rate": 1e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.8814066052436829,
      "num_tokens": 538662046.0,
      "step": 15480
    },
    {
      "epoch": 2.874837511606314,
      "grad_norm": 1.6190252572509634,
      "learning_rate": 1e-06,
      "loss": 0.3367,
      "mean_token_accuracy": 0.8812499046325684,
      "num_tokens": 538695189.0,
      "step": 15481
    },
    {
      "epoch": 2.8750232126276694,
      "grad_norm": 1.6787355891900881,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8754673004150391,
      "num_tokens": 538725656.0,
      "step": 15482
    },
    {
      "epoch": 2.875208913649025,
      "grad_norm": 1.445223969911418,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.889601469039917,
      "num_tokens": 538767369.0,
      "step": 15483
    },
    {
      "epoch": 2.875394614670381,
      "grad_norm": 1.6511169265420085,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8809137940406799,
      "num_tokens": 538800852.0,
      "step": 15484
    },
    {
      "epoch": 2.875580315691736,
      "grad_norm": 1.4842143409567206,
      "learning_rate": 1e-06,
      "loss": 0.2779,
      "mean_token_accuracy": 0.9007829427719116,
      "num_tokens": 538835242.0,
      "step": 15485
    },
    {
      "epoch": 2.875766016713092,
      "grad_norm": 1.6459430911164314,
      "learning_rate": 1e-06,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8870484232902527,
      "num_tokens": 538867899.0,
      "step": 15486
    },
    {
      "epoch": 2.8759517177344476,
      "grad_norm": 1.7154726816436494,
      "learning_rate": 1e-06,
      "loss": 0.3558,
      "mean_token_accuracy": 0.877197802066803,
      "num_tokens": 538900206.0,
      "step": 15487
    },
    {
      "epoch": 2.8761374187558033,
      "grad_norm": 1.4640620123057486,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8905539512634277,
      "num_tokens": 538938484.0,
      "step": 15488
    },
    {
      "epoch": 2.876323119777159,
      "grad_norm": 1.5022813565097963,
      "learning_rate": 1e-06,
      "loss": 0.2935,
      "mean_token_accuracy": 0.8954390287399292,
      "num_tokens": 538975654.0,
      "step": 15489
    },
    {
      "epoch": 2.8765088207985143,
      "grad_norm": 1.6512944566097494,
      "learning_rate": 1e-06,
      "loss": 0.3598,
      "mean_token_accuracy": 0.8788356781005859,
      "num_tokens": 539013304.0,
      "step": 15490
    },
    {
      "epoch": 2.87669452181987,
      "grad_norm": 1.4353688365051047,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.8926306962966919,
      "num_tokens": 539048356.0,
      "step": 15491
    },
    {
      "epoch": 2.8768802228412254,
      "grad_norm": 1.4244994550956178,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8950257301330566,
      "num_tokens": 539085819.0,
      "step": 15492
    },
    {
      "epoch": 2.877065923862581,
      "grad_norm": 1.4667952852861905,
      "learning_rate": 1e-06,
      "loss": 0.3459,
      "mean_token_accuracy": 0.8801611661911011,
      "num_tokens": 539126015.0,
      "step": 15493
    },
    {
      "epoch": 2.877251624883937,
      "grad_norm": 1.5315579777760926,
      "learning_rate": 1e-06,
      "loss": 0.3168,
      "mean_token_accuracy": 0.8849097490310669,
      "num_tokens": 539163136.0,
      "step": 15494
    },
    {
      "epoch": 2.8774373259052926,
      "grad_norm": 1.5625490946193479,
      "learning_rate": 1e-06,
      "loss": 0.3024,
      "mean_token_accuracy": 0.8925593495368958,
      "num_tokens": 539197143.0,
      "step": 15495
    },
    {
      "epoch": 2.8776230269266483,
      "grad_norm": 1.468635927103877,
      "learning_rate": 1e-06,
      "loss": 0.3112,
      "mean_token_accuracy": 0.8923935890197754,
      "num_tokens": 539238635.0,
      "step": 15496
    },
    {
      "epoch": 2.8778087279480036,
      "grad_norm": 1.490215004819929,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8953837156295776,
      "num_tokens": 539274815.0,
      "step": 15497
    },
    {
      "epoch": 2.8779944289693593,
      "grad_norm": 1.683169173149813,
      "learning_rate": 1e-06,
      "loss": 0.3077,
      "mean_token_accuracy": 0.8934131860733032,
      "num_tokens": 539303683.0,
      "step": 15498
    },
    {
      "epoch": 2.878180129990715,
      "grad_norm": 1.5548490141917808,
      "learning_rate": 1e-06,
      "loss": 0.3618,
      "mean_token_accuracy": 0.8727731704711914,
      "num_tokens": 539340379.0,
      "step": 15499
    },
    {
      "epoch": 2.8783658310120703,
      "grad_norm": 1.6594098974075921,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8884352445602417,
      "num_tokens": 539372077.0,
      "step": 15500
    },
    {
      "epoch": 2.878551532033426,
      "grad_norm": 1.4836065969977426,
      "learning_rate": 1e-06,
      "loss": 0.3547,
      "mean_token_accuracy": 0.8759337067604065,
      "num_tokens": 539413557.0,
      "step": 15501
    },
    {
      "epoch": 2.878737233054782,
      "grad_norm": 1.6073254191363122,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.891494631767273,
      "num_tokens": 539444848.0,
      "step": 15502
    },
    {
      "epoch": 2.8789229340761375,
      "grad_norm": 1.625424290134304,
      "learning_rate": 1e-06,
      "loss": 0.3145,
      "mean_token_accuracy": 0.8902089595794678,
      "num_tokens": 539474566.0,
      "step": 15503
    },
    {
      "epoch": 2.8791086350974933,
      "grad_norm": 1.4693409626111729,
      "learning_rate": 1e-06,
      "loss": 0.2857,
      "mean_token_accuracy": 0.8990386724472046,
      "num_tokens": 539510738.0,
      "step": 15504
    },
    {
      "epoch": 2.8792943361188486,
      "grad_norm": 1.6071738885400297,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8795650005340576,
      "num_tokens": 539547406.0,
      "step": 15505
    },
    {
      "epoch": 2.8794800371402043,
      "grad_norm": 1.590774948269028,
      "learning_rate": 1e-06,
      "loss": 0.2873,
      "mean_token_accuracy": 0.8967692852020264,
      "num_tokens": 539581667.0,
      "step": 15506
    },
    {
      "epoch": 2.87966573816156,
      "grad_norm": 1.5403146776260872,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8829311728477478,
      "num_tokens": 539622494.0,
      "step": 15507
    },
    {
      "epoch": 2.8798514391829153,
      "grad_norm": 1.4778859059896052,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8906773328781128,
      "num_tokens": 539660762.0,
      "step": 15508
    },
    {
      "epoch": 2.880037140204271,
      "grad_norm": 1.6855711520735581,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.8855330944061279,
      "num_tokens": 539689558.0,
      "step": 15509
    },
    {
      "epoch": 2.8802228412256268,
      "grad_norm": 1.5610163908800576,
      "learning_rate": 1e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8693758249282837,
      "num_tokens": 539728307.0,
      "step": 15510
    },
    {
      "epoch": 2.8804085422469825,
      "grad_norm": 1.7087766588909443,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8801571130752563,
      "num_tokens": 539761181.0,
      "step": 15511
    },
    {
      "epoch": 2.8805942432683382,
      "grad_norm": 1.5835882370688346,
      "learning_rate": 1e-06,
      "loss": 0.2999,
      "mean_token_accuracy": 0.8915581703186035,
      "num_tokens": 539793631.0,
      "step": 15512
    },
    {
      "epoch": 2.8807799442896935,
      "grad_norm": 1.4871870382461467,
      "learning_rate": 1e-06,
      "loss": 0.2934,
      "mean_token_accuracy": 0.8990460634231567,
      "num_tokens": 539832470.0,
      "step": 15513
    },
    {
      "epoch": 2.8809656453110493,
      "grad_norm": 1.6656929778770593,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8812421560287476,
      "num_tokens": 539864281.0,
      "step": 15514
    },
    {
      "epoch": 2.881151346332405,
      "grad_norm": 1.5051452942484702,
      "learning_rate": 1e-06,
      "loss": 0.3023,
      "mean_token_accuracy": 0.8931583166122437,
      "num_tokens": 539897940.0,
      "step": 15515
    },
    {
      "epoch": 2.8813370473537603,
      "grad_norm": 1.572857967712675,
      "learning_rate": 1e-06,
      "loss": 0.3217,
      "mean_token_accuracy": 0.8854359984397888,
      "num_tokens": 539929710.0,
      "step": 15516
    },
    {
      "epoch": 2.881522748375116,
      "grad_norm": 1.4261040696185463,
      "learning_rate": 1e-06,
      "loss": 0.3052,
      "mean_token_accuracy": 0.8954163789749146,
      "num_tokens": 539969447.0,
      "step": 15517
    },
    {
      "epoch": 2.8817084493964717,
      "grad_norm": 1.657638354856084,
      "learning_rate": 1e-06,
      "loss": 0.3385,
      "mean_token_accuracy": 0.8804671168327332,
      "num_tokens": 539997675.0,
      "step": 15518
    },
    {
      "epoch": 2.8818941504178275,
      "grad_norm": 1.5189560663225912,
      "learning_rate": 1e-06,
      "loss": 0.2604,
      "mean_token_accuracy": 0.9093860983848572,
      "num_tokens": 540035510.0,
      "step": 15519
    },
    {
      "epoch": 2.8820798514391828,
      "grad_norm": 1.5635855401924186,
      "learning_rate": 1e-06,
      "loss": 0.3299,
      "mean_token_accuracy": 0.8864745497703552,
      "num_tokens": 540070984.0,
      "step": 15520
    },
    {
      "epoch": 2.8822655524605385,
      "grad_norm": 1.465262822620672,
      "learning_rate": 1e-06,
      "loss": 0.2762,
      "mean_token_accuracy": 0.9027965068817139,
      "num_tokens": 540104489.0,
      "step": 15521
    },
    {
      "epoch": 2.8824512534818942,
      "grad_norm": 1.7174454554418608,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.8950760364532471,
      "num_tokens": 540132812.0,
      "step": 15522
    },
    {
      "epoch": 2.8826369545032495,
      "grad_norm": 1.437655185668005,
      "learning_rate": 1e-06,
      "loss": 0.2656,
      "mean_token_accuracy": 0.9034153819084167,
      "num_tokens": 540169333.0,
      "step": 15523
    },
    {
      "epoch": 2.8828226555246053,
      "grad_norm": 1.4236763488303614,
      "learning_rate": 1e-06,
      "loss": 0.326,
      "mean_token_accuracy": 0.8851560354232788,
      "num_tokens": 540212676.0,
      "step": 15524
    },
    {
      "epoch": 2.883008356545961,
      "grad_norm": 1.5457402370518365,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8836752772331238,
      "num_tokens": 540248128.0,
      "step": 15525
    },
    {
      "epoch": 2.8831940575673167,
      "grad_norm": 1.446584195446058,
      "learning_rate": 1e-06,
      "loss": 0.3042,
      "mean_token_accuracy": 0.8967059850692749,
      "num_tokens": 540286493.0,
      "step": 15526
    },
    {
      "epoch": 2.8833797585886725,
      "grad_norm": 1.5251504892878274,
      "learning_rate": 1e-06,
      "loss": 0.3184,
      "mean_token_accuracy": 0.8881106376647949,
      "num_tokens": 540321390.0,
      "step": 15527
    },
    {
      "epoch": 2.8835654596100277,
      "grad_norm": 1.6370565498273688,
      "learning_rate": 1e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8739550113677979,
      "num_tokens": 540353351.0,
      "step": 15528
    },
    {
      "epoch": 2.8837511606313835,
      "grad_norm": 1.567994989668805,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8917973637580872,
      "num_tokens": 540386572.0,
      "step": 15529
    },
    {
      "epoch": 2.883936861652739,
      "grad_norm": 1.519733023453949,
      "learning_rate": 1e-06,
      "loss": 0.3123,
      "mean_token_accuracy": 0.8890681266784668,
      "num_tokens": 540423427.0,
      "step": 15530
    },
    {
      "epoch": 2.8841225626740945,
      "grad_norm": 1.644112058721984,
      "learning_rate": 1e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8658038377761841,
      "num_tokens": 540461749.0,
      "step": 15531
    },
    {
      "epoch": 2.8843082636954502,
      "grad_norm": 1.6068828095788525,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.893333375453949,
      "num_tokens": 540492746.0,
      "step": 15532
    },
    {
      "epoch": 2.884493964716806,
      "grad_norm": 1.522762154395994,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8796274662017822,
      "num_tokens": 540532604.0,
      "step": 15533
    },
    {
      "epoch": 2.8846796657381617,
      "grad_norm": 1.7396224076257945,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8868454694747925,
      "num_tokens": 540559502.0,
      "step": 15534
    },
    {
      "epoch": 2.8848653667595174,
      "grad_norm": 1.8338657601612245,
      "learning_rate": 1e-06,
      "loss": 0.3376,
      "mean_token_accuracy": 0.8852218985557556,
      "num_tokens": 540588605.0,
      "step": 15535
    },
    {
      "epoch": 2.8850510677808727,
      "grad_norm": 1.6132221521352927,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.8843984007835388,
      "num_tokens": 540622297.0,
      "step": 15536
    },
    {
      "epoch": 2.8852367688022285,
      "grad_norm": 1.6422114545501942,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.879773736000061,
      "num_tokens": 540654009.0,
      "step": 15537
    },
    {
      "epoch": 2.885422469823584,
      "grad_norm": 1.4388988838989698,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.890865683555603,
      "num_tokens": 540694910.0,
      "step": 15538
    },
    {
      "epoch": 2.8856081708449395,
      "grad_norm": 1.387368808694831,
      "learning_rate": 1e-06,
      "loss": 0.281,
      "mean_token_accuracy": 0.8982075452804565,
      "num_tokens": 540733037.0,
      "step": 15539
    },
    {
      "epoch": 2.885793871866295,
      "grad_norm": 1.5436096791028153,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8861403465270996,
      "num_tokens": 540768197.0,
      "step": 15540
    },
    {
      "epoch": 2.885979572887651,
      "grad_norm": 1.5348041512384458,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8846816420555115,
      "num_tokens": 540804678.0,
      "step": 15541
    },
    {
      "epoch": 2.8861652739090067,
      "grad_norm": 1.46292402731084,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.886728823184967,
      "num_tokens": 540845405.0,
      "step": 15542
    },
    {
      "epoch": 2.886350974930362,
      "grad_norm": 1.656982619083459,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8770656585693359,
      "num_tokens": 540877907.0,
      "step": 15543
    },
    {
      "epoch": 2.8865366759517177,
      "grad_norm": 1.4650185995399754,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8852250576019287,
      "num_tokens": 540920266.0,
      "step": 15544
    },
    {
      "epoch": 2.8867223769730734,
      "grad_norm": 1.5002539283899456,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.880843997001648,
      "num_tokens": 540957345.0,
      "step": 15545
    },
    {
      "epoch": 2.8869080779944287,
      "grad_norm": 1.7467572688748154,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8866350650787354,
      "num_tokens": 540985855.0,
      "step": 15546
    },
    {
      "epoch": 2.8870937790157845,
      "grad_norm": 1.7290167368457852,
      "learning_rate": 1e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8781689405441284,
      "num_tokens": 541017070.0,
      "step": 15547
    },
    {
      "epoch": 2.88727948003714,
      "grad_norm": 1.5002772125009496,
      "learning_rate": 1e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.8875994086265564,
      "num_tokens": 541052498.0,
      "step": 15548
    },
    {
      "epoch": 2.887465181058496,
      "grad_norm": 1.7009048695585576,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8790267705917358,
      "num_tokens": 541086262.0,
      "step": 15549
    },
    {
      "epoch": 2.8876508820798517,
      "grad_norm": 1.6343515368475958,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8828946948051453,
      "num_tokens": 541117445.0,
      "step": 15550
    },
    {
      "epoch": 2.887836583101207,
      "grad_norm": 1.7112021185520498,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8821677565574646,
      "num_tokens": 541148836.0,
      "step": 15551
    },
    {
      "epoch": 2.8880222841225627,
      "grad_norm": 1.53703442797093,
      "learning_rate": 1e-06,
      "loss": 0.2927,
      "mean_token_accuracy": 0.8949057459831238,
      "num_tokens": 541185636.0,
      "step": 15552
    },
    {
      "epoch": 2.8882079851439184,
      "grad_norm": 1.5433064000470047,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8912715911865234,
      "num_tokens": 541220835.0,
      "step": 15553
    },
    {
      "epoch": 2.8883936861652737,
      "grad_norm": 1.5496901061537336,
      "learning_rate": 1e-06,
      "loss": 0.2895,
      "mean_token_accuracy": 0.8953865766525269,
      "num_tokens": 541250680.0,
      "step": 15554
    },
    {
      "epoch": 2.8885793871866294,
      "grad_norm": 1.6503363812352887,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8885884284973145,
      "num_tokens": 541285354.0,
      "step": 15555
    },
    {
      "epoch": 2.888765088207985,
      "grad_norm": 1.574940755629621,
      "learning_rate": 1e-06,
      "loss": 0.2795,
      "mean_token_accuracy": 0.8987922668457031,
      "num_tokens": 541316092.0,
      "step": 15556
    },
    {
      "epoch": 2.888950789229341,
      "grad_norm": 1.5734550321494685,
      "learning_rate": 1e-06,
      "loss": 0.2951,
      "mean_token_accuracy": 0.8955442309379578,
      "num_tokens": 541350154.0,
      "step": 15557
    },
    {
      "epoch": 2.8891364902506966,
      "grad_norm": 1.5452265935508276,
      "learning_rate": 1e-06,
      "loss": 0.3118,
      "mean_token_accuracy": 0.888222873210907,
      "num_tokens": 541384243.0,
      "step": 15558
    },
    {
      "epoch": 2.889322191272052,
      "grad_norm": 1.4707110960222987,
      "learning_rate": 1e-06,
      "loss": 0.357,
      "mean_token_accuracy": 0.8741037845611572,
      "num_tokens": 541427095.0,
      "step": 15559
    },
    {
      "epoch": 2.8895078922934077,
      "grad_norm": 1.7316931499189179,
      "learning_rate": 1e-06,
      "loss": 0.3317,
      "mean_token_accuracy": 0.8856205344200134,
      "num_tokens": 541458329.0,
      "step": 15560
    },
    {
      "epoch": 2.8896935933147634,
      "grad_norm": 1.5886026060161391,
      "learning_rate": 1e-06,
      "loss": 0.3163,
      "mean_token_accuracy": 0.890182614326477,
      "num_tokens": 541492522.0,
      "step": 15561
    },
    {
      "epoch": 2.8898792943361187,
      "grad_norm": 1.6316893111973128,
      "learning_rate": 1e-06,
      "loss": 0.3676,
      "mean_token_accuracy": 0.8715894222259521,
      "num_tokens": 541526057.0,
      "step": 15562
    },
    {
      "epoch": 2.8900649953574744,
      "grad_norm": 1.5365215265604497,
      "learning_rate": 1e-06,
      "loss": 0.3016,
      "mean_token_accuracy": 0.8932355642318726,
      "num_tokens": 541558504.0,
      "step": 15563
    },
    {
      "epoch": 2.89025069637883,
      "grad_norm": 1.7105395975475983,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8837233781814575,
      "num_tokens": 541591705.0,
      "step": 15564
    },
    {
      "epoch": 2.890436397400186,
      "grad_norm": 1.573411910540093,
      "learning_rate": 1e-06,
      "loss": 0.3318,
      "mean_token_accuracy": 0.8858668804168701,
      "num_tokens": 541626497.0,
      "step": 15565
    },
    {
      "epoch": 2.890622098421541,
      "grad_norm": 1.508090748233167,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8809927701950073,
      "num_tokens": 541667675.0,
      "step": 15566
    },
    {
      "epoch": 2.890807799442897,
      "grad_norm": 1.533140234205773,
      "learning_rate": 1e-06,
      "loss": 0.3172,
      "mean_token_accuracy": 0.8888176679611206,
      "num_tokens": 541700056.0,
      "step": 15567
    },
    {
      "epoch": 2.8909935004642526,
      "grad_norm": 1.673514037057682,
      "learning_rate": 1e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.8754682540893555,
      "num_tokens": 541732477.0,
      "step": 15568
    },
    {
      "epoch": 2.891179201485608,
      "grad_norm": 1.703623282905566,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.880885124206543,
      "num_tokens": 541759791.0,
      "step": 15569
    },
    {
      "epoch": 2.8913649025069637,
      "grad_norm": 1.5156501137923115,
      "learning_rate": 1e-06,
      "loss": 0.2919,
      "mean_token_accuracy": 0.8956765532493591,
      "num_tokens": 541794581.0,
      "step": 15570
    },
    {
      "epoch": 2.8915506035283194,
      "grad_norm": 1.4617895883640366,
      "learning_rate": 1e-06,
      "loss": 0.3152,
      "mean_token_accuracy": 0.8927335739135742,
      "num_tokens": 541830117.0,
      "step": 15571
    },
    {
      "epoch": 2.891736304549675,
      "grad_norm": 1.5664686070021596,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8807882070541382,
      "num_tokens": 541864930.0,
      "step": 15572
    },
    {
      "epoch": 2.891922005571031,
      "grad_norm": 1.4391144042409265,
      "learning_rate": 1e-06,
      "loss": 0.2883,
      "mean_token_accuracy": 0.8970984220504761,
      "num_tokens": 541907961.0,
      "step": 15573
    },
    {
      "epoch": 2.892107706592386,
      "grad_norm": 1.4968314553046604,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8919469714164734,
      "num_tokens": 541944720.0,
      "step": 15574
    },
    {
      "epoch": 2.892293407613742,
      "grad_norm": 1.5599501770615993,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8809489011764526,
      "num_tokens": 541979010.0,
      "step": 15575
    },
    {
      "epoch": 2.8924791086350976,
      "grad_norm": 1.7440170354283624,
      "learning_rate": 1e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8849773406982422,
      "num_tokens": 542009400.0,
      "step": 15576
    },
    {
      "epoch": 2.892664809656453,
      "grad_norm": 1.77137964963325,
      "learning_rate": 1e-06,
      "loss": 0.3153,
      "mean_token_accuracy": 0.8877030611038208,
      "num_tokens": 542042176.0,
      "step": 15577
    },
    {
      "epoch": 2.8928505106778086,
      "grad_norm": 1.5473793253129444,
      "learning_rate": 1e-06,
      "loss": 0.3161,
      "mean_token_accuracy": 0.8887501358985901,
      "num_tokens": 542074844.0,
      "step": 15578
    },
    {
      "epoch": 2.8930362116991644,
      "grad_norm": 1.7084966131868584,
      "learning_rate": 1e-06,
      "loss": 0.28,
      "mean_token_accuracy": 0.8986914157867432,
      "num_tokens": 542106816.0,
      "step": 15579
    },
    {
      "epoch": 2.89322191272052,
      "grad_norm": 1.5728753405825115,
      "learning_rate": 1e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8744235038757324,
      "num_tokens": 542142747.0,
      "step": 15580
    },
    {
      "epoch": 2.893407613741876,
      "grad_norm": 1.5640277427150446,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8811551332473755,
      "num_tokens": 542178544.0,
      "step": 15581
    },
    {
      "epoch": 2.893593314763231,
      "grad_norm": 1.5247185460683568,
      "learning_rate": 1e-06,
      "loss": 0.2946,
      "mean_token_accuracy": 0.8965436220169067,
      "num_tokens": 542217895.0,
      "step": 15582
    },
    {
      "epoch": 2.893779015784587,
      "grad_norm": 1.6449832256485533,
      "learning_rate": 1e-06,
      "loss": 0.3315,
      "mean_token_accuracy": 0.8845441341400146,
      "num_tokens": 542255906.0,
      "step": 15583
    },
    {
      "epoch": 2.8939647168059426,
      "grad_norm": 1.6433999998835807,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8890783786773682,
      "num_tokens": 542291929.0,
      "step": 15584
    },
    {
      "epoch": 2.894150417827298,
      "grad_norm": 1.5411960430295504,
      "learning_rate": 1e-06,
      "loss": 0.3314,
      "mean_token_accuracy": 0.8822275400161743,
      "num_tokens": 542326960.0,
      "step": 15585
    },
    {
      "epoch": 2.8943361188486536,
      "grad_norm": 1.583824655661653,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8936444520950317,
      "num_tokens": 542360647.0,
      "step": 15586
    },
    {
      "epoch": 2.8945218198700093,
      "grad_norm": 1.5870924808318656,
      "learning_rate": 1e-06,
      "loss": 0.3066,
      "mean_token_accuracy": 0.891861081123352,
      "num_tokens": 542394374.0,
      "step": 15587
    },
    {
      "epoch": 2.894707520891365,
      "grad_norm": 1.4928252832545605,
      "learning_rate": 1e-06,
      "loss": 0.3491,
      "mean_token_accuracy": 0.8797284364700317,
      "num_tokens": 542437821.0,
      "step": 15588
    },
    {
      "epoch": 2.8948932219127204,
      "grad_norm": 1.8158095224377493,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8807787895202637,
      "num_tokens": 542466667.0,
      "step": 15589
    },
    {
      "epoch": 2.895078922934076,
      "grad_norm": 1.6199039600844454,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8921982049942017,
      "num_tokens": 542500297.0,
      "step": 15590
    },
    {
      "epoch": 2.895264623955432,
      "grad_norm": 1.7126791740554337,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8842660188674927,
      "num_tokens": 542527999.0,
      "step": 15591
    },
    {
      "epoch": 2.895450324976787,
      "grad_norm": 1.6094717156568141,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.892548680305481,
      "num_tokens": 542561153.0,
      "step": 15592
    },
    {
      "epoch": 2.895636025998143,
      "grad_norm": 1.6271030429832523,
      "learning_rate": 1e-06,
      "loss": 0.3121,
      "mean_token_accuracy": 0.8895286917686462,
      "num_tokens": 542592575.0,
      "step": 15593
    },
    {
      "epoch": 2.8958217270194986,
      "grad_norm": 1.523433960805917,
      "learning_rate": 1e-06,
      "loss": 0.294,
      "mean_token_accuracy": 0.8935480117797852,
      "num_tokens": 542626429.0,
      "step": 15594
    },
    {
      "epoch": 2.8960074280408543,
      "grad_norm": 1.7188264766808237,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8813649415969849,
      "num_tokens": 542659489.0,
      "step": 15595
    },
    {
      "epoch": 2.89619312906221,
      "grad_norm": 1.4136097350010386,
      "learning_rate": 1e-06,
      "loss": 0.2643,
      "mean_token_accuracy": 0.9058455228805542,
      "num_tokens": 542692850.0,
      "step": 15596
    },
    {
      "epoch": 2.8963788300835653,
      "grad_norm": 1.3998903179449211,
      "learning_rate": 1e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.89182448387146,
      "num_tokens": 542729968.0,
      "step": 15597
    },
    {
      "epoch": 2.896564531104921,
      "grad_norm": 1.3967548637800646,
      "learning_rate": 1e-06,
      "loss": 0.2748,
      "mean_token_accuracy": 0.9007778167724609,
      "num_tokens": 542769905.0,
      "step": 15598
    },
    {
      "epoch": 2.896750232126277,
      "grad_norm": 1.410616331525018,
      "learning_rate": 1e-06,
      "loss": 0.307,
      "mean_token_accuracy": 0.8927571177482605,
      "num_tokens": 542810755.0,
      "step": 15599
    },
    {
      "epoch": 2.896935933147632,
      "grad_norm": 1.6915668462058433,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.8829221725463867,
      "num_tokens": 542840867.0,
      "step": 15600
    },
    {
      "epoch": 2.897121634168988,
      "grad_norm": 1.4103721473037447,
      "learning_rate": 1e-06,
      "loss": 0.2757,
      "mean_token_accuracy": 0.9020012617111206,
      "num_tokens": 542877262.0,
      "step": 15601
    },
    {
      "epoch": 2.8973073351903436,
      "grad_norm": 1.5086596724301387,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8808778524398804,
      "num_tokens": 542916355.0,
      "step": 15602
    },
    {
      "epoch": 2.8974930362116993,
      "grad_norm": 1.6024864957350438,
      "learning_rate": 1e-06,
      "loss": 0.3283,
      "mean_token_accuracy": 0.8847503662109375,
      "num_tokens": 542949372.0,
      "step": 15603
    },
    {
      "epoch": 2.897678737233055,
      "grad_norm": 1.5413067011750523,
      "learning_rate": 1e-06,
      "loss": 0.3296,
      "mean_token_accuracy": 0.885932445526123,
      "num_tokens": 542985400.0,
      "step": 15604
    },
    {
      "epoch": 2.8978644382544103,
      "grad_norm": 1.3949790327877025,
      "learning_rate": 1e-06,
      "loss": 0.3199,
      "mean_token_accuracy": 0.8851673603057861,
      "num_tokens": 543030264.0,
      "step": 15605
    },
    {
      "epoch": 2.898050139275766,
      "grad_norm": 1.5382089795440228,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8930149674415588,
      "num_tokens": 543063320.0,
      "step": 15606
    },
    {
      "epoch": 2.898235840297122,
      "grad_norm": 1.5225897101580277,
      "learning_rate": 1e-06,
      "loss": 0.3542,
      "mean_token_accuracy": 0.8771927356719971,
      "num_tokens": 543103997.0,
      "step": 15607
    },
    {
      "epoch": 2.898421541318477,
      "grad_norm": 1.4855759260151131,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8883072137832642,
      "num_tokens": 543144052.0,
      "step": 15608
    },
    {
      "epoch": 2.898607242339833,
      "grad_norm": 1.5737890694257484,
      "learning_rate": 1e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8806681632995605,
      "num_tokens": 543186213.0,
      "step": 15609
    },
    {
      "epoch": 2.8987929433611885,
      "grad_norm": 1.5638334503113154,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8762582540512085,
      "num_tokens": 543224276.0,
      "step": 15610
    },
    {
      "epoch": 2.8989786443825443,
      "grad_norm": 1.7176553746012535,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8800129890441895,
      "num_tokens": 543254053.0,
      "step": 15611
    },
    {
      "epoch": 2.8991643454038996,
      "grad_norm": 1.4833732503356758,
      "learning_rate": 1e-06,
      "loss": 0.3246,
      "mean_token_accuracy": 0.8855304718017578,
      "num_tokens": 543289884.0,
      "step": 15612
    },
    {
      "epoch": 2.8993500464252553,
      "grad_norm": 1.4183675484664284,
      "learning_rate": 1e-06,
      "loss": 0.2914,
      "mean_token_accuracy": 0.8950351476669312,
      "num_tokens": 543328978.0,
      "step": 15613
    },
    {
      "epoch": 2.899535747446611,
      "grad_norm": 1.724609422056062,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8753935694694519,
      "num_tokens": 543358412.0,
      "step": 15614
    },
    {
      "epoch": 2.8997214484679663,
      "grad_norm": 1.4969568447547112,
      "learning_rate": 1e-06,
      "loss": 0.2769,
      "mean_token_accuracy": 0.90123450756073,
      "num_tokens": 543389427.0,
      "step": 15615
    },
    {
      "epoch": 2.899907149489322,
      "grad_norm": 1.5579501369325697,
      "learning_rate": 1e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8817645907402039,
      "num_tokens": 543425210.0,
      "step": 15616
    },
    {
      "epoch": 2.9000928505106778,
      "grad_norm": 1.5478040730032336,
      "learning_rate": 1e-06,
      "loss": 0.2751,
      "mean_token_accuracy": 0.9018611907958984,
      "num_tokens": 543454947.0,
      "step": 15617
    },
    {
      "epoch": 2.9002785515320335,
      "grad_norm": 1.4857119703799837,
      "learning_rate": 1e-06,
      "loss": 0.3248,
      "mean_token_accuracy": 0.8834758400917053,
      "num_tokens": 543494034.0,
      "step": 15618
    },
    {
      "epoch": 2.9004642525533892,
      "grad_norm": 1.5021271540629701,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8918232321739197,
      "num_tokens": 543531104.0,
      "step": 15619
    },
    {
      "epoch": 2.9006499535747445,
      "grad_norm": 1.4936630123401071,
      "learning_rate": 1e-06,
      "loss": 0.2885,
      "mean_token_accuracy": 0.8994418382644653,
      "num_tokens": 543567516.0,
      "step": 15620
    },
    {
      "epoch": 2.9008356545961003,
      "grad_norm": 1.6013154540451369,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8851406574249268,
      "num_tokens": 543602955.0,
      "step": 15621
    },
    {
      "epoch": 2.901021355617456,
      "grad_norm": 1.4961713125700429,
      "learning_rate": 1e-06,
      "loss": 0.352,
      "mean_token_accuracy": 0.875591516494751,
      "num_tokens": 543644203.0,
      "step": 15622
    },
    {
      "epoch": 2.9012070566388113,
      "grad_norm": 1.7461787245507023,
      "learning_rate": 1e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8694621920585632,
      "num_tokens": 543675602.0,
      "step": 15623
    },
    {
      "epoch": 2.901392757660167,
      "grad_norm": 1.5299618995307067,
      "learning_rate": 1e-06,
      "loss": 0.2871,
      "mean_token_accuracy": 0.8978016972541809,
      "num_tokens": 543708686.0,
      "step": 15624
    },
    {
      "epoch": 2.9015784586815228,
      "grad_norm": 1.6701258224960442,
      "learning_rate": 1e-06,
      "loss": 0.3516,
      "mean_token_accuracy": 0.8767127394676208,
      "num_tokens": 543742030.0,
      "step": 15625
    },
    {
      "epoch": 2.9017641597028785,
      "grad_norm": 1.3888747662880494,
      "learning_rate": 1e-06,
      "loss": 0.2646,
      "mean_token_accuracy": 0.9045063257217407,
      "num_tokens": 543779986.0,
      "step": 15626
    },
    {
      "epoch": 2.901949860724234,
      "grad_norm": 1.6742142456949507,
      "learning_rate": 1e-06,
      "loss": 0.342,
      "mean_token_accuracy": 0.8839172124862671,
      "num_tokens": 543810424.0,
      "step": 15627
    },
    {
      "epoch": 2.9021355617455895,
      "grad_norm": 1.583023162872451,
      "learning_rate": 1e-06,
      "loss": 0.3162,
      "mean_token_accuracy": 0.8879914283752441,
      "num_tokens": 543840999.0,
      "step": 15628
    },
    {
      "epoch": 2.9023212627669452,
      "grad_norm": 1.6474177639563945,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.886309027671814,
      "num_tokens": 543873654.0,
      "step": 15629
    },
    {
      "epoch": 2.902506963788301,
      "grad_norm": 1.454513331539301,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8957014083862305,
      "num_tokens": 543914486.0,
      "step": 15630
    },
    {
      "epoch": 2.9026926648096563,
      "grad_norm": 1.5512523233889692,
      "learning_rate": 1e-06,
      "loss": 0.2936,
      "mean_token_accuracy": 0.8935322761535645,
      "num_tokens": 543950081.0,
      "step": 15631
    },
    {
      "epoch": 2.902878365831012,
      "grad_norm": 1.5759645904698703,
      "learning_rate": 1e-06,
      "loss": 0.2963,
      "mean_token_accuracy": 0.8949241638183594,
      "num_tokens": 543981767.0,
      "step": 15632
    },
    {
      "epoch": 2.9030640668523677,
      "grad_norm": 1.4474385644519312,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8940931558609009,
      "num_tokens": 544019447.0,
      "step": 15633
    },
    {
      "epoch": 2.9032497678737235,
      "grad_norm": 1.7282550286348386,
      "learning_rate": 1e-06,
      "loss": 0.2954,
      "mean_token_accuracy": 0.8960875272750854,
      "num_tokens": 544046737.0,
      "step": 15634
    },
    {
      "epoch": 2.903435468895079,
      "grad_norm": 1.6622033202675792,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.892011821269989,
      "num_tokens": 544075869.0,
      "step": 15635
    },
    {
      "epoch": 2.9036211699164345,
      "grad_norm": 1.459456754730986,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8901338577270508,
      "num_tokens": 544111577.0,
      "step": 15636
    },
    {
      "epoch": 2.90380687093779,
      "grad_norm": 1.5338524411509333,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8860938549041748,
      "num_tokens": 544146096.0,
      "step": 15637
    },
    {
      "epoch": 2.9039925719591455,
      "grad_norm": 1.575050666940227,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8887125253677368,
      "num_tokens": 544181493.0,
      "step": 15638
    },
    {
      "epoch": 2.9041782729805012,
      "grad_norm": 1.56089144036512,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8878334164619446,
      "num_tokens": 544214914.0,
      "step": 15639
    },
    {
      "epoch": 2.904363974001857,
      "grad_norm": 1.4655545639990932,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8849425911903381,
      "num_tokens": 544255361.0,
      "step": 15640
    },
    {
      "epoch": 2.9045496750232127,
      "grad_norm": 1.4822402173974347,
      "learning_rate": 1e-06,
      "loss": 0.3546,
      "mean_token_accuracy": 0.8749251961708069,
      "num_tokens": 544300332.0,
      "step": 15641
    },
    {
      "epoch": 2.9047353760445684,
      "grad_norm": 1.426857616420953,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8919166922569275,
      "num_tokens": 544338002.0,
      "step": 15642
    },
    {
      "epoch": 2.9049210770659237,
      "grad_norm": 1.4967105756180654,
      "learning_rate": 1e-06,
      "loss": 0.3036,
      "mean_token_accuracy": 0.8918023109436035,
      "num_tokens": 544373387.0,
      "step": 15643
    },
    {
      "epoch": 2.9051067780872795,
      "grad_norm": 1.5429527604826716,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8853581547737122,
      "num_tokens": 544412883.0,
      "step": 15644
    },
    {
      "epoch": 2.905292479108635,
      "grad_norm": 1.8326901606655086,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8822349309921265,
      "num_tokens": 544440772.0,
      "step": 15645
    },
    {
      "epoch": 2.9054781801299905,
      "grad_norm": 1.7249708801269643,
      "learning_rate": 1e-06,
      "loss": 0.3554,
      "mean_token_accuracy": 0.8743081092834473,
      "num_tokens": 544472895.0,
      "step": 15646
    },
    {
      "epoch": 2.905663881151346,
      "grad_norm": 1.4956005713347065,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8823494911193848,
      "num_tokens": 544514631.0,
      "step": 15647
    },
    {
      "epoch": 2.905849582172702,
      "grad_norm": 1.6768332120220346,
      "learning_rate": 1e-06,
      "loss": 0.304,
      "mean_token_accuracy": 0.8922690749168396,
      "num_tokens": 544546731.0,
      "step": 15648
    },
    {
      "epoch": 2.9060352831940577,
      "grad_norm": 1.8171084701457374,
      "learning_rate": 1e-06,
      "loss": 0.366,
      "mean_token_accuracy": 0.8759661912918091,
      "num_tokens": 544575210.0,
      "step": 15649
    },
    {
      "epoch": 2.9062209842154134,
      "grad_norm": 1.6581334573765403,
      "learning_rate": 1e-06,
      "loss": 0.3319,
      "mean_token_accuracy": 0.8866223096847534,
      "num_tokens": 544609301.0,
      "step": 15650
    },
    {
      "epoch": 2.9064066852367687,
      "grad_norm": 1.4142898694875676,
      "learning_rate": 1e-06,
      "loss": 0.3155,
      "mean_token_accuracy": 0.8848758935928345,
      "num_tokens": 544651197.0,
      "step": 15651
    },
    {
      "epoch": 2.9065923862581244,
      "grad_norm": 1.3623431115077076,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.892410159111023,
      "num_tokens": 544692454.0,
      "step": 15652
    },
    {
      "epoch": 2.90677808727948,
      "grad_norm": 1.575397471840925,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8896531462669373,
      "num_tokens": 544726659.0,
      "step": 15653
    },
    {
      "epoch": 2.9069637883008355,
      "grad_norm": 1.4753730939264629,
      "learning_rate": 1e-06,
      "loss": 0.3018,
      "mean_token_accuracy": 0.8916328549385071,
      "num_tokens": 544761181.0,
      "step": 15654
    },
    {
      "epoch": 2.907149489322191,
      "grad_norm": 1.3990083561342093,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8897168040275574,
      "num_tokens": 544801092.0,
      "step": 15655
    },
    {
      "epoch": 2.907335190343547,
      "grad_norm": 1.5245531221228064,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8906688690185547,
      "num_tokens": 544836964.0,
      "step": 15656
    },
    {
      "epoch": 2.9075208913649027,
      "grad_norm": 1.5019163656325007,
      "learning_rate": 1e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8790469765663147,
      "num_tokens": 544873719.0,
      "step": 15657
    },
    {
      "epoch": 2.9077065923862584,
      "grad_norm": 1.6570947996738778,
      "learning_rate": 1e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8685088157653809,
      "num_tokens": 544909962.0,
      "step": 15658
    },
    {
      "epoch": 2.9078922934076137,
      "grad_norm": 1.3996586641463955,
      "learning_rate": 1e-06,
      "loss": 0.2827,
      "mean_token_accuracy": 0.9003647565841675,
      "num_tokens": 544950081.0,
      "step": 15659
    },
    {
      "epoch": 2.9080779944289694,
      "grad_norm": 1.4489432672017746,
      "learning_rate": 1e-06,
      "loss": 0.2766,
      "mean_token_accuracy": 0.8988691568374634,
      "num_tokens": 544984793.0,
      "step": 15660
    },
    {
      "epoch": 2.9082636954503247,
      "grad_norm": 1.7488904846318973,
      "learning_rate": 1e-06,
      "loss": 0.3407,
      "mean_token_accuracy": 0.8879765868186951,
      "num_tokens": 545013653.0,
      "step": 15661
    },
    {
      "epoch": 2.9084493964716804,
      "grad_norm": 1.4073911554899,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8940283060073853,
      "num_tokens": 545054644.0,
      "step": 15662
    },
    {
      "epoch": 2.908635097493036,
      "grad_norm": 1.6521779817187787,
      "learning_rate": 1e-06,
      "loss": 0.3567,
      "mean_token_accuracy": 0.876017689704895,
      "num_tokens": 545087866.0,
      "step": 15663
    },
    {
      "epoch": 2.908820798514392,
      "grad_norm": 1.534437975564622,
      "learning_rate": 1e-06,
      "loss": 0.3304,
      "mean_token_accuracy": 0.88304603099823,
      "num_tokens": 545124180.0,
      "step": 15664
    },
    {
      "epoch": 2.9090064995357476,
      "grad_norm": 1.5549661622096764,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8754660487174988,
      "num_tokens": 545161513.0,
      "step": 15665
    },
    {
      "epoch": 2.909192200557103,
      "grad_norm": 1.5352323126378657,
      "learning_rate": 1e-06,
      "loss": 0.344,
      "mean_token_accuracy": 0.8834143280982971,
      "num_tokens": 545199738.0,
      "step": 15666
    },
    {
      "epoch": 2.9093779015784587,
      "grad_norm": 1.5703966242420457,
      "learning_rate": 1e-06,
      "loss": 0.2876,
      "mean_token_accuracy": 0.8976658582687378,
      "num_tokens": 545231230.0,
      "step": 15667
    },
    {
      "epoch": 2.9095636025998144,
      "grad_norm": 1.5625628931727087,
      "learning_rate": 1e-06,
      "loss": 0.2879,
      "mean_token_accuracy": 0.8958833813667297,
      "num_tokens": 545262270.0,
      "step": 15668
    },
    {
      "epoch": 2.9097493036211697,
      "grad_norm": 1.7069587898177168,
      "learning_rate": 1e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8775880336761475,
      "num_tokens": 545297262.0,
      "step": 15669
    },
    {
      "epoch": 2.9099350046425254,
      "grad_norm": 1.4975385321110575,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.8897726535797119,
      "num_tokens": 545334311.0,
      "step": 15670
    },
    {
      "epoch": 2.910120705663881,
      "grad_norm": 1.5839459678852414,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8764687180519104,
      "num_tokens": 545368583.0,
      "step": 15671
    },
    {
      "epoch": 2.910306406685237,
      "grad_norm": 1.5300565534085777,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8873623013496399,
      "num_tokens": 545401486.0,
      "step": 15672
    },
    {
      "epoch": 2.9104921077065926,
      "grad_norm": 1.499639587600636,
      "learning_rate": 1e-06,
      "loss": 0.3113,
      "mean_token_accuracy": 0.891230046749115,
      "num_tokens": 545440574.0,
      "step": 15673
    },
    {
      "epoch": 2.910677808727948,
      "grad_norm": 1.4280416735734818,
      "learning_rate": 1e-06,
      "loss": 0.2944,
      "mean_token_accuracy": 0.8957010507583618,
      "num_tokens": 545478864.0,
      "step": 15674
    },
    {
      "epoch": 2.9108635097493036,
      "grad_norm": 1.5105735900730912,
      "learning_rate": 1e-06,
      "loss": 0.2814,
      "mean_token_accuracy": 0.899368166923523,
      "num_tokens": 545513657.0,
      "step": 15675
    },
    {
      "epoch": 2.9110492107706594,
      "grad_norm": 1.4922757457036482,
      "learning_rate": 1e-06,
      "loss": 0.3228,
      "mean_token_accuracy": 0.8849144577980042,
      "num_tokens": 545552414.0,
      "step": 15676
    },
    {
      "epoch": 2.9112349117920147,
      "grad_norm": 1.465290257440104,
      "learning_rate": 1e-06,
      "loss": 0.2907,
      "mean_token_accuracy": 0.8960756659507751,
      "num_tokens": 545590669.0,
      "step": 15677
    },
    {
      "epoch": 2.9114206128133704,
      "grad_norm": 1.5275881687119925,
      "learning_rate": 1e-06,
      "loss": 0.2671,
      "mean_token_accuracy": 0.904163122177124,
      "num_tokens": 545624296.0,
      "step": 15678
    },
    {
      "epoch": 2.911606313834726,
      "grad_norm": 1.5646231864523794,
      "learning_rate": 1e-06,
      "loss": 0.3105,
      "mean_token_accuracy": 0.8913389444351196,
      "num_tokens": 545655938.0,
      "step": 15679
    },
    {
      "epoch": 2.911792014856082,
      "grad_norm": 1.7695360967619205,
      "learning_rate": 1e-06,
      "loss": 0.3447,
      "mean_token_accuracy": 0.8779604434967041,
      "num_tokens": 545683563.0,
      "step": 15680
    },
    {
      "epoch": 2.9119777158774376,
      "grad_norm": 1.421725426246588,
      "learning_rate": 1e-06,
      "loss": 0.2975,
      "mean_token_accuracy": 0.894725501537323,
      "num_tokens": 545725319.0,
      "step": 15681
    },
    {
      "epoch": 2.912163416898793,
      "grad_norm": 1.5760253997425442,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8855414986610413,
      "num_tokens": 545761711.0,
      "step": 15682
    },
    {
      "epoch": 2.9123491179201486,
      "grad_norm": 1.6455751584857814,
      "learning_rate": 1e-06,
      "loss": 0.3232,
      "mean_token_accuracy": 0.8883504867553711,
      "num_tokens": 545793054.0,
      "step": 15683
    },
    {
      "epoch": 2.9125348189415043,
      "grad_norm": 1.5748980300070077,
      "learning_rate": 1e-06,
      "loss": 0.2936,
      "mean_token_accuracy": 0.8933820724487305,
      "num_tokens": 545825996.0,
      "step": 15684
    },
    {
      "epoch": 2.9127205199628596,
      "grad_norm": 1.5489972761942257,
      "learning_rate": 1e-06,
      "loss": 0.3209,
      "mean_token_accuracy": 0.885273814201355,
      "num_tokens": 545862453.0,
      "step": 15685
    },
    {
      "epoch": 2.9129062209842154,
      "grad_norm": 1.5723184280138394,
      "learning_rate": 1e-06,
      "loss": 0.3187,
      "mean_token_accuracy": 0.8892428278923035,
      "num_tokens": 545895359.0,
      "step": 15686
    },
    {
      "epoch": 2.913091922005571,
      "grad_norm": 1.5777126296484507,
      "learning_rate": 1e-06,
      "loss": 0.3171,
      "mean_token_accuracy": 0.888338565826416,
      "num_tokens": 545929838.0,
      "step": 15687
    },
    {
      "epoch": 2.913277623026927,
      "grad_norm": 1.6492122054655962,
      "learning_rate": 1e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8712589740753174,
      "num_tokens": 545966869.0,
      "step": 15688
    },
    {
      "epoch": 2.913463324048282,
      "grad_norm": 1.5954152983213057,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8900792598724365,
      "num_tokens": 546002167.0,
      "step": 15689
    },
    {
      "epoch": 2.913649025069638,
      "grad_norm": 1.7236059483503536,
      "learning_rate": 1e-06,
      "loss": 0.3147,
      "mean_token_accuracy": 0.8890571594238281,
      "num_tokens": 546029999.0,
      "step": 15690
    },
    {
      "epoch": 2.9138347260909936,
      "grad_norm": 1.514803474945325,
      "learning_rate": 1e-06,
      "loss": 0.3015,
      "mean_token_accuracy": 0.892598032951355,
      "num_tokens": 546063433.0,
      "step": 15691
    },
    {
      "epoch": 2.914020427112349,
      "grad_norm": 1.5868747070700222,
      "learning_rate": 1e-06,
      "loss": 0.3471,
      "mean_token_accuracy": 0.8795801997184753,
      "num_tokens": 546105138.0,
      "step": 15692
    },
    {
      "epoch": 2.9142061281337046,
      "grad_norm": 1.5387028921086399,
      "learning_rate": 1e-06,
      "loss": 0.2779,
      "mean_token_accuracy": 0.9023420810699463,
      "num_tokens": 546137408.0,
      "step": 15693
    },
    {
      "epoch": 2.9143918291550603,
      "grad_norm": 1.7293157873796945,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8890054225921631,
      "num_tokens": 546164166.0,
      "step": 15694
    },
    {
      "epoch": 2.914577530176416,
      "grad_norm": 1.4912066131533421,
      "learning_rate": 1e-06,
      "loss": 0.3067,
      "mean_token_accuracy": 0.8899931311607361,
      "num_tokens": 546203375.0,
      "step": 15695
    },
    {
      "epoch": 2.914763231197772,
      "grad_norm": 1.5306281807518947,
      "learning_rate": 1e-06,
      "loss": 0.3691,
      "mean_token_accuracy": 0.8706724643707275,
      "num_tokens": 546243451.0,
      "step": 15696
    },
    {
      "epoch": 2.914948932219127,
      "grad_norm": 1.5474284907882088,
      "learning_rate": 1e-06,
      "loss": 0.3056,
      "mean_token_accuracy": 0.893714189529419,
      "num_tokens": 546281430.0,
      "step": 15697
    },
    {
      "epoch": 2.915134633240483,
      "grad_norm": 1.4815389315701901,
      "learning_rate": 1e-06,
      "loss": 0.2983,
      "mean_token_accuracy": 0.8952798247337341,
      "num_tokens": 546320622.0,
      "step": 15698
    },
    {
      "epoch": 2.9153203342618386,
      "grad_norm": 1.6186840368227136,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.878459095954895,
      "num_tokens": 546355267.0,
      "step": 15699
    },
    {
      "epoch": 2.915506035283194,
      "grad_norm": 1.5223144955117214,
      "learning_rate": 1e-06,
      "loss": 0.3117,
      "mean_token_accuracy": 0.8934271335601807,
      "num_tokens": 546390536.0,
      "step": 15700
    },
    {
      "epoch": 2.9156917363045496,
      "grad_norm": 1.487685636069701,
      "learning_rate": 1e-06,
      "loss": 0.3335,
      "mean_token_accuracy": 0.8858348727226257,
      "num_tokens": 546430795.0,
      "step": 15701
    },
    {
      "epoch": 2.9158774373259053,
      "grad_norm": 1.5140381632138402,
      "learning_rate": 1e-06,
      "loss": 0.3331,
      "mean_token_accuracy": 0.8833856582641602,
      "num_tokens": 546467956.0,
      "step": 15702
    },
    {
      "epoch": 2.916063138347261,
      "grad_norm": 1.4771060585123155,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8818755745887756,
      "num_tokens": 546506047.0,
      "step": 15703
    },
    {
      "epoch": 2.916248839368617,
      "grad_norm": 1.4699911604897327,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8954601287841797,
      "num_tokens": 546543187.0,
      "step": 15704
    },
    {
      "epoch": 2.916434540389972,
      "grad_norm": 1.5833104889499803,
      "learning_rate": 1e-06,
      "loss": 0.3562,
      "mean_token_accuracy": 0.8748331069946289,
      "num_tokens": 546576978.0,
      "step": 15705
    },
    {
      "epoch": 2.916620241411328,
      "grad_norm": 1.9884026519635793,
      "learning_rate": 1e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8726885914802551,
      "num_tokens": 546603213.0,
      "step": 15706
    },
    {
      "epoch": 2.9168059424326835,
      "grad_norm": 1.6771642292816726,
      "learning_rate": 1e-06,
      "loss": 0.3267,
      "mean_token_accuracy": 0.8881444931030273,
      "num_tokens": 546632486.0,
      "step": 15707
    },
    {
      "epoch": 2.916991643454039,
      "grad_norm": 1.5489075538748303,
      "learning_rate": 1e-06,
      "loss": 0.3395,
      "mean_token_accuracy": 0.8800239562988281,
      "num_tokens": 546670171.0,
      "step": 15708
    },
    {
      "epoch": 2.9171773444753946,
      "grad_norm": 1.4138229410306566,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8905583620071411,
      "num_tokens": 546708966.0,
      "step": 15709
    },
    {
      "epoch": 2.9173630454967503,
      "grad_norm": 1.7494754786684061,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8966509103775024,
      "num_tokens": 546736868.0,
      "step": 15710
    },
    {
      "epoch": 2.917548746518106,
      "grad_norm": 1.7303612903342875,
      "learning_rate": 1e-06,
      "loss": 0.2922,
      "mean_token_accuracy": 0.8966612815856934,
      "num_tokens": 546763993.0,
      "step": 15711
    },
    {
      "epoch": 2.9177344475394613,
      "grad_norm": 1.53262456859773,
      "learning_rate": 1e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8736934661865234,
      "num_tokens": 546803612.0,
      "step": 15712
    },
    {
      "epoch": 2.917920148560817,
      "grad_norm": 1.4221913287856065,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.894730806350708,
      "num_tokens": 546841424.0,
      "step": 15713
    },
    {
      "epoch": 2.918105849582173,
      "grad_norm": 1.6424093451334616,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8788882493972778,
      "num_tokens": 546876085.0,
      "step": 15714
    },
    {
      "epoch": 2.918291550603528,
      "grad_norm": 1.5733493037336055,
      "learning_rate": 1e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8780707120895386,
      "num_tokens": 546910506.0,
      "step": 15715
    },
    {
      "epoch": 2.918477251624884,
      "grad_norm": 1.5738154608309078,
      "learning_rate": 1e-06,
      "loss": 0.3528,
      "mean_token_accuracy": 0.8785661458969116,
      "num_tokens": 546949509.0,
      "step": 15716
    },
    {
      "epoch": 2.9186629526462395,
      "grad_norm": 1.5223911007353894,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8910167813301086,
      "num_tokens": 546983568.0,
      "step": 15717
    },
    {
      "epoch": 2.9188486536675953,
      "grad_norm": 1.540779154222672,
      "learning_rate": 1e-06,
      "loss": 0.34,
      "mean_token_accuracy": 0.8801407217979431,
      "num_tokens": 547021648.0,
      "step": 15718
    },
    {
      "epoch": 2.919034354688951,
      "grad_norm": 1.5120660276887636,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8879497051239014,
      "num_tokens": 547057575.0,
      "step": 15719
    },
    {
      "epoch": 2.9192200557103063,
      "grad_norm": 1.5189215850351068,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8799747228622437,
      "num_tokens": 547094567.0,
      "step": 15720
    },
    {
      "epoch": 2.919405756731662,
      "grad_norm": 1.6277363067227486,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8767607808113098,
      "num_tokens": 547125675.0,
      "step": 15721
    },
    {
      "epoch": 2.9195914577530178,
      "grad_norm": 1.4411123625967657,
      "learning_rate": 1e-06,
      "loss": 0.2633,
      "mean_token_accuracy": 0.9049414992332458,
      "num_tokens": 547159510.0,
      "step": 15722
    },
    {
      "epoch": 2.919777158774373,
      "grad_norm": 1.5610873414407676,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8914834856987,
      "num_tokens": 547192126.0,
      "step": 15723
    },
    {
      "epoch": 2.919962859795729,
      "grad_norm": 1.5178058025434915,
      "learning_rate": 1e-06,
      "loss": 0.3254,
      "mean_token_accuracy": 0.884348452091217,
      "num_tokens": 547228623.0,
      "step": 15724
    },
    {
      "epoch": 2.9201485608170845,
      "grad_norm": 1.6875589710784114,
      "learning_rate": 1e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8814383745193481,
      "num_tokens": 547262156.0,
      "step": 15725
    },
    {
      "epoch": 2.9203342618384402,
      "grad_norm": 1.6366645161229496,
      "learning_rate": 1e-06,
      "loss": 0.3215,
      "mean_token_accuracy": 0.8881365656852722,
      "num_tokens": 547295498.0,
      "step": 15726
    },
    {
      "epoch": 2.920519962859796,
      "grad_norm": 1.5456501217749774,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8913203477859497,
      "num_tokens": 547328323.0,
      "step": 15727
    },
    {
      "epoch": 2.9207056638811513,
      "grad_norm": 1.5592200999630434,
      "learning_rate": 1e-06,
      "loss": 0.3347,
      "mean_token_accuracy": 0.8820874094963074,
      "num_tokens": 547364367.0,
      "step": 15728
    },
    {
      "epoch": 2.920891364902507,
      "grad_norm": 1.5343666874717283,
      "learning_rate": 1e-06,
      "loss": 0.3559,
      "mean_token_accuracy": 0.8746237754821777,
      "num_tokens": 547403223.0,
      "step": 15729
    },
    {
      "epoch": 2.9210770659238627,
      "grad_norm": 1.7393365817728266,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8852742910385132,
      "num_tokens": 547433870.0,
      "step": 15730
    },
    {
      "epoch": 2.921262766945218,
      "grad_norm": 1.673733621613849,
      "learning_rate": 1e-06,
      "loss": 0.3507,
      "mean_token_accuracy": 0.8856480121612549,
      "num_tokens": 547468187.0,
      "step": 15731
    },
    {
      "epoch": 2.9214484679665738,
      "grad_norm": 1.4838386761818514,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8864346742630005,
      "num_tokens": 547502563.0,
      "step": 15732
    },
    {
      "epoch": 2.9216341689879295,
      "grad_norm": 1.537243257169702,
      "learning_rate": 1e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.886530876159668,
      "num_tokens": 547540105.0,
      "step": 15733
    },
    {
      "epoch": 2.9218198700092852,
      "grad_norm": 1.4285964836962939,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8860245943069458,
      "num_tokens": 547583057.0,
      "step": 15734
    },
    {
      "epoch": 2.9220055710306405,
      "grad_norm": 1.565366786120635,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8956310749053955,
      "num_tokens": 547617700.0,
      "step": 15735
    },
    {
      "epoch": 2.9221912720519962,
      "grad_norm": 1.7245214873867485,
      "learning_rate": 1e-06,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8791323304176331,
      "num_tokens": 547648386.0,
      "step": 15736
    },
    {
      "epoch": 2.922376973073352,
      "grad_norm": 1.4974395572045023,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.8846422433853149,
      "num_tokens": 547686992.0,
      "step": 15737
    },
    {
      "epoch": 2.9225626740947073,
      "grad_norm": 1.639415609208378,
      "learning_rate": 1e-06,
      "loss": 0.2976,
      "mean_token_accuracy": 0.8916422128677368,
      "num_tokens": 547716709.0,
      "step": 15738
    },
    {
      "epoch": 2.922748375116063,
      "grad_norm": 1.4662171843547576,
      "learning_rate": 1e-06,
      "loss": 0.3434,
      "mean_token_accuracy": 0.8759864568710327,
      "num_tokens": 547756946.0,
      "step": 15739
    },
    {
      "epoch": 2.9229340761374187,
      "grad_norm": 1.448116407622598,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8948856592178345,
      "num_tokens": 547794945.0,
      "step": 15740
    },
    {
      "epoch": 2.9231197771587745,
      "grad_norm": 1.526048151504499,
      "learning_rate": 1e-06,
      "loss": 0.2986,
      "mean_token_accuracy": 0.8941411972045898,
      "num_tokens": 547826492.0,
      "step": 15741
    },
    {
      "epoch": 2.92330547818013,
      "grad_norm": 1.522518905376614,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8736367225646973,
      "num_tokens": 547868280.0,
      "step": 15742
    },
    {
      "epoch": 2.9234911792014855,
      "grad_norm": 1.4368269911599854,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8786314725875854,
      "num_tokens": 547911916.0,
      "step": 15743
    },
    {
      "epoch": 2.9236768802228412,
      "grad_norm": 1.7891133052432626,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8950031399726868,
      "num_tokens": 547935650.0,
      "step": 15744
    },
    {
      "epoch": 2.923862581244197,
      "grad_norm": 1.6230839963889105,
      "learning_rate": 1e-06,
      "loss": 0.3389,
      "mean_token_accuracy": 0.8842551112174988,
      "num_tokens": 547970788.0,
      "step": 15745
    },
    {
      "epoch": 2.9240482822655522,
      "grad_norm": 1.544497816445551,
      "learning_rate": 1e-06,
      "loss": 0.2988,
      "mean_token_accuracy": 0.8942838907241821,
      "num_tokens": 548003695.0,
      "step": 15746
    },
    {
      "epoch": 2.924233983286908,
      "grad_norm": 1.4956626486487143,
      "learning_rate": 1e-06,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8906763792037964,
      "num_tokens": 548041886.0,
      "step": 15747
    },
    {
      "epoch": 2.9244196843082637,
      "grad_norm": 1.6939615646926236,
      "learning_rate": 1e-06,
      "loss": 0.3181,
      "mean_token_accuracy": 0.8896665573120117,
      "num_tokens": 548069219.0,
      "step": 15748
    },
    {
      "epoch": 2.9246053853296194,
      "grad_norm": 1.5663705693363659,
      "learning_rate": 1e-06,
      "loss": 0.328,
      "mean_token_accuracy": 0.8842328786849976,
      "num_tokens": 548103700.0,
      "step": 15749
    },
    {
      "epoch": 2.924791086350975,
      "grad_norm": 1.5378890484481826,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8865962028503418,
      "num_tokens": 548145724.0,
      "step": 15750
    },
    {
      "epoch": 2.9249767873723305,
      "grad_norm": 1.5402697663841725,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8924410343170166,
      "num_tokens": 548178210.0,
      "step": 15751
    },
    {
      "epoch": 2.925162488393686,
      "grad_norm": 1.8599334813443298,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8916044235229492,
      "num_tokens": 548217000.0,
      "step": 15752
    },
    {
      "epoch": 2.925348189415042,
      "grad_norm": 1.7656633061880502,
      "learning_rate": 1e-06,
      "loss": 0.3178,
      "mean_token_accuracy": 0.8882527947425842,
      "num_tokens": 548245810.0,
      "step": 15753
    },
    {
      "epoch": 2.925533890436397,
      "grad_norm": 1.4740964991062953,
      "learning_rate": 1e-06,
      "loss": 0.2695,
      "mean_token_accuracy": 0.9035781621932983,
      "num_tokens": 548282509.0,
      "step": 15754
    },
    {
      "epoch": 2.925719591457753,
      "grad_norm": 1.5328270457535944,
      "learning_rate": 1e-06,
      "loss": 0.3157,
      "mean_token_accuracy": 0.8912572264671326,
      "num_tokens": 548315679.0,
      "step": 15755
    },
    {
      "epoch": 2.9259052924791087,
      "grad_norm": 1.5675573182319202,
      "learning_rate": 1e-06,
      "loss": 0.3287,
      "mean_token_accuracy": 0.886834979057312,
      "num_tokens": 548351219.0,
      "step": 15756
    },
    {
      "epoch": 2.9260909935004644,
      "grad_norm": 1.5367423218726346,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8939040899276733,
      "num_tokens": 548383706.0,
      "step": 15757
    },
    {
      "epoch": 2.9262766945218197,
      "grad_norm": 1.4753058966130508,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.891992449760437,
      "num_tokens": 548422330.0,
      "step": 15758
    },
    {
      "epoch": 2.9264623955431754,
      "grad_norm": 1.616916958460536,
      "learning_rate": 1e-06,
      "loss": 0.3462,
      "mean_token_accuracy": 0.8761382699012756,
      "num_tokens": 548459903.0,
      "step": 15759
    },
    {
      "epoch": 2.926648096564531,
      "grad_norm": 1.4826659960481219,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8912761211395264,
      "num_tokens": 548497402.0,
      "step": 15760
    },
    {
      "epoch": 2.9268337975858865,
      "grad_norm": 1.4362395901330198,
      "learning_rate": 1e-06,
      "loss": 0.2841,
      "mean_token_accuracy": 0.8982065320014954,
      "num_tokens": 548533398.0,
      "step": 15761
    },
    {
      "epoch": 2.927019498607242,
      "grad_norm": 1.4874083556658133,
      "learning_rate": 1e-06,
      "loss": 0.284,
      "mean_token_accuracy": 0.8979848027229309,
      "num_tokens": 548568774.0,
      "step": 15762
    },
    {
      "epoch": 2.927205199628598,
      "grad_norm": 1.4837853469923103,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8926987051963806,
      "num_tokens": 548606074.0,
      "step": 15763
    },
    {
      "epoch": 2.9273909006499537,
      "grad_norm": 1.5357689317690855,
      "learning_rate": 1e-06,
      "loss": 0.3428,
      "mean_token_accuracy": 0.879223108291626,
      "num_tokens": 548642555.0,
      "step": 15764
    },
    {
      "epoch": 2.9275766016713094,
      "grad_norm": 1.6549444900367758,
      "learning_rate": 1e-06,
      "loss": 0.3497,
      "mean_token_accuracy": 0.8855069875717163,
      "num_tokens": 548677511.0,
      "step": 15765
    },
    {
      "epoch": 2.9277623026926647,
      "grad_norm": 1.6971118960676443,
      "learning_rate": 1e-06,
      "loss": 0.313,
      "mean_token_accuracy": 0.8885579109191895,
      "num_tokens": 548706545.0,
      "step": 15766
    },
    {
      "epoch": 2.9279480037140204,
      "grad_norm": 1.5718718192673216,
      "learning_rate": 1e-06,
      "loss": 0.3357,
      "mean_token_accuracy": 0.8801887631416321,
      "num_tokens": 548743465.0,
      "step": 15767
    },
    {
      "epoch": 2.928133704735376,
      "grad_norm": 1.4506452401288454,
      "learning_rate": 1e-06,
      "loss": 0.3325,
      "mean_token_accuracy": 0.8824766874313354,
      "num_tokens": 548783942.0,
      "step": 15768
    },
    {
      "epoch": 2.9283194057567314,
      "grad_norm": 1.6126733292034774,
      "learning_rate": 1e-06,
      "loss": 0.297,
      "mean_token_accuracy": 0.8938933610916138,
      "num_tokens": 548815700.0,
      "step": 15769
    },
    {
      "epoch": 2.928505106778087,
      "grad_norm": 1.8790882241756806,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8636997938156128,
      "num_tokens": 548845876.0,
      "step": 15770
    },
    {
      "epoch": 2.928690807799443,
      "grad_norm": 1.6205335317421752,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8775429725646973,
      "num_tokens": 548878221.0,
      "step": 15771
    },
    {
      "epoch": 2.9288765088207986,
      "grad_norm": 1.5991861758956927,
      "learning_rate": 1e-06,
      "loss": 0.3665,
      "mean_token_accuracy": 0.8702864050865173,
      "num_tokens": 548913083.0,
      "step": 15772
    },
    {
      "epoch": 2.9290622098421544,
      "grad_norm": 1.7358070095468279,
      "learning_rate": 1e-06,
      "loss": 0.3279,
      "mean_token_accuracy": 0.8859553337097168,
      "num_tokens": 548942647.0,
      "step": 15773
    },
    {
      "epoch": 2.9292479108635097,
      "grad_norm": 1.6090893186230792,
      "learning_rate": 1e-06,
      "loss": 0.3144,
      "mean_token_accuracy": 0.8889545202255249,
      "num_tokens": 548973428.0,
      "step": 15774
    },
    {
      "epoch": 2.9294336118848654,
      "grad_norm": 1.6207243260384672,
      "learning_rate": 1e-06,
      "loss": 0.3501,
      "mean_token_accuracy": 0.8778924942016602,
      "num_tokens": 549006471.0,
      "step": 15775
    },
    {
      "epoch": 2.929619312906221,
      "grad_norm": 1.5641463788602052,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8941453695297241,
      "num_tokens": 549038710.0,
      "step": 15776
    },
    {
      "epoch": 2.9298050139275764,
      "grad_norm": 1.5450920127591015,
      "learning_rate": 1e-06,
      "loss": 0.2831,
      "mean_token_accuracy": 0.8996148705482483,
      "num_tokens": 549068827.0,
      "step": 15777
    },
    {
      "epoch": 2.929990714948932,
      "grad_norm": 1.6201619965973277,
      "learning_rate": 1e-06,
      "loss": 0.375,
      "mean_token_accuracy": 0.8676267266273499,
      "num_tokens": 549103451.0,
      "step": 15778
    },
    {
      "epoch": 2.930176415970288,
      "grad_norm": 1.398080378612398,
      "learning_rate": 1e-06,
      "loss": 0.312,
      "mean_token_accuracy": 0.8911311030387878,
      "num_tokens": 549143273.0,
      "step": 15779
    },
    {
      "epoch": 2.9303621169916436,
      "grad_norm": 1.435004936479876,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8870713710784912,
      "num_tokens": 549185074.0,
      "step": 15780
    },
    {
      "epoch": 2.930547818012999,
      "grad_norm": 1.492400214409349,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8879958391189575,
      "num_tokens": 549223429.0,
      "step": 15781
    },
    {
      "epoch": 2.9307335190343546,
      "grad_norm": 1.6997147162707833,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8824796676635742,
      "num_tokens": 549255728.0,
      "step": 15782
    },
    {
      "epoch": 2.9309192200557104,
      "grad_norm": 1.6473504666513352,
      "learning_rate": 1e-06,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8833295106887817,
      "num_tokens": 549289308.0,
      "step": 15783
    },
    {
      "epoch": 2.9311049210770657,
      "grad_norm": 1.6668699580959658,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8933743238449097,
      "num_tokens": 549317230.0,
      "step": 15784
    },
    {
      "epoch": 2.9312906220984214,
      "grad_norm": 1.662926538746697,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.8810016512870789,
      "num_tokens": 549346920.0,
      "step": 15785
    },
    {
      "epoch": 2.931476323119777,
      "grad_norm": 1.6312404972823118,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8806362152099609,
      "num_tokens": 549379690.0,
      "step": 15786
    },
    {
      "epoch": 2.931662024141133,
      "grad_norm": 1.5035484064405933,
      "learning_rate": 1e-06,
      "loss": 0.3345,
      "mean_token_accuracy": 0.883664608001709,
      "num_tokens": 549416580.0,
      "step": 15787
    },
    {
      "epoch": 2.9318477251624886,
      "grad_norm": 1.6404596780508784,
      "learning_rate": 1e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8786689043045044,
      "num_tokens": 549450103.0,
      "step": 15788
    },
    {
      "epoch": 2.932033426183844,
      "grad_norm": 1.6116492439259031,
      "learning_rate": 1e-06,
      "loss": 0.3369,
      "mean_token_accuracy": 0.8808140754699707,
      "num_tokens": 549482647.0,
      "step": 15789
    },
    {
      "epoch": 2.9322191272051996,
      "grad_norm": 1.5743164698151806,
      "learning_rate": 1e-06,
      "loss": 0.308,
      "mean_token_accuracy": 0.8899757266044617,
      "num_tokens": 549518510.0,
      "step": 15790
    },
    {
      "epoch": 2.9324048282265553,
      "grad_norm": 1.6240595569843734,
      "learning_rate": 1e-06,
      "loss": 0.3572,
      "mean_token_accuracy": 0.8766289353370667,
      "num_tokens": 549554417.0,
      "step": 15791
    },
    {
      "epoch": 2.9325905292479106,
      "grad_norm": 1.5145167648680085,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.8943266868591309,
      "num_tokens": 549590693.0,
      "step": 15792
    },
    {
      "epoch": 2.9327762302692664,
      "grad_norm": 1.60547166736118,
      "learning_rate": 1e-06,
      "loss": 0.3235,
      "mean_token_accuracy": 0.8888919353485107,
      "num_tokens": 549625982.0,
      "step": 15793
    },
    {
      "epoch": 2.932961931290622,
      "grad_norm": 1.4657350537593816,
      "learning_rate": 1e-06,
      "loss": 0.3057,
      "mean_token_accuracy": 0.8932647109031677,
      "num_tokens": 549664396.0,
      "step": 15794
    },
    {
      "epoch": 2.933147632311978,
      "grad_norm": 1.511121840934249,
      "learning_rate": 1e-06,
      "loss": 0.3018,
      "mean_token_accuracy": 0.8950895071029663,
      "num_tokens": 549698895.0,
      "step": 15795
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 1.558682229504336,
      "learning_rate": 1e-06,
      "loss": 0.3207,
      "mean_token_accuracy": 0.8860712647438049,
      "num_tokens": 549733597.0,
      "step": 15796
    },
    {
      "epoch": 2.933519034354689,
      "grad_norm": 1.7921974654695842,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.88587486743927,
      "num_tokens": 549759579.0,
      "step": 15797
    },
    {
      "epoch": 2.9337047353760446,
      "grad_norm": 1.5091845229848921,
      "learning_rate": 1e-06,
      "loss": 0.2955,
      "mean_token_accuracy": 0.8931675553321838,
      "num_tokens": 549794494.0,
      "step": 15798
    },
    {
      "epoch": 2.9338904363974003,
      "grad_norm": 1.3999977013075742,
      "learning_rate": 1e-06,
      "loss": 0.285,
      "mean_token_accuracy": 0.9024381637573242,
      "num_tokens": 549833829.0,
      "step": 15799
    },
    {
      "epoch": 2.9340761374187556,
      "grad_norm": 1.5892667037806834,
      "learning_rate": 1e-06,
      "loss": 0.3202,
      "mean_token_accuracy": 0.8881828188896179,
      "num_tokens": 549866807.0,
      "step": 15800
    },
    {
      "epoch": 2.9342618384401113,
      "grad_norm": 1.7837161335003153,
      "learning_rate": 1e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.8726474046707153,
      "num_tokens": 549894129.0,
      "step": 15801
    },
    {
      "epoch": 2.934447539461467,
      "grad_norm": 1.4594575474159701,
      "learning_rate": 1e-06,
      "loss": 0.3029,
      "mean_token_accuracy": 0.8951870203018188,
      "num_tokens": 549930762.0,
      "step": 15802
    },
    {
      "epoch": 2.934633240482823,
      "grad_norm": 1.8035939508982852,
      "learning_rate": 1e-06,
      "loss": 0.3409,
      "mean_token_accuracy": 0.8819375038146973,
      "num_tokens": 549958837.0,
      "step": 15803
    },
    {
      "epoch": 2.9348189415041785,
      "grad_norm": 1.6624182201044582,
      "learning_rate": 1e-06,
      "loss": 0.318,
      "mean_token_accuracy": 0.8888909220695496,
      "num_tokens": 549991221.0,
      "step": 15804
    },
    {
      "epoch": 2.935004642525534,
      "grad_norm": 1.5407826634925534,
      "learning_rate": 1e-06,
      "loss": 0.2941,
      "mean_token_accuracy": 0.8966519832611084,
      "num_tokens": 550026821.0,
      "step": 15805
    },
    {
      "epoch": 2.9351903435468896,
      "grad_norm": 1.6958618378139083,
      "learning_rate": 1e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.8835249543190002,
      "num_tokens": 550058495.0,
      "step": 15806
    },
    {
      "epoch": 2.935376044568245,
      "grad_norm": 1.5799381243229407,
      "learning_rate": 1e-06,
      "loss": 0.2991,
      "mean_token_accuracy": 0.8903219699859619,
      "num_tokens": 550092380.0,
      "step": 15807
    },
    {
      "epoch": 2.9355617455896006,
      "grad_norm": 1.6231290544623753,
      "learning_rate": 1e-06,
      "loss": 0.2785,
      "mean_token_accuracy": 0.9053640365600586,
      "num_tokens": 550119563.0,
      "step": 15808
    },
    {
      "epoch": 2.9357474466109563,
      "grad_norm": 1.6953681168635655,
      "learning_rate": 1e-06,
      "loss": 0.2948,
      "mean_token_accuracy": 0.8970731496810913,
      "num_tokens": 550151718.0,
      "step": 15809
    },
    {
      "epoch": 2.935933147632312,
      "grad_norm": 1.7347557100982374,
      "learning_rate": 1e-06,
      "loss": 0.3444,
      "mean_token_accuracy": 0.8852831125259399,
      "num_tokens": 550180061.0,
      "step": 15810
    },
    {
      "epoch": 2.936118848653668,
      "grad_norm": 1.5128550578809008,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8946871757507324,
      "num_tokens": 550214529.0,
      "step": 15811
    },
    {
      "epoch": 2.936304549675023,
      "grad_norm": 1.64140436716936,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8821908831596375,
      "num_tokens": 550250864.0,
      "step": 15812
    },
    {
      "epoch": 2.936490250696379,
      "grad_norm": 1.7378999129383976,
      "learning_rate": 1e-06,
      "loss": 0.3363,
      "mean_token_accuracy": 0.8863251209259033,
      "num_tokens": 550280808.0,
      "step": 15813
    },
    {
      "epoch": 2.9366759517177345,
      "grad_norm": 1.3558873731626409,
      "learning_rate": 1e-06,
      "loss": 0.2738,
      "mean_token_accuracy": 0.9019812941551208,
      "num_tokens": 550320983.0,
      "step": 15814
    },
    {
      "epoch": 2.93686165273909,
      "grad_norm": 1.5024039102560287,
      "learning_rate": 1e-06,
      "loss": 0.2936,
      "mean_token_accuracy": 0.8938112854957581,
      "num_tokens": 550355956.0,
      "step": 15815
    },
    {
      "epoch": 2.9370473537604456,
      "grad_norm": 1.5579634080542255,
      "learning_rate": 1e-06,
      "loss": 0.3613,
      "mean_token_accuracy": 0.8767223954200745,
      "num_tokens": 550390704.0,
      "step": 15816
    },
    {
      "epoch": 2.9372330547818013,
      "grad_norm": 1.4560763745009149,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8835529088973999,
      "num_tokens": 550428562.0,
      "step": 15817
    },
    {
      "epoch": 2.937418755803157,
      "grad_norm": 1.5605707124831174,
      "learning_rate": 1e-06,
      "loss": 0.3342,
      "mean_token_accuracy": 0.8804070949554443,
      "num_tokens": 550463603.0,
      "step": 15818
    },
    {
      "epoch": 2.9376044568245128,
      "grad_norm": 1.4958995493627572,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8801803588867188,
      "num_tokens": 550497356.0,
      "step": 15819
    },
    {
      "epoch": 2.937790157845868,
      "grad_norm": 1.4724543541544994,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8937026262283325,
      "num_tokens": 550531462.0,
      "step": 15820
    },
    {
      "epoch": 2.937975858867224,
      "grad_norm": 1.5368748491699493,
      "learning_rate": 1e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8856902718544006,
      "num_tokens": 550566575.0,
      "step": 15821
    },
    {
      "epoch": 2.9381615598885795,
      "grad_norm": 1.5756520530451612,
      "learning_rate": 1e-06,
      "loss": 0.3285,
      "mean_token_accuracy": 0.885259747505188,
      "num_tokens": 550600242.0,
      "step": 15822
    },
    {
      "epoch": 2.938347260909935,
      "grad_norm": 1.6051713938207788,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8774057626724243,
      "num_tokens": 550637031.0,
      "step": 15823
    },
    {
      "epoch": 2.9385329619312905,
      "grad_norm": 1.418799465263913,
      "learning_rate": 1e-06,
      "loss": 0.3069,
      "mean_token_accuracy": 0.8895716667175293,
      "num_tokens": 550676350.0,
      "step": 15824
    },
    {
      "epoch": 2.9387186629526463,
      "grad_norm": 1.5811734576209862,
      "learning_rate": 1e-06,
      "loss": 0.3274,
      "mean_token_accuracy": 0.8882684707641602,
      "num_tokens": 550707631.0,
      "step": 15825
    },
    {
      "epoch": 2.938904363974002,
      "grad_norm": 1.759804058179991,
      "learning_rate": 1e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.8760528564453125,
      "num_tokens": 550738118.0,
      "step": 15826
    },
    {
      "epoch": 2.9390900649953577,
      "grad_norm": 1.558179952735411,
      "learning_rate": 1e-06,
      "loss": 0.3673,
      "mean_token_accuracy": 0.872492790222168,
      "num_tokens": 550775862.0,
      "step": 15827
    },
    {
      "epoch": 2.939275766016713,
      "grad_norm": 1.4429731336165417,
      "learning_rate": 1e-06,
      "loss": 0.2884,
      "mean_token_accuracy": 0.8987947106361389,
      "num_tokens": 550811763.0,
      "step": 15828
    },
    {
      "epoch": 2.9394614670380688,
      "grad_norm": 1.596204106672449,
      "learning_rate": 1e-06,
      "loss": 0.2855,
      "mean_token_accuracy": 0.897566556930542,
      "num_tokens": 550841496.0,
      "step": 15829
    },
    {
      "epoch": 2.939647168059424,
      "grad_norm": 1.4716124186420598,
      "learning_rate": 1e-06,
      "loss": 0.2885,
      "mean_token_accuracy": 0.8969557285308838,
      "num_tokens": 550876122.0,
      "step": 15830
    },
    {
      "epoch": 2.93983286908078,
      "grad_norm": 1.378948629642639,
      "learning_rate": 1e-06,
      "loss": 0.2998,
      "mean_token_accuracy": 0.8951363563537598,
      "num_tokens": 550915916.0,
      "step": 15831
    },
    {
      "epoch": 2.9400185701021355,
      "grad_norm": 1.6981484146576415,
      "learning_rate": 1e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8789747953414917,
      "num_tokens": 550944413.0,
      "step": 15832
    },
    {
      "epoch": 2.9402042711234913,
      "grad_norm": 1.5556853635014898,
      "learning_rate": 1e-06,
      "loss": 0.3174,
      "mean_token_accuracy": 0.8899298906326294,
      "num_tokens": 550976426.0,
      "step": 15833
    },
    {
      "epoch": 2.940389972144847,
      "grad_norm": 1.5286847640790875,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8805942535400391,
      "num_tokens": 551014975.0,
      "step": 15834
    },
    {
      "epoch": 2.9405756731662023,
      "grad_norm": 1.6189472215180831,
      "learning_rate": 1e-06,
      "loss": 0.3132,
      "mean_token_accuracy": 0.8924537897109985,
      "num_tokens": 551046045.0,
      "step": 15835
    },
    {
      "epoch": 2.940761374187558,
      "grad_norm": 1.4834529809076495,
      "learning_rate": 1e-06,
      "loss": 0.2781,
      "mean_token_accuracy": 0.9008561968803406,
      "num_tokens": 551078933.0,
      "step": 15836
    },
    {
      "epoch": 2.9409470752089137,
      "grad_norm": 1.6135663676823786,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8945932388305664,
      "num_tokens": 551105595.0,
      "step": 15837
    },
    {
      "epoch": 2.941132776230269,
      "grad_norm": 1.7283572240151874,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8858720064163208,
      "num_tokens": 551138119.0,
      "step": 15838
    },
    {
      "epoch": 2.9413184772516248,
      "grad_norm": 1.6169839506644474,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8793215751647949,
      "num_tokens": 551177251.0,
      "step": 15839
    },
    {
      "epoch": 2.9415041782729805,
      "grad_norm": 1.650677646899924,
      "learning_rate": 1e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8695314526557922,
      "num_tokens": 551211717.0,
      "step": 15840
    },
    {
      "epoch": 2.9416898792943362,
      "grad_norm": 1.5052302429630617,
      "learning_rate": 1e-06,
      "loss": 0.3141,
      "mean_token_accuracy": 0.8887531161308289,
      "num_tokens": 551251622.0,
      "step": 15841
    },
    {
      "epoch": 2.941875580315692,
      "grad_norm": 1.5050811485653792,
      "learning_rate": 1e-06,
      "loss": 0.3237,
      "mean_token_accuracy": 0.8868027925491333,
      "num_tokens": 551287754.0,
      "step": 15842
    },
    {
      "epoch": 2.9420612813370473,
      "grad_norm": 1.617082966345992,
      "learning_rate": 1e-06,
      "loss": 0.2992,
      "mean_token_accuracy": 0.8942471742630005,
      "num_tokens": 551319053.0,
      "step": 15843
    },
    {
      "epoch": 2.942246982358403,
      "grad_norm": 1.529173802488696,
      "learning_rate": 1e-06,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8777875900268555,
      "num_tokens": 551358812.0,
      "step": 15844
    },
    {
      "epoch": 2.9424326833797587,
      "grad_norm": 1.6820532195180133,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8802601099014282,
      "num_tokens": 551394667.0,
      "step": 15845
    },
    {
      "epoch": 2.942618384401114,
      "grad_norm": 1.5357654571527513,
      "learning_rate": 1e-06,
      "loss": 0.2869,
      "mean_token_accuracy": 0.8990566730499268,
      "num_tokens": 551425269.0,
      "step": 15846
    },
    {
      "epoch": 2.9428040854224697,
      "grad_norm": 1.4912080424770335,
      "learning_rate": 1e-06,
      "loss": 0.3511,
      "mean_token_accuracy": 0.878666877746582,
      "num_tokens": 551467154.0,
      "step": 15847
    },
    {
      "epoch": 2.9429897864438255,
      "grad_norm": 1.6745135223048744,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8819707632064819,
      "num_tokens": 551497199.0,
      "step": 15848
    },
    {
      "epoch": 2.943175487465181,
      "grad_norm": 1.3995271512892997,
      "learning_rate": 1e-06,
      "loss": 0.2921,
      "mean_token_accuracy": 0.8979239463806152,
      "num_tokens": 551537224.0,
      "step": 15849
    },
    {
      "epoch": 2.943361188486537,
      "grad_norm": 1.5504931052883182,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8912777900695801,
      "num_tokens": 551568347.0,
      "step": 15850
    },
    {
      "epoch": 2.9435468895078922,
      "grad_norm": 1.498999691712853,
      "learning_rate": 1e-06,
      "loss": 0.3111,
      "mean_token_accuracy": 0.8944005966186523,
      "num_tokens": 551605706.0,
      "step": 15851
    },
    {
      "epoch": 2.943732590529248,
      "grad_norm": 1.4294088082903242,
      "learning_rate": 1e-06,
      "loss": 0.2766,
      "mean_token_accuracy": 0.9005728363990784,
      "num_tokens": 551639236.0,
      "step": 15852
    },
    {
      "epoch": 2.9439182915506037,
      "grad_norm": 1.5564556345055083,
      "learning_rate": 1e-06,
      "loss": 0.2856,
      "mean_token_accuracy": 0.8974973559379578,
      "num_tokens": 551668745.0,
      "step": 15853
    },
    {
      "epoch": 2.944103992571959,
      "grad_norm": 1.408699740461105,
      "learning_rate": 1e-06,
      "loss": 0.3227,
      "mean_token_accuracy": 0.8871651887893677,
      "num_tokens": 551707604.0,
      "step": 15854
    },
    {
      "epoch": 2.9442896935933147,
      "grad_norm": 1.5890611537575456,
      "learning_rate": 1e-06,
      "loss": 0.3054,
      "mean_token_accuracy": 0.8901457190513611,
      "num_tokens": 551741456.0,
      "step": 15855
    },
    {
      "epoch": 2.9444753946146704,
      "grad_norm": 1.5440769754934023,
      "learning_rate": 1e-06,
      "loss": 0.3245,
      "mean_token_accuracy": 0.8852578401565552,
      "num_tokens": 551779706.0,
      "step": 15856
    },
    {
      "epoch": 2.944661095636026,
      "grad_norm": 1.526162608145649,
      "learning_rate": 1e-06,
      "loss": 0.3025,
      "mean_token_accuracy": 0.8943735361099243,
      "num_tokens": 551813870.0,
      "step": 15857
    },
    {
      "epoch": 2.9448467966573815,
      "grad_norm": 1.5226232695847328,
      "learning_rate": 1e-06,
      "loss": 0.3422,
      "mean_token_accuracy": 0.87954181432724,
      "num_tokens": 551850581.0,
      "step": 15858
    },
    {
      "epoch": 2.945032497678737,
      "grad_norm": 1.5092331710814006,
      "learning_rate": 1e-06,
      "loss": 0.3055,
      "mean_token_accuracy": 0.8920717239379883,
      "num_tokens": 551885448.0,
      "step": 15859
    },
    {
      "epoch": 2.945218198700093,
      "grad_norm": 1.5292495591597914,
      "learning_rate": 1e-06,
      "loss": 0.3104,
      "mean_token_accuracy": 0.8908030986785889,
      "num_tokens": 551918176.0,
      "step": 15860
    },
    {
      "epoch": 2.9454038997214482,
      "grad_norm": 1.6371680832240083,
      "learning_rate": 1e-06,
      "loss": 0.3007,
      "mean_token_accuracy": 0.8951514959335327,
      "num_tokens": 551948965.0,
      "step": 15861
    },
    {
      "epoch": 2.945589600742804,
      "grad_norm": 1.628966121564375,
      "learning_rate": 1e-06,
      "loss": 0.2967,
      "mean_token_accuracy": 0.8947296142578125,
      "num_tokens": 551979287.0,
      "step": 15862
    },
    {
      "epoch": 2.9457753017641597,
      "grad_norm": 1.5882988576338037,
      "learning_rate": 1e-06,
      "loss": 0.3261,
      "mean_token_accuracy": 0.8875234127044678,
      "num_tokens": 552012782.0,
      "step": 15863
    },
    {
      "epoch": 2.9459610027855154,
      "grad_norm": 1.7354682776516497,
      "learning_rate": 1e-06,
      "loss": 0.3416,
      "mean_token_accuracy": 0.8795562386512756,
      "num_tokens": 552042247.0,
      "step": 15864
    },
    {
      "epoch": 2.946146703806871,
      "grad_norm": 1.6135929504924638,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8802807927131653,
      "num_tokens": 552077668.0,
      "step": 15865
    },
    {
      "epoch": 2.9463324048282264,
      "grad_norm": 1.5532857169705585,
      "learning_rate": 1e-06,
      "loss": 0.3295,
      "mean_token_accuracy": 0.8846298456192017,
      "num_tokens": 552115848.0,
      "step": 15866
    },
    {
      "epoch": 2.946518105849582,
      "grad_norm": 1.4924140980745888,
      "learning_rate": 1e-06,
      "loss": 0.2838,
      "mean_token_accuracy": 0.8997024297714233,
      "num_tokens": 552151632.0,
      "step": 15867
    },
    {
      "epoch": 2.946703806870938,
      "grad_norm": 1.5457880014022551,
      "learning_rate": 1e-06,
      "loss": 0.3148,
      "mean_token_accuracy": 0.8890058994293213,
      "num_tokens": 552183547.0,
      "step": 15868
    },
    {
      "epoch": 2.946889507892293,
      "grad_norm": 1.6891773147880114,
      "learning_rate": 1e-06,
      "loss": 0.29,
      "mean_token_accuracy": 0.9022570848464966,
      "num_tokens": 552211223.0,
      "step": 15869
    },
    {
      "epoch": 2.947075208913649,
      "grad_norm": 1.7642565880799908,
      "learning_rate": 1e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8717957735061646,
      "num_tokens": 552243168.0,
      "step": 15870
    },
    {
      "epoch": 2.9472609099350047,
      "grad_norm": 1.5924908428815714,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8907018899917603,
      "num_tokens": 552275599.0,
      "step": 15871
    },
    {
      "epoch": 2.9474466109563604,
      "grad_norm": 1.7111390928296804,
      "learning_rate": 1e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8780449032783508,
      "num_tokens": 552307263.0,
      "step": 15872
    },
    {
      "epoch": 2.947632311977716,
      "grad_norm": 1.4787883895729959,
      "learning_rate": 1e-06,
      "loss": 0.2959,
      "mean_token_accuracy": 0.8955265879631042,
      "num_tokens": 552342701.0,
      "step": 15873
    },
    {
      "epoch": 2.9478180129990714,
      "grad_norm": 1.5093460403161527,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8808666467666626,
      "num_tokens": 552380582.0,
      "step": 15874
    },
    {
      "epoch": 2.948003714020427,
      "grad_norm": 1.4642587599643757,
      "learning_rate": 1e-06,
      "loss": 0.3058,
      "mean_token_accuracy": 0.8921075463294983,
      "num_tokens": 552415019.0,
      "step": 15875
    },
    {
      "epoch": 2.948189415041783,
      "grad_norm": 1.551129994561578,
      "learning_rate": 1e-06,
      "loss": 0.32,
      "mean_token_accuracy": 0.8870171904563904,
      "num_tokens": 552448254.0,
      "step": 15876
    },
    {
      "epoch": 2.948375116063138,
      "grad_norm": 1.602424877180297,
      "learning_rate": 1e-06,
      "loss": 0.3636,
      "mean_token_accuracy": 0.8731886148452759,
      "num_tokens": 552484773.0,
      "step": 15877
    },
    {
      "epoch": 2.948560817084494,
      "grad_norm": 1.5780201479366773,
      "learning_rate": 1e-06,
      "loss": 0.2835,
      "mean_token_accuracy": 0.8980716466903687,
      "num_tokens": 552521572.0,
      "step": 15878
    },
    {
      "epoch": 2.9487465181058496,
      "grad_norm": 1.4806577923305118,
      "learning_rate": 1e-06,
      "loss": 0.2904,
      "mean_token_accuracy": 0.8981112837791443,
      "num_tokens": 552557936.0,
      "step": 15879
    },
    {
      "epoch": 2.9489322191272054,
      "grad_norm": 1.505063681607296,
      "learning_rate": 1e-06,
      "loss": 0.3328,
      "mean_token_accuracy": 0.883651852607727,
      "num_tokens": 552594159.0,
      "step": 15880
    },
    {
      "epoch": 2.9491179201485607,
      "grad_norm": 1.5901767241145908,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8795104622840881,
      "num_tokens": 552630132.0,
      "step": 15881
    },
    {
      "epoch": 2.9493036211699164,
      "grad_norm": 1.6133294270814371,
      "learning_rate": 1e-06,
      "loss": 0.3273,
      "mean_token_accuracy": 0.8847264051437378,
      "num_tokens": 552663030.0,
      "step": 15882
    },
    {
      "epoch": 2.949489322191272,
      "grad_norm": 1.7553424262376551,
      "learning_rate": 1e-06,
      "loss": 0.3607,
      "mean_token_accuracy": 0.8773831725120544,
      "num_tokens": 552692899.0,
      "step": 15883
    },
    {
      "epoch": 2.9496750232126274,
      "grad_norm": 1.544422708455152,
      "learning_rate": 1e-06,
      "loss": 0.3424,
      "mean_token_accuracy": 0.8793289065361023,
      "num_tokens": 552730325.0,
      "step": 15884
    },
    {
      "epoch": 2.949860724233983,
      "grad_norm": 1.4995926320461974,
      "learning_rate": 1e-06,
      "loss": 0.3106,
      "mean_token_accuracy": 0.8898871541023254,
      "num_tokens": 552765426.0,
      "step": 15885
    },
    {
      "epoch": 2.950046425255339,
      "grad_norm": 1.4595621023395742,
      "learning_rate": 1e-06,
      "loss": 0.3312,
      "mean_token_accuracy": 0.8838992118835449,
      "num_tokens": 552804281.0,
      "step": 15886
    },
    {
      "epoch": 2.9502321262766946,
      "grad_norm": 1.5550736047168874,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8918333053588867,
      "num_tokens": 552839934.0,
      "step": 15887
    },
    {
      "epoch": 2.9504178272980504,
      "grad_norm": 1.546129322726798,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8897479176521301,
      "num_tokens": 552871836.0,
      "step": 15888
    },
    {
      "epoch": 2.9506035283194056,
      "grad_norm": 1.6115852296042557,
      "learning_rate": 1e-06,
      "loss": 0.3211,
      "mean_token_accuracy": 0.8864147663116455,
      "num_tokens": 552904795.0,
      "step": 15889
    },
    {
      "epoch": 2.9507892293407614,
      "grad_norm": 1.4920622404980826,
      "learning_rate": 1e-06,
      "loss": 0.3094,
      "mean_token_accuracy": 0.889399528503418,
      "num_tokens": 552941901.0,
      "step": 15890
    },
    {
      "epoch": 2.950974930362117,
      "grad_norm": 1.624121618884503,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8863376975059509,
      "num_tokens": 552977018.0,
      "step": 15891
    },
    {
      "epoch": 2.9511606313834724,
      "grad_norm": 1.5767233397581482,
      "learning_rate": 1e-06,
      "loss": 0.3124,
      "mean_token_accuracy": 0.8903733491897583,
      "num_tokens": 553012660.0,
      "step": 15892
    },
    {
      "epoch": 2.951346332404828,
      "grad_norm": 1.6616470836749582,
      "learning_rate": 1e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8749470710754395,
      "num_tokens": 553047231.0,
      "step": 15893
    },
    {
      "epoch": 2.951532033426184,
      "grad_norm": 1.619493618509401,
      "learning_rate": 1e-06,
      "loss": 0.3449,
      "mean_token_accuracy": 0.878121018409729,
      "num_tokens": 553080748.0,
      "step": 15894
    },
    {
      "epoch": 2.9517177344475396,
      "grad_norm": 1.5554993552903904,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.8891316056251526,
      "num_tokens": 553115922.0,
      "step": 15895
    },
    {
      "epoch": 2.9519034354688953,
      "grad_norm": 1.8551608029786193,
      "learning_rate": 1e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8802405595779419,
      "num_tokens": 553142441.0,
      "step": 15896
    },
    {
      "epoch": 2.9520891364902506,
      "grad_norm": 1.6406984873438373,
      "learning_rate": 1e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8749838471412659,
      "num_tokens": 553174116.0,
      "step": 15897
    },
    {
      "epoch": 2.9522748375116064,
      "grad_norm": 1.4473407528996427,
      "learning_rate": 1e-06,
      "loss": 0.311,
      "mean_token_accuracy": 0.8894835710525513,
      "num_tokens": 553212965.0,
      "step": 15898
    },
    {
      "epoch": 2.952460538532962,
      "grad_norm": 1.8808556362148916,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8771909475326538,
      "num_tokens": 553238881.0,
      "step": 15899
    },
    {
      "epoch": 2.9526462395543174,
      "grad_norm": 1.606615801746153,
      "learning_rate": 1e-06,
      "loss": 0.3046,
      "mean_token_accuracy": 0.8930813670158386,
      "num_tokens": 553270355.0,
      "step": 15900
    },
    {
      "epoch": 2.952831940575673,
      "grad_norm": 1.5947635210870423,
      "learning_rate": 1e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8901596665382385,
      "num_tokens": 553302346.0,
      "step": 15901
    },
    {
      "epoch": 2.953017641597029,
      "grad_norm": 1.6282625188024555,
      "learning_rate": 1e-06,
      "loss": 0.3539,
      "mean_token_accuracy": 0.8777735233306885,
      "num_tokens": 553337920.0,
      "step": 15902
    },
    {
      "epoch": 2.9532033426183846,
      "grad_norm": 1.5195101204172339,
      "learning_rate": 1e-06,
      "loss": 0.3233,
      "mean_token_accuracy": 0.885073721408844,
      "num_tokens": 553373571.0,
      "step": 15903
    },
    {
      "epoch": 2.95338904363974,
      "grad_norm": 1.5811320743599573,
      "learning_rate": 1e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8713517189025879,
      "num_tokens": 553413807.0,
      "step": 15904
    },
    {
      "epoch": 2.9535747446610956,
      "grad_norm": 1.498269343557169,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8896349668502808,
      "num_tokens": 553446438.0,
      "step": 15905
    },
    {
      "epoch": 2.9537604456824513,
      "grad_norm": 1.6316363610696,
      "learning_rate": 1e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8758509159088135,
      "num_tokens": 553480666.0,
      "step": 15906
    },
    {
      "epoch": 2.9539461467038066,
      "grad_norm": 1.477098000092673,
      "learning_rate": 1e-06,
      "loss": 0.2894,
      "mean_token_accuracy": 0.8962908983230591,
      "num_tokens": 553515971.0,
      "step": 15907
    },
    {
      "epoch": 2.9541318477251624,
      "grad_norm": 1.575599002254709,
      "learning_rate": 1e-06,
      "loss": 0.2923,
      "mean_token_accuracy": 0.8959673047065735,
      "num_tokens": 553547511.0,
      "step": 15908
    },
    {
      "epoch": 2.954317548746518,
      "grad_norm": 1.4434216621241873,
      "learning_rate": 1e-06,
      "loss": 0.2745,
      "mean_token_accuracy": 0.9013705253601074,
      "num_tokens": 553584525.0,
      "step": 15909
    },
    {
      "epoch": 2.954503249767874,
      "grad_norm": 1.6295547446611902,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8841595649719238,
      "num_tokens": 553619165.0,
      "step": 15910
    },
    {
      "epoch": 2.9546889507892296,
      "grad_norm": 1.417420503443795,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8865939378738403,
      "num_tokens": 553661041.0,
      "step": 15911
    },
    {
      "epoch": 2.954874651810585,
      "grad_norm": 1.6524292534849352,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8859305381774902,
      "num_tokens": 553698436.0,
      "step": 15912
    },
    {
      "epoch": 2.9550603528319406,
      "grad_norm": 1.5719061570121995,
      "learning_rate": 1e-06,
      "loss": 0.3435,
      "mean_token_accuracy": 0.8807916641235352,
      "num_tokens": 553732678.0,
      "step": 15913
    },
    {
      "epoch": 2.9552460538532963,
      "grad_norm": 1.5770368823706207,
      "learning_rate": 1e-06,
      "loss": 0.3493,
      "mean_token_accuracy": 0.8776538372039795,
      "num_tokens": 553767925.0,
      "step": 15914
    },
    {
      "epoch": 2.9554317548746516,
      "grad_norm": 1.411650265349087,
      "learning_rate": 1e-06,
      "loss": 0.2854,
      "mean_token_accuracy": 0.900348424911499,
      "num_tokens": 553806894.0,
      "step": 15915
    },
    {
      "epoch": 2.9556174558960073,
      "grad_norm": 1.5130902217894953,
      "learning_rate": 1e-06,
      "loss": 0.3301,
      "mean_token_accuracy": 0.8872495293617249,
      "num_tokens": 553844038.0,
      "step": 15916
    },
    {
      "epoch": 2.955803156917363,
      "grad_norm": 1.585593626714333,
      "learning_rate": 1e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.8723297715187073,
      "num_tokens": 553880190.0,
      "step": 15917
    },
    {
      "epoch": 2.955988857938719,
      "grad_norm": 1.3394179576992076,
      "learning_rate": 1e-06,
      "loss": 0.2865,
      "mean_token_accuracy": 0.8992326259613037,
      "num_tokens": 553922542.0,
      "step": 15918
    },
    {
      "epoch": 2.9561745589600745,
      "grad_norm": 1.4334976248732896,
      "learning_rate": 1e-06,
      "loss": 0.2581,
      "mean_token_accuracy": 0.9070445895195007,
      "num_tokens": 553959024.0,
      "step": 15919
    },
    {
      "epoch": 2.95636025998143,
      "grad_norm": 1.5854358280263792,
      "learning_rate": 1e-06,
      "loss": 0.273,
      "mean_token_accuracy": 0.901258647441864,
      "num_tokens": 553987079.0,
      "step": 15920
    },
    {
      "epoch": 2.9565459610027855,
      "grad_norm": 1.7063743046055093,
      "learning_rate": 1e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.8787618279457092,
      "num_tokens": 554017703.0,
      "step": 15921
    },
    {
      "epoch": 2.9567316620241413,
      "grad_norm": 1.6059100671540656,
      "learning_rate": 1e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.882500410079956,
      "num_tokens": 554050791.0,
      "step": 15922
    },
    {
      "epoch": 2.9569173630454966,
      "grad_norm": 1.550522556858043,
      "learning_rate": 1e-06,
      "loss": 0.2918,
      "mean_token_accuracy": 0.8951418399810791,
      "num_tokens": 554084768.0,
      "step": 15923
    },
    {
      "epoch": 2.9571030640668523,
      "grad_norm": 1.5641384800748914,
      "learning_rate": 1e-06,
      "loss": 0.3489,
      "mean_token_accuracy": 0.8802088499069214,
      "num_tokens": 554125170.0,
      "step": 15924
    },
    {
      "epoch": 2.957288765088208,
      "grad_norm": 1.3532827784494468,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8906093835830688,
      "num_tokens": 554169537.0,
      "step": 15925
    },
    {
      "epoch": 2.9574744661095638,
      "grad_norm": 1.6214497580576985,
      "learning_rate": 1e-06,
      "loss": 0.3777,
      "mean_token_accuracy": 0.8681171536445618,
      "num_tokens": 554208488.0,
      "step": 15926
    },
    {
      "epoch": 2.957660167130919,
      "grad_norm": 1.5970643176784198,
      "learning_rate": 1e-06,
      "loss": 0.2636,
      "mean_token_accuracy": 0.9030910730361938,
      "num_tokens": 554237344.0,
      "step": 15927
    },
    {
      "epoch": 2.957845868152275,
      "grad_norm": 1.3855408886684275,
      "learning_rate": 1e-06,
      "loss": 0.2762,
      "mean_token_accuracy": 0.900112509727478,
      "num_tokens": 554277415.0,
      "step": 15928
    },
    {
      "epoch": 2.9580315691736305,
      "grad_norm": 1.5088848421198038,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.8831719160079956,
      "num_tokens": 554312134.0,
      "step": 15929
    },
    {
      "epoch": 2.958217270194986,
      "grad_norm": 1.4217374426597411,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.8973879814147949,
      "num_tokens": 554352279.0,
      "step": 15930
    },
    {
      "epoch": 2.9584029712163415,
      "grad_norm": 1.561671599446215,
      "learning_rate": 1e-06,
      "loss": 0.2853,
      "mean_token_accuracy": 0.8984411358833313,
      "num_tokens": 554385429.0,
      "step": 15931
    },
    {
      "epoch": 2.9585886722376973,
      "grad_norm": 1.5842290258213252,
      "learning_rate": 1e-06,
      "loss": 0.3579,
      "mean_token_accuracy": 0.8741626143455505,
      "num_tokens": 554420065.0,
      "step": 15932
    },
    {
      "epoch": 2.958774373259053,
      "grad_norm": 1.5116867775212204,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8899650573730469,
      "num_tokens": 554458002.0,
      "step": 15933
    },
    {
      "epoch": 2.9589600742804087,
      "grad_norm": 1.5814551125790615,
      "learning_rate": 1e-06,
      "loss": 0.3413,
      "mean_token_accuracy": 0.8830270767211914,
      "num_tokens": 554495159.0,
      "step": 15934
    },
    {
      "epoch": 2.959145775301764,
      "grad_norm": 1.5609790455698387,
      "learning_rate": 1e-06,
      "loss": 0.3336,
      "mean_token_accuracy": 0.8826847076416016,
      "num_tokens": 554529508.0,
      "step": 15935
    },
    {
      "epoch": 2.9593314763231198,
      "grad_norm": 1.4242173569127012,
      "learning_rate": 1e-06,
      "loss": 0.2634,
      "mean_token_accuracy": 0.905654788017273,
      "num_tokens": 554568782.0,
      "step": 15936
    },
    {
      "epoch": 2.9595171773444755,
      "grad_norm": 1.4989108451991002,
      "learning_rate": 1e-06,
      "loss": 0.3356,
      "mean_token_accuracy": 0.8830897808074951,
      "num_tokens": 554606521.0,
      "step": 15937
    },
    {
      "epoch": 2.959702878365831,
      "grad_norm": 1.6485263586561483,
      "learning_rate": 1e-06,
      "loss": 0.3533,
      "mean_token_accuracy": 0.879544198513031,
      "num_tokens": 554639727.0,
      "step": 15938
    },
    {
      "epoch": 2.9598885793871865,
      "grad_norm": 1.6696875414042986,
      "learning_rate": 1e-06,
      "loss": 0.3276,
      "mean_token_accuracy": 0.8851941227912903,
      "num_tokens": 554672650.0,
      "step": 15939
    },
    {
      "epoch": 2.9600742804085423,
      "grad_norm": 1.5152723995988417,
      "learning_rate": 1e-06,
      "loss": 0.3243,
      "mean_token_accuracy": 0.8879381418228149,
      "num_tokens": 554710844.0,
      "step": 15940
    },
    {
      "epoch": 2.960259981429898,
      "grad_norm": 1.5399688562410976,
      "learning_rate": 1e-06,
      "loss": 0.2966,
      "mean_token_accuracy": 0.8961359858512878,
      "num_tokens": 554743532.0,
      "step": 15941
    },
    {
      "epoch": 2.9604456824512537,
      "grad_norm": 1.5640102524986115,
      "learning_rate": 1e-06,
      "loss": 0.3484,
      "mean_token_accuracy": 0.8790094256401062,
      "num_tokens": 554780578.0,
      "step": 15942
    },
    {
      "epoch": 2.960631383472609,
      "grad_norm": 1.5991544433812666,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8813395500183105,
      "num_tokens": 554814444.0,
      "step": 15943
    },
    {
      "epoch": 2.9608170844939647,
      "grad_norm": 1.5681814932887357,
      "learning_rate": 1e-06,
      "loss": 0.288,
      "mean_token_accuracy": 0.8981000781059265,
      "num_tokens": 554846465.0,
      "step": 15944
    },
    {
      "epoch": 2.9610027855153205,
      "grad_norm": 1.574253506447094,
      "learning_rate": 1e-06,
      "loss": 0.3253,
      "mean_token_accuracy": 0.8849895596504211,
      "num_tokens": 554883003.0,
      "step": 15945
    },
    {
      "epoch": 2.9611884865366758,
      "grad_norm": 1.5861494889109886,
      "learning_rate": 1e-06,
      "loss": 0.3322,
      "mean_token_accuracy": 0.8815052509307861,
      "num_tokens": 554913650.0,
      "step": 15946
    },
    {
      "epoch": 2.9613741875580315,
      "grad_norm": 1.642835202325384,
      "learning_rate": 1e-06,
      "loss": 0.3391,
      "mean_token_accuracy": 0.8819627165794373,
      "num_tokens": 554945153.0,
      "step": 15947
    },
    {
      "epoch": 2.9615598885793872,
      "grad_norm": 1.397974958098451,
      "learning_rate": 1e-06,
      "loss": 0.3278,
      "mean_token_accuracy": 0.8857219219207764,
      "num_tokens": 554984748.0,
      "step": 15948
    },
    {
      "epoch": 2.961745589600743,
      "grad_norm": 1.620681028457076,
      "learning_rate": 1e-06,
      "loss": 0.3079,
      "mean_token_accuracy": 0.8881781101226807,
      "num_tokens": 555016861.0,
      "step": 15949
    },
    {
      "epoch": 2.9619312906220983,
      "grad_norm": 1.671502428589109,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8883037567138672,
      "num_tokens": 555048925.0,
      "step": 15950
    },
    {
      "epoch": 2.962116991643454,
      "grad_norm": 1.5144653088911437,
      "learning_rate": 1e-06,
      "loss": 0.306,
      "mean_token_accuracy": 0.8933230042457581,
      "num_tokens": 555088258.0,
      "step": 15951
    },
    {
      "epoch": 2.9623026926648097,
      "grad_norm": 1.486668266895201,
      "learning_rate": 1e-06,
      "loss": 0.315,
      "mean_token_accuracy": 0.8909623622894287,
      "num_tokens": 555125796.0,
      "step": 15952
    },
    {
      "epoch": 2.962488393686165,
      "grad_norm": 1.5332933081715276,
      "learning_rate": 1e-06,
      "loss": 0.2814,
      "mean_token_accuracy": 0.8999266028404236,
      "num_tokens": 555157666.0,
      "step": 15953
    },
    {
      "epoch": 2.9626740947075207,
      "grad_norm": 1.6634433903723862,
      "learning_rate": 1e-06,
      "loss": 0.3461,
      "mean_token_accuracy": 0.8764592409133911,
      "num_tokens": 555191774.0,
      "step": 15954
    },
    {
      "epoch": 2.9628597957288765,
      "grad_norm": 1.6680501091427937,
      "learning_rate": 1e-06,
      "loss": 0.3326,
      "mean_token_accuracy": 0.8791476488113403,
      "num_tokens": 555225099.0,
      "step": 15955
    },
    {
      "epoch": 2.963045496750232,
      "grad_norm": 1.5521549746557743,
      "learning_rate": 1e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8640501499176025,
      "num_tokens": 555264452.0,
      "step": 15956
    },
    {
      "epoch": 2.963231197771588,
      "grad_norm": 1.6703451658754804,
      "learning_rate": 1e-06,
      "loss": 0.3263,
      "mean_token_accuracy": 0.8846811652183533,
      "num_tokens": 555296667.0,
      "step": 15957
    },
    {
      "epoch": 2.9634168987929432,
      "grad_norm": 1.5998365460024433,
      "learning_rate": 1e-06,
      "loss": 0.3122,
      "mean_token_accuracy": 0.8857674598693848,
      "num_tokens": 555329045.0,
      "step": 15958
    },
    {
      "epoch": 2.963602599814299,
      "grad_norm": 1.5278616037177746,
      "learning_rate": 1e-06,
      "loss": 0.2503,
      "mean_token_accuracy": 0.9102216958999634,
      "num_tokens": 555359183.0,
      "step": 15959
    },
    {
      "epoch": 2.9637883008356547,
      "grad_norm": 1.5637829964131262,
      "learning_rate": 1e-06,
      "loss": 0.3441,
      "mean_token_accuracy": 0.8872880935668945,
      "num_tokens": 555395550.0,
      "step": 15960
    },
    {
      "epoch": 2.96397400185701,
      "grad_norm": 1.5246320159257867,
      "learning_rate": 1e-06,
      "loss": 0.2841,
      "mean_token_accuracy": 0.9035601615905762,
      "num_tokens": 555429852.0,
      "step": 15961
    },
    {
      "epoch": 2.9641597028783657,
      "grad_norm": 1.5225036549433604,
      "learning_rate": 1e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8877763152122498,
      "num_tokens": 555465752.0,
      "step": 15962
    },
    {
      "epoch": 2.9643454038997215,
      "grad_norm": 1.6423077360288423,
      "learning_rate": 1e-06,
      "loss": 0.3175,
      "mean_token_accuracy": 0.8876492977142334,
      "num_tokens": 555493921.0,
      "step": 15963
    },
    {
      "epoch": 2.964531104921077,
      "grad_norm": 1.7476301974124628,
      "learning_rate": 1e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8646928071975708,
      "num_tokens": 555528124.0,
      "step": 15964
    },
    {
      "epoch": 2.964716805942433,
      "grad_norm": 1.6910525189451424,
      "learning_rate": 1e-06,
      "loss": 0.2931,
      "mean_token_accuracy": 0.8964368104934692,
      "num_tokens": 555558683.0,
      "step": 15965
    },
    {
      "epoch": 2.964902506963788,
      "grad_norm": 1.687636373408029,
      "learning_rate": 1e-06,
      "loss": 0.3203,
      "mean_token_accuracy": 0.8874996304512024,
      "num_tokens": 555587243.0,
      "step": 15966
    },
    {
      "epoch": 2.965088207985144,
      "grad_norm": 1.630086796293285,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.8844794034957886,
      "num_tokens": 555622490.0,
      "step": 15967
    },
    {
      "epoch": 2.9652739090064997,
      "grad_norm": 1.5299789880613854,
      "learning_rate": 1e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8820122480392456,
      "num_tokens": 555661430.0,
      "step": 15968
    },
    {
      "epoch": 2.965459610027855,
      "grad_norm": 1.5414562196247954,
      "learning_rate": 1e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8678960800170898,
      "num_tokens": 555700936.0,
      "step": 15969
    },
    {
      "epoch": 2.9656453110492107,
      "grad_norm": 1.3870356158925363,
      "learning_rate": 1e-06,
      "loss": 0.2851,
      "mean_token_accuracy": 0.8976386785507202,
      "num_tokens": 555740004.0,
      "step": 15970
    },
    {
      "epoch": 2.9658310120705664,
      "grad_norm": 1.5197601833132381,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8878453373908997,
      "num_tokens": 555776824.0,
      "step": 15971
    },
    {
      "epoch": 2.966016713091922,
      "grad_norm": 1.6405821986317122,
      "learning_rate": 1e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8741475939750671,
      "num_tokens": 555809013.0,
      "step": 15972
    },
    {
      "epoch": 2.966202414113278,
      "grad_norm": 1.602154698697163,
      "learning_rate": 1e-06,
      "loss": 0.322,
      "mean_token_accuracy": 0.885353147983551,
      "num_tokens": 555843195.0,
      "step": 15973
    },
    {
      "epoch": 2.966388115134633,
      "grad_norm": 1.6439978676230056,
      "learning_rate": 1e-06,
      "loss": 0.3298,
      "mean_token_accuracy": 0.8822714686393738,
      "num_tokens": 555873898.0,
      "step": 15974
    },
    {
      "epoch": 2.966573816155989,
      "grad_norm": 1.4816954521793009,
      "learning_rate": 1e-06,
      "loss": 0.2898,
      "mean_token_accuracy": 0.8935985565185547,
      "num_tokens": 555911953.0,
      "step": 15975
    },
    {
      "epoch": 2.966759517177344,
      "grad_norm": 1.6351836838092317,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8902081251144409,
      "num_tokens": 555943699.0,
      "step": 15976
    },
    {
      "epoch": 2.9669452181987,
      "grad_norm": 1.5570835098470843,
      "learning_rate": 1e-06,
      "loss": 0.3041,
      "mean_token_accuracy": 0.8931299448013306,
      "num_tokens": 555979372.0,
      "step": 15977
    },
    {
      "epoch": 2.9671309192200557,
      "grad_norm": 1.636272556250109,
      "learning_rate": 1e-06,
      "loss": 0.3438,
      "mean_token_accuracy": 0.8788525462150574,
      "num_tokens": 556013301.0,
      "step": 15978
    },
    {
      "epoch": 2.9673166202414114,
      "grad_norm": 1.4559590296468805,
      "learning_rate": 1e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.875219464302063,
      "num_tokens": 556055547.0,
      "step": 15979
    },
    {
      "epoch": 2.967502321262767,
      "grad_norm": 1.495103132314835,
      "learning_rate": 1e-06,
      "loss": 0.3368,
      "mean_token_accuracy": 0.8847837448120117,
      "num_tokens": 556095431.0,
      "step": 15980
    },
    {
      "epoch": 2.9676880222841224,
      "grad_norm": 1.661938642136909,
      "learning_rate": 1e-06,
      "loss": 0.382,
      "mean_token_accuracy": 0.8643885850906372,
      "num_tokens": 556128230.0,
      "step": 15981
    },
    {
      "epoch": 2.967873723305478,
      "grad_norm": 1.4069993793259088,
      "learning_rate": 1e-06,
      "loss": 0.2825,
      "mean_token_accuracy": 0.9009303450584412,
      "num_tokens": 556167509.0,
      "step": 15982
    },
    {
      "epoch": 2.968059424326834,
      "grad_norm": 1.6286508131302533,
      "learning_rate": 1e-06,
      "loss": 0.3258,
      "mean_token_accuracy": 0.8866909146308899,
      "num_tokens": 556201282.0,
      "step": 15983
    },
    {
      "epoch": 2.968245125348189,
      "grad_norm": 1.6336661980713143,
      "learning_rate": 1e-06,
      "loss": 0.2964,
      "mean_token_accuracy": 0.8959307074546814,
      "num_tokens": 556231267.0,
      "step": 15984
    },
    {
      "epoch": 2.968430826369545,
      "grad_norm": 1.5807340751665,
      "learning_rate": 1e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8787112236022949,
      "num_tokens": 556268707.0,
      "step": 15985
    },
    {
      "epoch": 2.9686165273909007,
      "grad_norm": 1.6611822328339847,
      "learning_rate": 1e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.866165041923523,
      "num_tokens": 556305155.0,
      "step": 15986
    },
    {
      "epoch": 2.9688022284122564,
      "grad_norm": 1.5720194625990245,
      "learning_rate": 1e-06,
      "loss": 0.3224,
      "mean_token_accuracy": 0.8852689266204834,
      "num_tokens": 556339888.0,
      "step": 15987
    },
    {
      "epoch": 2.968987929433612,
      "grad_norm": 1.6015485875681483,
      "learning_rate": 1e-06,
      "loss": 0.3366,
      "mean_token_accuracy": 0.882416844367981,
      "num_tokens": 556375047.0,
      "step": 15988
    },
    {
      "epoch": 2.9691736304549674,
      "grad_norm": 1.6095445107381354,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8835391998291016,
      "num_tokens": 556408450.0,
      "step": 15989
    },
    {
      "epoch": 2.969359331476323,
      "grad_norm": 1.6066800325525752,
      "learning_rate": 1e-06,
      "loss": 0.3663,
      "mean_token_accuracy": 0.873015820980072,
      "num_tokens": 556442579.0,
      "step": 15990
    },
    {
      "epoch": 2.969545032497679,
      "grad_norm": 1.6341799841617686,
      "learning_rate": 1e-06,
      "loss": 0.3282,
      "mean_token_accuracy": 0.8827453255653381,
      "num_tokens": 556475032.0,
      "step": 15991
    },
    {
      "epoch": 2.969730733519034,
      "grad_norm": 1.550333659633001,
      "learning_rate": 1e-06,
      "loss": 0.3075,
      "mean_token_accuracy": 0.8922936916351318,
      "num_tokens": 556506358.0,
      "step": 15992
    },
    {
      "epoch": 2.96991643454039,
      "grad_norm": 1.533078492750717,
      "learning_rate": 1e-06,
      "loss": 0.3103,
      "mean_token_accuracy": 0.8927234411239624,
      "num_tokens": 556539046.0,
      "step": 15993
    },
    {
      "epoch": 2.9701021355617456,
      "grad_norm": 1.67626381361356,
      "learning_rate": 1e-06,
      "loss": 0.3316,
      "mean_token_accuracy": 0.8839142918586731,
      "num_tokens": 556569453.0,
      "step": 15994
    },
    {
      "epoch": 2.9702878365831014,
      "grad_norm": 1.4356673606313053,
      "learning_rate": 1e-06,
      "loss": 0.3135,
      "mean_token_accuracy": 0.8890798687934875,
      "num_tokens": 556608935.0,
      "step": 15995
    },
    {
      "epoch": 2.970473537604457,
      "grad_norm": 1.4943683262929415,
      "learning_rate": 1e-06,
      "loss": 0.3208,
      "mean_token_accuracy": 0.8849667310714722,
      "num_tokens": 556648253.0,
      "step": 15996
    },
    {
      "epoch": 2.9706592386258124,
      "grad_norm": 1.7508238915714103,
      "learning_rate": 1e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8681752681732178,
      "num_tokens": 556680560.0,
      "step": 15997
    },
    {
      "epoch": 2.970844939647168,
      "grad_norm": 1.5465398354164002,
      "learning_rate": 1e-06,
      "loss": 0.2847,
      "mean_token_accuracy": 0.8972458839416504,
      "num_tokens": 556713695.0,
      "step": 15998
    },
    {
      "epoch": 2.9710306406685234,
      "grad_norm": 1.4844145172238097,
      "learning_rate": 1e-06,
      "loss": 0.2888,
      "mean_token_accuracy": 0.897504448890686,
      "num_tokens": 556749657.0,
      "step": 15999
    },
    {
      "epoch": 2.971216341689879,
      "grad_norm": 1.5724351545924171,
      "learning_rate": 1e-06,
      "loss": 0.3311,
      "mean_token_accuracy": 0.885229229927063,
      "num_tokens": 556787933.0,
      "step": 16000
    },
    {
      "epoch": 2.971402042711235,
      "grad_norm": 1.5861612499994122,
      "learning_rate": 1e-06,
      "loss": 0.3238,
      "mean_token_accuracy": 0.8848316669464111,
      "num_tokens": 556822505.0,
      "step": 16001
    },
    {
      "epoch": 2.9715877437325906,
      "grad_norm": 1.5712005461076688,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.8951613903045654,
      "num_tokens": 556851373.0,
      "step": 16002
    },
    {
      "epoch": 2.9717734447539463,
      "grad_norm": 1.4906206576254188,
      "learning_rate": 1e-06,
      "loss": 0.3269,
      "mean_token_accuracy": 0.887549877166748,
      "num_tokens": 556892690.0,
      "step": 16003
    },
    {
      "epoch": 2.9719591457753016,
      "grad_norm": 1.8054724500051063,
      "learning_rate": 1e-06,
      "loss": 0.3384,
      "mean_token_accuracy": 0.880592942237854,
      "num_tokens": 556917458.0,
      "step": 16004
    },
    {
      "epoch": 2.9721448467966574,
      "grad_norm": 1.4016601106371203,
      "learning_rate": 1e-06,
      "loss": 0.28,
      "mean_token_accuracy": 0.9000157117843628,
      "num_tokens": 556954503.0,
      "step": 16005
    },
    {
      "epoch": 2.972330547818013,
      "grad_norm": 1.4236181361226712,
      "learning_rate": 1e-06,
      "loss": 0.292,
      "mean_token_accuracy": 0.8950058221817017,
      "num_tokens": 556991850.0,
      "step": 16006
    },
    {
      "epoch": 2.9725162488393684,
      "grad_norm": 1.6015036536224085,
      "learning_rate": 1e-06,
      "loss": 0.3185,
      "mean_token_accuracy": 0.890389084815979,
      "num_tokens": 557023868.0,
      "step": 16007
    },
    {
      "epoch": 2.972701949860724,
      "grad_norm": 1.4626797899738826,
      "learning_rate": 1e-06,
      "loss": 0.3061,
      "mean_token_accuracy": 0.891484260559082,
      "num_tokens": 557061409.0,
      "step": 16008
    },
    {
      "epoch": 2.97288765088208,
      "grad_norm": 1.4767237191150742,
      "learning_rate": 1e-06,
      "loss": 0.3084,
      "mean_token_accuracy": 0.8899592161178589,
      "num_tokens": 557099092.0,
      "step": 16009
    },
    {
      "epoch": 2.9730733519034356,
      "grad_norm": 1.483055644199236,
      "learning_rate": 1e-06,
      "loss": 0.3037,
      "mean_token_accuracy": 0.8938982486724854,
      "num_tokens": 557136392.0,
      "step": 16010
    },
    {
      "epoch": 2.9732590529247913,
      "grad_norm": 1.5827599829229992,
      "learning_rate": 1e-06,
      "loss": 0.3225,
      "mean_token_accuracy": 0.8892499804496765,
      "num_tokens": 557169228.0,
      "step": 16011
    },
    {
      "epoch": 2.9734447539461466,
      "grad_norm": 1.5960827601760639,
      "learning_rate": 1e-06,
      "loss": 0.3047,
      "mean_token_accuracy": 0.8924427032470703,
      "num_tokens": 557206040.0,
      "step": 16012
    },
    {
      "epoch": 2.9736304549675023,
      "grad_norm": 1.530459967434464,
      "learning_rate": 1e-06,
      "loss": 0.3083,
      "mean_token_accuracy": 0.8915724754333496,
      "num_tokens": 557241579.0,
      "step": 16013
    },
    {
      "epoch": 2.973816155988858,
      "grad_norm": 1.7600923884734214,
      "learning_rate": 1e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8793643712997437,
      "num_tokens": 557269124.0,
      "step": 16014
    },
    {
      "epoch": 2.9740018570102134,
      "grad_norm": 1.4874107570347357,
      "learning_rate": 1e-06,
      "loss": 0.2737,
      "mean_token_accuracy": 0.8996421098709106,
      "num_tokens": 557301173.0,
      "step": 16015
    },
    {
      "epoch": 2.974187558031569,
      "grad_norm": 1.706476739817219,
      "learning_rate": 1e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.876248836517334,
      "num_tokens": 557333707.0,
      "step": 16016
    },
    {
      "epoch": 2.974373259052925,
      "grad_norm": 1.4552357134257414,
      "learning_rate": 1e-06,
      "loss": 0.298,
      "mean_token_accuracy": 0.895435094833374,
      "num_tokens": 557371170.0,
      "step": 16017
    },
    {
      "epoch": 2.9745589600742806,
      "grad_norm": 1.4891310374791824,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.889693021774292,
      "num_tokens": 557410846.0,
      "step": 16018
    },
    {
      "epoch": 2.9747446610956363,
      "grad_norm": 1.5864074940963122,
      "learning_rate": 1e-06,
      "loss": 0.3226,
      "mean_token_accuracy": 0.8859537839889526,
      "num_tokens": 557444634.0,
      "step": 16019
    },
    {
      "epoch": 2.9749303621169916,
      "grad_norm": 1.5414958258901597,
      "learning_rate": 1e-06,
      "loss": 0.3346,
      "mean_token_accuracy": 0.8862806558609009,
      "num_tokens": 557483790.0,
      "step": 16020
    },
    {
      "epoch": 2.9751160631383473,
      "grad_norm": 1.594359548506007,
      "learning_rate": 1e-06,
      "loss": 0.358,
      "mean_token_accuracy": 0.8762151002883911,
      "num_tokens": 557523146.0,
      "step": 16021
    },
    {
      "epoch": 2.975301764159703,
      "grad_norm": 1.4727428092248094,
      "learning_rate": 1e-06,
      "loss": 0.3031,
      "mean_token_accuracy": 0.8935242891311646,
      "num_tokens": 557559636.0,
      "step": 16022
    },
    {
      "epoch": 2.9754874651810583,
      "grad_norm": 1.4528492438378382,
      "learning_rate": 1e-06,
      "loss": 0.3244,
      "mean_token_accuracy": 0.886033296585083,
      "num_tokens": 557599190.0,
      "step": 16023
    },
    {
      "epoch": 2.975673166202414,
      "grad_norm": 1.4171460855071112,
      "learning_rate": 1e-06,
      "loss": 0.2657,
      "mean_token_accuracy": 0.9064033031463623,
      "num_tokens": 557632996.0,
      "step": 16024
    },
    {
      "epoch": 2.97585886722377,
      "grad_norm": 1.498176727757101,
      "learning_rate": 1e-06,
      "loss": 0.3499,
      "mean_token_accuracy": 0.8767890334129333,
      "num_tokens": 557673422.0,
      "step": 16025
    },
    {
      "epoch": 2.9760445682451255,
      "grad_norm": 1.6213078965408756,
      "learning_rate": 1e-06,
      "loss": 0.3234,
      "mean_token_accuracy": 0.8888360261917114,
      "num_tokens": 557706068.0,
      "step": 16026
    },
    {
      "epoch": 2.976230269266481,
      "grad_norm": 1.524394598342905,
      "learning_rate": 1e-06,
      "loss": 0.3011,
      "mean_token_accuracy": 0.8959618806838989,
      "num_tokens": 557739427.0,
      "step": 16027
    },
    {
      "epoch": 2.9764159702878366,
      "grad_norm": 1.6026401354668438,
      "learning_rate": 1e-06,
      "loss": 0.3119,
      "mean_token_accuracy": 0.8888037204742432,
      "num_tokens": 557771088.0,
      "step": 16028
    },
    {
      "epoch": 2.9766016713091923,
      "grad_norm": 1.5821565342010737,
      "learning_rate": 1e-06,
      "loss": 0.3051,
      "mean_token_accuracy": 0.8942165970802307,
      "num_tokens": 557804329.0,
      "step": 16029
    },
    {
      "epoch": 2.9767873723305476,
      "grad_norm": 1.6175660359323836,
      "learning_rate": 1e-06,
      "loss": 0.3043,
      "mean_token_accuracy": 0.891724705696106,
      "num_tokens": 557836669.0,
      "step": 16030
    },
    {
      "epoch": 2.9769730733519033,
      "grad_norm": 1.652673396834634,
      "learning_rate": 1e-06,
      "loss": 0.3189,
      "mean_token_accuracy": 0.8879373669624329,
      "num_tokens": 557866737.0,
      "step": 16031
    },
    {
      "epoch": 2.977158774373259,
      "grad_norm": 1.6336956020637659,
      "learning_rate": 1e-06,
      "loss": 0.3002,
      "mean_token_accuracy": 0.8929877281188965,
      "num_tokens": 557897144.0,
      "step": 16032
    },
    {
      "epoch": 2.9773444753946148,
      "grad_norm": 1.6417051285216415,
      "learning_rate": 1e-06,
      "loss": 0.3262,
      "mean_token_accuracy": 0.8852614164352417,
      "num_tokens": 557935317.0,
      "step": 16033
    },
    {
      "epoch": 2.9775301764159705,
      "grad_norm": 1.6317251454069681,
      "learning_rate": 1e-06,
      "loss": 0.2939,
      "mean_token_accuracy": 0.8961162567138672,
      "num_tokens": 557967373.0,
      "step": 16034
    },
    {
      "epoch": 2.977715877437326,
      "grad_norm": 1.5181018916633866,
      "learning_rate": 1e-06,
      "loss": 0.2983,
      "mean_token_accuracy": 0.8953924179077148,
      "num_tokens": 558001685.0,
      "step": 16035
    },
    {
      "epoch": 2.9779015784586815,
      "grad_norm": 1.5461032275551962,
      "learning_rate": 1e-06,
      "loss": 0.2833,
      "mean_token_accuracy": 0.8970695734024048,
      "num_tokens": 558033580.0,
      "step": 16036
    },
    {
      "epoch": 2.9780872794800373,
      "grad_norm": 1.5980593060974486,
      "learning_rate": 1e-06,
      "loss": 0.3592,
      "mean_token_accuracy": 0.878176212310791,
      "num_tokens": 558068908.0,
      "step": 16037
    },
    {
      "epoch": 2.9782729805013926,
      "grad_norm": 1.5370830965555493,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8916556239128113,
      "num_tokens": 558106095.0,
      "step": 16038
    },
    {
      "epoch": 2.9784586815227483,
      "grad_norm": 1.5667301811631802,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8937072157859802,
      "num_tokens": 558137659.0,
      "step": 16039
    },
    {
      "epoch": 2.978644382544104,
      "grad_norm": 1.5325804896907762,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.892868161201477,
      "num_tokens": 558172643.0,
      "step": 16040
    },
    {
      "epoch": 2.9788300835654598,
      "grad_norm": 1.595056719517026,
      "learning_rate": 1e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8847838640213013,
      "num_tokens": 558204598.0,
      "step": 16041
    },
    {
      "epoch": 2.9790157845868155,
      "grad_norm": 1.4629208639484659,
      "learning_rate": 1e-06,
      "loss": 0.2931,
      "mean_token_accuracy": 0.8953781127929688,
      "num_tokens": 558243169.0,
      "step": 16042
    },
    {
      "epoch": 2.9792014856081708,
      "grad_norm": 1.7131183391820004,
      "learning_rate": 1e-06,
      "loss": 0.3039,
      "mean_token_accuracy": 0.8945974707603455,
      "num_tokens": 558270734.0,
      "step": 16043
    },
    {
      "epoch": 2.9793871866295265,
      "grad_norm": 1.6867838261346262,
      "learning_rate": 1e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.879366397857666,
      "num_tokens": 558303969.0,
      "step": 16044
    },
    {
      "epoch": 2.9795728876508822,
      "grad_norm": 1.7650005357494423,
      "learning_rate": 1e-06,
      "loss": 0.3445,
      "mean_token_accuracy": 0.8789799213409424,
      "num_tokens": 558333204.0,
      "step": 16045
    },
    {
      "epoch": 2.9797585886722375,
      "grad_norm": 1.5630221326611053,
      "learning_rate": 1e-06,
      "loss": 0.3306,
      "mean_token_accuracy": 0.8834360837936401,
      "num_tokens": 558371086.0,
      "step": 16046
    },
    {
      "epoch": 2.9799442896935933,
      "grad_norm": 1.6235377601508156,
      "learning_rate": 1e-06,
      "loss": 0.3415,
      "mean_token_accuracy": 0.8844128847122192,
      "num_tokens": 558406288.0,
      "step": 16047
    },
    {
      "epoch": 2.980129990714949,
      "grad_norm": 1.620655549370811,
      "learning_rate": 1e-06,
      "loss": 0.2958,
      "mean_token_accuracy": 0.8959097862243652,
      "num_tokens": 558435621.0,
      "step": 16048
    },
    {
      "epoch": 2.9803156917363047,
      "grad_norm": 1.6499928649077664,
      "learning_rate": 1e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8719555139541626,
      "num_tokens": 558469240.0,
      "step": 16049
    },
    {
      "epoch": 2.98050139275766,
      "grad_norm": 1.7212289816939343,
      "learning_rate": 1e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.882777214050293,
      "num_tokens": 558500305.0,
      "step": 16050
    },
    {
      "epoch": 2.9806870937790158,
      "grad_norm": 1.5576484901388044,
      "learning_rate": 1e-06,
      "loss": 0.3134,
      "mean_token_accuracy": 0.889353334903717,
      "num_tokens": 558535279.0,
      "step": 16051
    },
    {
      "epoch": 2.9808727948003715,
      "grad_norm": 1.5629662895766139,
      "learning_rate": 1e-06,
      "loss": 0.3374,
      "mean_token_accuracy": 0.8809905052185059,
      "num_tokens": 558568622.0,
      "step": 16052
    },
    {
      "epoch": 2.9810584958217268,
      "grad_norm": 1.4503082258696294,
      "learning_rate": 1e-06,
      "loss": 0.2529,
      "mean_token_accuracy": 0.9077696800231934,
      "num_tokens": 558598971.0,
      "step": 16053
    },
    {
      "epoch": 2.9812441968430825,
      "grad_norm": 1.7909402282741176,
      "learning_rate": 1e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.878819465637207,
      "num_tokens": 558627344.0,
      "step": 16054
    },
    {
      "epoch": 2.9814298978644382,
      "grad_norm": 1.4317667925213227,
      "learning_rate": 1e-06,
      "loss": 0.2971,
      "mean_token_accuracy": 0.8977000713348389,
      "num_tokens": 558664635.0,
      "step": 16055
    },
    {
      "epoch": 2.981615598885794,
      "grad_norm": 1.8032889949790416,
      "learning_rate": 1e-06,
      "loss": 0.3403,
      "mean_token_accuracy": 0.8870224356651306,
      "num_tokens": 558695963.0,
      "step": 16056
    },
    {
      "epoch": 2.9818012999071497,
      "grad_norm": 1.716483134950654,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.8847886323928833,
      "num_tokens": 558727355.0,
      "step": 16057
    },
    {
      "epoch": 2.981987000928505,
      "grad_norm": 1.474683863764191,
      "learning_rate": 1e-06,
      "loss": 0.3231,
      "mean_token_accuracy": 0.8867571353912354,
      "num_tokens": 558766971.0,
      "step": 16058
    },
    {
      "epoch": 2.9821727019498607,
      "grad_norm": 1.6095806868904172,
      "learning_rate": 1e-06,
      "loss": 0.3143,
      "mean_token_accuracy": 0.8912094831466675,
      "num_tokens": 558800528.0,
      "step": 16059
    },
    {
      "epoch": 2.9823584029712165,
      "grad_norm": 1.5566130235468638,
      "learning_rate": 1e-06,
      "loss": 0.3247,
      "mean_token_accuracy": 0.8854222893714905,
      "num_tokens": 558836760.0,
      "step": 16060
    },
    {
      "epoch": 2.9825441039925717,
      "grad_norm": 1.7195757989041838,
      "learning_rate": 1e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8774853348731995,
      "num_tokens": 558868997.0,
      "step": 16061
    },
    {
      "epoch": 2.9827298050139275,
      "grad_norm": 1.5108134173132943,
      "learning_rate": 1e-06,
      "loss": 0.293,
      "mean_token_accuracy": 0.8931254148483276,
      "num_tokens": 558901137.0,
      "step": 16062
    },
    {
      "epoch": 2.982915506035283,
      "grad_norm": 1.5159697272576351,
      "learning_rate": 1e-06,
      "loss": 0.3002,
      "mean_token_accuracy": 0.8929191827774048,
      "num_tokens": 558934966.0,
      "step": 16063
    },
    {
      "epoch": 2.983101207056639,
      "grad_norm": 1.4967482034161883,
      "learning_rate": 1e-06,
      "loss": 0.3474,
      "mean_token_accuracy": 0.8790993690490723,
      "num_tokens": 558974835.0,
      "step": 16064
    },
    {
      "epoch": 2.9832869080779947,
      "grad_norm": 1.5930039184738745,
      "learning_rate": 1e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8714011311531067,
      "num_tokens": 559009339.0,
      "step": 16065
    },
    {
      "epoch": 2.98347260909935,
      "grad_norm": 1.5813837086899099,
      "learning_rate": 1e-06,
      "loss": 0.3139,
      "mean_token_accuracy": 0.8874253034591675,
      "num_tokens": 559043987.0,
      "step": 16066
    },
    {
      "epoch": 2.9836583101207057,
      "grad_norm": 1.5858139513531129,
      "learning_rate": 1e-06,
      "loss": 0.2893,
      "mean_token_accuracy": 0.8960762023925781,
      "num_tokens": 559075160.0,
      "step": 16067
    },
    {
      "epoch": 2.9838440111420614,
      "grad_norm": 1.7272772644684538,
      "learning_rate": 1e-06,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8778070211410522,
      "num_tokens": 559105458.0,
      "step": 16068
    },
    {
      "epoch": 2.9840297121634167,
      "grad_norm": 1.5403256847455604,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8776159286499023,
      "num_tokens": 559144308.0,
      "step": 16069
    },
    {
      "epoch": 2.9842154131847725,
      "grad_norm": 1.5082200659403167,
      "learning_rate": 1e-06,
      "loss": 0.2784,
      "mean_token_accuracy": 0.9003788232803345,
      "num_tokens": 559175277.0,
      "step": 16070
    },
    {
      "epoch": 2.984401114206128,
      "grad_norm": 1.6360295473242907,
      "learning_rate": 1e-06,
      "loss": 0.3717,
      "mean_token_accuracy": 0.8719223141670227,
      "num_tokens": 559212483.0,
      "step": 16071
    },
    {
      "epoch": 2.984586815227484,
      "grad_norm": 1.3992684173155903,
      "learning_rate": 1e-06,
      "loss": 0.319,
      "mean_token_accuracy": 0.8855639696121216,
      "num_tokens": 559256183.0,
      "step": 16072
    },
    {
      "epoch": 2.984772516248839,
      "grad_norm": 1.6224146697463593,
      "learning_rate": 1e-06,
      "loss": 0.3013,
      "mean_token_accuracy": 0.8946789503097534,
      "num_tokens": 559286874.0,
      "step": 16073
    },
    {
      "epoch": 2.984958217270195,
      "grad_norm": 1.6203164721535013,
      "learning_rate": 1e-06,
      "loss": 0.3452,
      "mean_token_accuracy": 0.8819326162338257,
      "num_tokens": 559318351.0,
      "step": 16074
    },
    {
      "epoch": 2.9851439182915507,
      "grad_norm": 1.4244623017343883,
      "learning_rate": 1e-06,
      "loss": 0.3252,
      "mean_token_accuracy": 0.8824354410171509,
      "num_tokens": 559360056.0,
      "step": 16075
    },
    {
      "epoch": 2.985329619312906,
      "grad_norm": 1.6091075185334134,
      "learning_rate": 1e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8858778476715088,
      "num_tokens": 559394828.0,
      "step": 16076
    },
    {
      "epoch": 2.9855153203342617,
      "grad_norm": 1.6885442566901998,
      "learning_rate": 1e-06,
      "loss": 0.3517,
      "mean_token_accuracy": 0.8792510628700256,
      "num_tokens": 559424202.0,
      "step": 16077
    },
    {
      "epoch": 2.9857010213556174,
      "grad_norm": 1.545589521596676,
      "learning_rate": 1e-06,
      "loss": 0.3183,
      "mean_token_accuracy": 0.8847667574882507,
      "num_tokens": 559457893.0,
      "step": 16078
    },
    {
      "epoch": 2.985886722376973,
      "grad_norm": 1.4591668208269746,
      "learning_rate": 1e-06,
      "loss": 0.2989,
      "mean_token_accuracy": 0.8956102132797241,
      "num_tokens": 559494803.0,
      "step": 16079
    },
    {
      "epoch": 2.986072423398329,
      "grad_norm": 1.506176457099073,
      "learning_rate": 1e-06,
      "loss": 0.3188,
      "mean_token_accuracy": 0.890485405921936,
      "num_tokens": 559533455.0,
      "step": 16080
    },
    {
      "epoch": 2.986258124419684,
      "grad_norm": 1.5509078898284572,
      "learning_rate": 1e-06,
      "loss": 0.302,
      "mean_token_accuracy": 0.8933998346328735,
      "num_tokens": 559566835.0,
      "step": 16081
    },
    {
      "epoch": 2.98644382544104,
      "grad_norm": 1.5352507180233135,
      "learning_rate": 1e-06,
      "loss": 0.3335,
      "mean_token_accuracy": 0.882652759552002,
      "num_tokens": 559606071.0,
      "step": 16082
    },
    {
      "epoch": 2.9866295264623957,
      "grad_norm": 1.6449945813742277,
      "learning_rate": 1e-06,
      "loss": 0.3081,
      "mean_token_accuracy": 0.8896340727806091,
      "num_tokens": 559640587.0,
      "step": 16083
    },
    {
      "epoch": 2.986815227483751,
      "grad_norm": 1.3263285661883517,
      "learning_rate": 1e-06,
      "loss": 0.2938,
      "mean_token_accuracy": 0.896133303642273,
      "num_tokens": 559683743.0,
      "step": 16084
    },
    {
      "epoch": 2.9870009285051067,
      "grad_norm": 1.4572207423807508,
      "learning_rate": 1e-06,
      "loss": 0.3194,
      "mean_token_accuracy": 0.8844712972640991,
      "num_tokens": 559720370.0,
      "step": 16085
    },
    {
      "epoch": 2.9871866295264624,
      "grad_norm": 1.5745265422960135,
      "learning_rate": 1e-06,
      "loss": 0.3327,
      "mean_token_accuracy": 0.8831204175949097,
      "num_tokens": 559752946.0,
      "step": 16086
    },
    {
      "epoch": 2.987372330547818,
      "grad_norm": 1.5117403545177692,
      "learning_rate": 1e-06,
      "loss": 0.3101,
      "mean_token_accuracy": 0.8906697034835815,
      "num_tokens": 559789128.0,
      "step": 16087
    },
    {
      "epoch": 2.987558031569174,
      "grad_norm": 1.8860981068555576,
      "learning_rate": 1e-06,
      "loss": 0.3678,
      "mean_token_accuracy": 0.8730332851409912,
      "num_tokens": 559827172.0,
      "step": 16088
    },
    {
      "epoch": 2.987743732590529,
      "grad_norm": 1.6058145431093236,
      "learning_rate": 1e-06,
      "loss": 0.3097,
      "mean_token_accuracy": 0.8923341035842896,
      "num_tokens": 559860415.0,
      "step": 16089
    },
    {
      "epoch": 2.987929433611885,
      "grad_norm": 1.4861997287181905,
      "learning_rate": 1e-06,
      "loss": 0.3323,
      "mean_token_accuracy": 0.8798260688781738,
      "num_tokens": 559898815.0,
      "step": 16090
    },
    {
      "epoch": 2.9881151346332406,
      "grad_norm": 1.5601447347002613,
      "learning_rate": 1e-06,
      "loss": 0.3284,
      "mean_token_accuracy": 0.8857985138893127,
      "num_tokens": 559934356.0,
      "step": 16091
    },
    {
      "epoch": 2.988300835654596,
      "grad_norm": 1.4956901005305994,
      "learning_rate": 1e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8784763813018799,
      "num_tokens": 559972014.0,
      "step": 16092
    },
    {
      "epoch": 2.9884865366759517,
      "grad_norm": 1.41133263720233,
      "learning_rate": 1e-06,
      "loss": 0.3198,
      "mean_token_accuracy": 0.8898186683654785,
      "num_tokens": 560014346.0,
      "step": 16093
    },
    {
      "epoch": 2.9886722376973074,
      "grad_norm": 1.605401576024436,
      "learning_rate": 1e-06,
      "loss": 0.3146,
      "mean_token_accuracy": 0.8836506605148315,
      "num_tokens": 560047235.0,
      "step": 16094
    },
    {
      "epoch": 2.988857938718663,
      "grad_norm": 1.5007431593957787,
      "learning_rate": 1e-06,
      "loss": 0.3108,
      "mean_token_accuracy": 0.8903768062591553,
      "num_tokens": 560082273.0,
      "step": 16095
    },
    {
      "epoch": 2.9890436397400184,
      "grad_norm": 1.6548281398199844,
      "learning_rate": 1e-06,
      "loss": 0.3126,
      "mean_token_accuracy": 0.8907951712608337,
      "num_tokens": 560110605.0,
      "step": 16096
    },
    {
      "epoch": 2.989229340761374,
      "grad_norm": 1.3935700176472756,
      "learning_rate": 1e-06,
      "loss": 0.3197,
      "mean_token_accuracy": 0.8867889642715454,
      "num_tokens": 560155107.0,
      "step": 16097
    },
    {
      "epoch": 2.98941504178273,
      "grad_norm": 1.5810974804844948,
      "learning_rate": 1e-06,
      "loss": 0.2919,
      "mean_token_accuracy": 0.8978273868560791,
      "num_tokens": 560188263.0,
      "step": 16098
    },
    {
      "epoch": 2.989600742804085,
      "grad_norm": 1.6296304212056958,
      "learning_rate": 1e-06,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8735840320587158,
      "num_tokens": 560221654.0,
      "step": 16099
    },
    {
      "epoch": 2.989786443825441,
      "grad_norm": 1.463140640839436,
      "learning_rate": 1e-06,
      "loss": 0.2935,
      "mean_token_accuracy": 0.8959554433822632,
      "num_tokens": 560259373.0,
      "step": 16100
    },
    {
      "epoch": 2.9899721448467966,
      "grad_norm": 1.585402876690148,
      "learning_rate": 1e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8865007162094116,
      "num_tokens": 560293889.0,
      "step": 16101
    },
    {
      "epoch": 2.9901578458681524,
      "grad_norm": 1.6998448309417393,
      "learning_rate": 1e-06,
      "loss": 0.3792,
      "mean_token_accuracy": 0.8681405186653137,
      "num_tokens": 560335093.0,
      "step": 16102
    },
    {
      "epoch": 2.990343546889508,
      "grad_norm": 1.689845804594132,
      "learning_rate": 1e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.8807692527770996,
      "num_tokens": 560365255.0,
      "step": 16103
    },
    {
      "epoch": 2.9905292479108634,
      "grad_norm": 1.5699692587953666,
      "learning_rate": 1e-06,
      "loss": 0.3206,
      "mean_token_accuracy": 0.8874884843826294,
      "num_tokens": 560398935.0,
      "step": 16104
    },
    {
      "epoch": 2.990714948932219,
      "grad_norm": 1.5566713876967764,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.894489049911499,
      "num_tokens": 560434299.0,
      "step": 16105
    },
    {
      "epoch": 2.990900649953575,
      "grad_norm": 1.565485164887395,
      "learning_rate": 1e-06,
      "loss": 0.3095,
      "mean_token_accuracy": 0.8909623026847839,
      "num_tokens": 560469159.0,
      "step": 16106
    },
    {
      "epoch": 2.99108635097493,
      "grad_norm": 1.5305050056765959,
      "learning_rate": 1e-06,
      "loss": 0.3458,
      "mean_token_accuracy": 0.8816745281219482,
      "num_tokens": 560506441.0,
      "step": 16107
    },
    {
      "epoch": 2.991272051996286,
      "grad_norm": 1.6417322477854723,
      "learning_rate": 1e-06,
      "loss": 0.338,
      "mean_token_accuracy": 0.8819674253463745,
      "num_tokens": 560538021.0,
      "step": 16108
    },
    {
      "epoch": 2.9914577530176416,
      "grad_norm": 1.5360318500646615,
      "learning_rate": 1e-06,
      "loss": 0.3044,
      "mean_token_accuracy": 0.8917319774627686,
      "num_tokens": 560571674.0,
      "step": 16109
    },
    {
      "epoch": 2.9916434540389973,
      "grad_norm": 1.6634688900298542,
      "learning_rate": 1e-06,
      "loss": 0.3332,
      "mean_token_accuracy": 0.8803014159202576,
      "num_tokens": 560605831.0,
      "step": 16110
    },
    {
      "epoch": 2.991829155060353,
      "grad_norm": 1.793618580282661,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8832228183746338,
      "num_tokens": 560637207.0,
      "step": 16111
    },
    {
      "epoch": 2.9920148560817084,
      "grad_norm": 1.543274573820571,
      "learning_rate": 1e-06,
      "loss": 0.2909,
      "mean_token_accuracy": 0.8950292468070984,
      "num_tokens": 560672943.0,
      "step": 16112
    },
    {
      "epoch": 2.992200557103064,
      "grad_norm": 1.554472745514544,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.8867158889770508,
      "num_tokens": 560711333.0,
      "step": 16113
    },
    {
      "epoch": 2.99238625812442,
      "grad_norm": 1.4565623135304202,
      "learning_rate": 1e-06,
      "loss": 0.3087,
      "mean_token_accuracy": 0.8933162093162537,
      "num_tokens": 560749405.0,
      "step": 16114
    },
    {
      "epoch": 2.992571959145775,
      "grad_norm": 1.741303196707265,
      "learning_rate": 1e-06,
      "loss": 0.3127,
      "mean_token_accuracy": 0.8924738168716431,
      "num_tokens": 560778260.0,
      "step": 16115
    },
    {
      "epoch": 2.992757660167131,
      "grad_norm": 1.4646276466544519,
      "learning_rate": 1e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8793357610702515,
      "num_tokens": 560818506.0,
      "step": 16116
    },
    {
      "epoch": 2.9929433611884866,
      "grad_norm": 1.6889595426739603,
      "learning_rate": 1e-06,
      "loss": 0.3406,
      "mean_token_accuracy": 0.8822000026702881,
      "num_tokens": 560847980.0,
      "step": 16117
    },
    {
      "epoch": 2.9931290622098423,
      "grad_norm": 1.5968386267536254,
      "learning_rate": 1e-06,
      "loss": 0.3053,
      "mean_token_accuracy": 0.8928943872451782,
      "num_tokens": 560877213.0,
      "step": 16118
    },
    {
      "epoch": 2.9933147632311976,
      "grad_norm": 1.7461923916605802,
      "learning_rate": 1e-06,
      "loss": 0.3074,
      "mean_token_accuracy": 0.8929917812347412,
      "num_tokens": 560904019.0,
      "step": 16119
    },
    {
      "epoch": 2.9935004642525533,
      "grad_norm": 1.7465242069522309,
      "learning_rate": 1e-06,
      "loss": 0.3624,
      "mean_token_accuracy": 0.8754417896270752,
      "num_tokens": 560933865.0,
      "step": 16120
    },
    {
      "epoch": 2.993686165273909,
      "grad_norm": 1.506977992888207,
      "learning_rate": 1e-06,
      "loss": 0.2957,
      "mean_token_accuracy": 0.895025372505188,
      "num_tokens": 560964835.0,
      "step": 16121
    },
    {
      "epoch": 2.9938718662952644,
      "grad_norm": 1.6116318340524305,
      "learning_rate": 1e-06,
      "loss": 0.31,
      "mean_token_accuracy": 0.8875333070755005,
      "num_tokens": 560995892.0,
      "step": 16122
    },
    {
      "epoch": 2.99405756731662,
      "grad_norm": 1.5618337063039598,
      "learning_rate": 1e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8709485530853271,
      "num_tokens": 561032788.0,
      "step": 16123
    },
    {
      "epoch": 2.994243268337976,
      "grad_norm": 1.7760878188682128,
      "learning_rate": 1e-06,
      "loss": 0.3519,
      "mean_token_accuracy": 0.8767707347869873,
      "num_tokens": 561060198.0,
      "step": 16124
    },
    {
      "epoch": 2.9944289693593316,
      "grad_norm": 1.4891187561136927,
      "learning_rate": 1e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.883170485496521,
      "num_tokens": 561096499.0,
      "step": 16125
    },
    {
      "epoch": 2.9946146703806873,
      "grad_norm": 1.4514180253449007,
      "learning_rate": 1e-06,
      "loss": 0.3078,
      "mean_token_accuracy": 0.8927328586578369,
      "num_tokens": 561133583.0,
      "step": 16126
    },
    {
      "epoch": 2.9948003714020426,
      "grad_norm": 1.601735688282664,
      "learning_rate": 1e-06,
      "loss": 0.335,
      "mean_token_accuracy": 0.886197030544281,
      "num_tokens": 561165009.0,
      "step": 16127
    },
    {
      "epoch": 2.9949860724233983,
      "grad_norm": 1.453476758200466,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8938583135604858,
      "num_tokens": 561203222.0,
      "step": 16128
    },
    {
      "epoch": 2.995171773444754,
      "grad_norm": 1.4295475448181036,
      "learning_rate": 1e-06,
      "loss": 0.3337,
      "mean_token_accuracy": 0.8838577270507812,
      "num_tokens": 561245155.0,
      "step": 16129
    },
    {
      "epoch": 2.9953574744661093,
      "grad_norm": 1.4724702368634928,
      "learning_rate": 1e-06,
      "loss": 0.3021,
      "mean_token_accuracy": 0.8901517987251282,
      "num_tokens": 561277963.0,
      "step": 16130
    },
    {
      "epoch": 2.995543175487465,
      "grad_norm": 1.6964263544907323,
      "learning_rate": 1e-06,
      "loss": 0.3089,
      "mean_token_accuracy": 0.8925666213035583,
      "num_tokens": 561306271.0,
      "step": 16131
    },
    {
      "epoch": 2.995728876508821,
      "grad_norm": 1.740819572965597,
      "learning_rate": 1e-06,
      "loss": 0.3082,
      "mean_token_accuracy": 0.8909575939178467,
      "num_tokens": 561335960.0,
      "step": 16132
    },
    {
      "epoch": 2.9959145775301765,
      "grad_norm": 1.5908651476941056,
      "learning_rate": 1e-06,
      "loss": 0.334,
      "mean_token_accuracy": 0.8874925374984741,
      "num_tokens": 561367904.0,
      "step": 16133
    },
    {
      "epoch": 2.9961002785515323,
      "grad_norm": 1.411803024064803,
      "learning_rate": 1e-06,
      "loss": 0.3353,
      "mean_token_accuracy": 0.8838807344436646,
      "num_tokens": 561411669.0,
      "step": 16134
    },
    {
      "epoch": 2.9962859795728876,
      "grad_norm": 1.4309920830243652,
      "learning_rate": 1e-06,
      "loss": 0.2943,
      "mean_token_accuracy": 0.8949699401855469,
      "num_tokens": 561448501.0,
      "step": 16135
    },
    {
      "epoch": 2.9964716805942433,
      "grad_norm": 1.540514308019623,
      "learning_rate": 1e-06,
      "loss": 0.3386,
      "mean_token_accuracy": 0.8811811208724976,
      "num_tokens": 561487101.0,
      "step": 16136
    },
    {
      "epoch": 2.996657381615599,
      "grad_norm": 1.5324031553716007,
      "learning_rate": 1e-06,
      "loss": 0.3472,
      "mean_token_accuracy": 0.8809106945991516,
      "num_tokens": 561524242.0,
      "step": 16137
    },
    {
      "epoch": 2.9968430826369543,
      "grad_norm": 1.6579650336330387,
      "learning_rate": 1e-06,
      "loss": 0.3382,
      "mean_token_accuracy": 0.883201003074646,
      "num_tokens": 561557269.0,
      "step": 16138
    },
    {
      "epoch": 2.99702878365831,
      "grad_norm": 1.5077547343110655,
      "learning_rate": 1e-06,
      "loss": 0.2631,
      "mean_token_accuracy": 0.9044901728630066,
      "num_tokens": 561588695.0,
      "step": 16139
    },
    {
      "epoch": 2.997214484679666,
      "grad_norm": 1.5505643626768084,
      "learning_rate": 1e-06,
      "loss": 0.2994,
      "mean_token_accuracy": 0.8940442204475403,
      "num_tokens": 561620743.0,
      "step": 16140
    },
    {
      "epoch": 2.9974001857010215,
      "grad_norm": 1.5514027092987983,
      "learning_rate": 1e-06,
      "loss": 0.2895,
      "mean_token_accuracy": 0.8965452909469604,
      "num_tokens": 561651554.0,
      "step": 16141
    },
    {
      "epoch": 2.9975858867223772,
      "grad_norm": 1.4616995636732657,
      "learning_rate": 1e-06,
      "loss": 0.2671,
      "mean_token_accuracy": 0.9035912752151489,
      "num_tokens": 561685929.0,
      "step": 16142
    },
    {
      "epoch": 2.9977715877437325,
      "grad_norm": 1.3962403839169515,
      "learning_rate": 1e-06,
      "loss": 0.303,
      "mean_token_accuracy": 0.8918302655220032,
      "num_tokens": 561725953.0,
      "step": 16143
    },
    {
      "epoch": 2.9979572887650883,
      "grad_norm": 1.5861215713262702,
      "learning_rate": 1e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8683451414108276,
      "num_tokens": 561761977.0,
      "step": 16144
    },
    {
      "epoch": 2.9981429897864436,
      "grad_norm": 1.5554758672462257,
      "learning_rate": 1e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8808101415634155,
      "num_tokens": 561798906.0,
      "step": 16145
    },
    {
      "epoch": 2.9983286908077993,
      "grad_norm": 1.6331062144778943,
      "learning_rate": 1e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8863834142684937,
      "num_tokens": 561834033.0,
      "step": 16146
    },
    {
      "epoch": 2.998514391829155,
      "grad_norm": 1.459299865468086,
      "learning_rate": 1e-06,
      "loss": 0.3582,
      "mean_token_accuracy": 0.8749105334281921,
      "num_tokens": 561877281.0,
      "step": 16147
    },
    {
      "epoch": 2.9987000928505108,
      "grad_norm": 1.5443901527401813,
      "learning_rate": 1e-06,
      "loss": 0.3059,
      "mean_token_accuracy": 0.8940643072128296,
      "num_tokens": 561912397.0,
      "step": 16148
    },
    {
      "epoch": 2.9988857938718665,
      "grad_norm": 1.6702012323361521,
      "learning_rate": 1e-06,
      "loss": 0.3635,
      "mean_token_accuracy": 0.8741455078125,
      "num_tokens": 561946207.0,
      "step": 16149
    },
    {
      "epoch": 2.999071494893222,
      "grad_norm": 1.687583541851725,
      "learning_rate": 1e-06,
      "loss": 0.3307,
      "mean_token_accuracy": 0.8856293559074402,
      "num_tokens": 561974115.0,
      "step": 16150
    },
    {
      "epoch": 2.9992571959145775,
      "grad_norm": 1.5474823626770984,
      "learning_rate": 1e-06,
      "loss": 0.3291,
      "mean_token_accuracy": 0.8830249309539795,
      "num_tokens": 562010340.0,
      "step": 16151
    },
    {
      "epoch": 2.9994428969359332,
      "grad_norm": 1.586752221532267,
      "learning_rate": 1e-06,
      "loss": 0.2622,
      "mean_token_accuracy": 0.9067530632019043,
      "num_tokens": 562038532.0,
      "step": 16152
    },
    {
      "epoch": 2.9996285979572885,
      "grad_norm": 1.6731413916553897,
      "learning_rate": 1e-06,
      "loss": 0.324,
      "mean_token_accuracy": 0.8849363923072815,
      "num_tokens": 562067942.0,
      "step": 16153
    },
    {
      "epoch": 2.9998142989786443,
      "grad_norm": 1.5447431004035534,
      "learning_rate": 1e-06,
      "loss": 0.3149,
      "mean_token_accuracy": 0.8861258625984192,
      "num_tokens": 562100531.0,
      "step": 16154
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3962260366837365,
      "learning_rate": 1e-06,
      "loss": 0.3063,
      "mean_token_accuracy": 0.889656662940979,
      "num_tokens": 562140732.0,
      "step": 16155
    },
    {
      "epoch": 3.0,
      "step": 16155,
      "total_flos": 897986090926080.0,
      "train_loss": 0.37490086142067125,
      "train_runtime": 56645.1982,
      "train_samples_per_second": 4.563,
      "train_steps_per_second": 0.285
    }
  ],
  "logging_steps": 1,
  "max_steps": 16155,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 3231,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 897986090926080.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}