diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9472 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1347,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022271714922048997,
+      "grad_norm": 57.42307920658086,
+      "learning_rate": 0.0,
+      "loss": 11.1055,
+      "step": 1
+    },
+    {
+      "epoch": 0.004454342984409799,
+      "grad_norm": 55.4142581004584,
+      "learning_rate": 3.703703703703704e-07,
+      "loss": 11.1441,
+      "step": 2
+    },
+    {
+      "epoch": 0.0066815144766146995,
+      "grad_norm": 56.04170034902961,
+      "learning_rate": 7.407407407407408e-07,
+      "loss": 11.1293,
+      "step": 3
+    },
+    {
+      "epoch": 0.008908685968819599,
+      "grad_norm": 54.95758433278382,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 11.0293,
+      "step": 4
+    },
+    {
+      "epoch": 0.011135857461024499,
+      "grad_norm": 57.029873140763165,
+      "learning_rate": 1.4814814814814817e-06,
+      "loss": 11.1025,
+      "step": 5
+    },
+    {
+      "epoch": 0.013363028953229399,
+      "grad_norm": 58.88736752344377,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 10.9889,
+      "step": 6
+    },
+    {
+      "epoch": 0.015590200445434299,
+      "grad_norm": 64.03379235623899,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 10.6547,
+      "step": 7
+    },
+    {
+      "epoch": 0.017817371937639197,
+      "grad_norm": 87.78214617261006,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 9.6951,
+      "step": 8
+    },
+    {
+      "epoch": 0.0200445434298441,
+      "grad_norm": 99.69188628493322,
+      "learning_rate": 2.9629629629629633e-06,
+      "loss": 9.124,
+      "step": 9
+    },
+    {
+      "epoch": 0.022271714922048998,
+      "grad_norm": 111.96002267549564,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 8.5551,
+      "step": 10
+    },
+    {
+      "epoch": 0.024498886414253896,
+      "grad_norm": 63.81628779520451,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 3.6613,
+      "step": 11
+    },
+    {
+      "epoch": 0.026726057906458798,
+      "grad_norm": 54.32876659142105,
+      "learning_rate": 4.074074074074075e-06,
+      "loss": 3.2366,
+      "step": 12
+    },
+    {
+      "epoch": 0.028953229398663696,
+      "grad_norm": 36.65458024792978,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 2.5217,
+      "step": 13
+    },
+    {
+      "epoch": 0.031180400890868598,
+      "grad_norm": 30.57221885493666,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 2.317,
+      "step": 14
+    },
+    {
+      "epoch": 0.0334075723830735,
+      "grad_norm": 6.25280699614946,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 1.3326,
+      "step": 15
+    },
+    {
+      "epoch": 0.035634743875278395,
+      "grad_norm": 4.862666131069493,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 1.2688,
+      "step": 16
+    },
+    {
+      "epoch": 0.0378619153674833,
+      "grad_norm": 3.488769157504774,
+      "learning_rate": 5.925925925925927e-06,
+      "loss": 1.1812,
+      "step": 17
+    },
+    {
+      "epoch": 0.0400890868596882,
+      "grad_norm": 2.8021514727775503,
+      "learning_rate": 6.296296296296296e-06,
+      "loss": 1.1316,
+      "step": 18
+    },
+    {
+      "epoch": 0.042316258351893093,
+      "grad_norm": 2.199388422339967,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.0683,
+      "step": 19
+    },
+    {
+      "epoch": 0.044543429844097995,
+      "grad_norm": 5.525393881777902,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.9047,
+      "step": 20
+    },
+    {
+      "epoch": 0.0467706013363029,
+      "grad_norm": 11.02138041002174,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.8999,
+      "step": 21
+    },
+    {
+      "epoch": 0.04899777282850779,
+      "grad_norm": 1.5453174995875891,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.9263,
+      "step": 22
+    },
+    {
+      "epoch": 0.051224944320712694,
+      "grad_norm": 1.1422921936760477,
+      "learning_rate": 8.14814814814815e-06,
+      "loss": 0.8617,
+      "step": 23
+    },
+    {
+      "epoch": 0.053452115812917596,
+      "grad_norm": 0.950098543532029,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.8579,
+      "step": 24
+    },
+    {
+      "epoch": 0.0556792873051225,
+      "grad_norm": 0.8460261969278372,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.8283,
+      "step": 25
+    },
+    {
+      "epoch": 0.05790645879732739,
+      "grad_norm": 0.7528175819870679,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.8063,
+      "step": 26
+    },
+    {
+      "epoch": 0.060133630289532294,
+      "grad_norm": 0.7930777664338834,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.7268,
+      "step": 27
+    },
+    {
+      "epoch": 0.062360801781737196,
+      "grad_norm": 0.7913200183157787,
+      "learning_rate": 1e-05,
+      "loss": 0.7574,
+      "step": 28
+    },
+    {
+      "epoch": 0.0645879732739421,
+      "grad_norm": 0.6627676019174363,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.7387,
+      "step": 29
+    },
+    {
+      "epoch": 0.066815144766147,
+      "grad_norm": 0.6189513082849771,
+      "learning_rate": 1.074074074074074e-05,
+      "loss": 0.719,
+      "step": 30
+    },
+    {
+      "epoch": 0.06904231625835189,
+      "grad_norm": 0.6392565046924722,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.6966,
+      "step": 31
+    },
+    {
+      "epoch": 0.07126948775055679,
+      "grad_norm": 0.7613554413263333,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.6991,
+      "step": 32
+    },
+    {
+      "epoch": 0.07349665924276169,
+      "grad_norm": 0.8495800994596282,
+      "learning_rate": 1.1851851851851853e-05,
+      "loss": 0.6658,
+      "step": 33
+    },
+    {
+      "epoch": 0.0757238307349666,
+      "grad_norm": 0.45250299215013506,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.6828,
+      "step": 34
+    },
+    {
+      "epoch": 0.0779510022271715,
+      "grad_norm": 0.47839604371584693,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.6472,
+      "step": 35
+    },
+    {
+      "epoch": 0.0801781737193764,
+      "grad_norm": 0.5706925261506277,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.6527,
+      "step": 36
+    },
+    {
+      "epoch": 0.08240534521158129,
+      "grad_norm": 0.5490722862903289,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6394,
+      "step": 37
+    },
+    {
+      "epoch": 0.08463251670378619,
+      "grad_norm": 0.3997578606482899,
+      "learning_rate": 1.3703703703703704e-05,
+      "loss": 0.6088,
+      "step": 38
+    },
+    {
+      "epoch": 0.08685968819599109,
+      "grad_norm": 0.3888538634363526,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.6347,
+      "step": 39
+    },
+    {
+      "epoch": 0.08908685968819599,
+      "grad_norm": 0.4842890885503436,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.6166,
+      "step": 40
+    },
+    {
+      "epoch": 0.09131403118040089,
+      "grad_norm": 0.518705372079969,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.62,
+      "step": 41
+    },
+    {
+      "epoch": 0.0935412026726058,
+      "grad_norm": 0.35931037655330894,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 0.6058,
+      "step": 42
+    },
+    {
+      "epoch": 0.0957683741648107,
+      "grad_norm": 0.33027312330089215,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.6302,
+      "step": 43
+    },
+    {
+      "epoch": 0.09799554565701558,
+      "grad_norm": 0.36257798617418846,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.6142,
+      "step": 44
+    },
+    {
+      "epoch": 0.10022271714922049,
+      "grad_norm": 0.37194278361077726,
+      "learning_rate": 1.62962962962963e-05,
+      "loss": 0.5843,
+      "step": 45
+    },
+    {
+      "epoch": 0.10244988864142539,
+      "grad_norm": 0.3371545105094365,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.5784,
+      "step": 46
+    },
+    {
+      "epoch": 0.10467706013363029,
+      "grad_norm": 0.3297761659143661,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.6282,
+      "step": 47
+    },
+    {
+      "epoch": 0.10690423162583519,
+      "grad_norm": 0.2835121491028332,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 0.5727,
+      "step": 48
+    },
+    {
+      "epoch": 0.1091314031180401,
+      "grad_norm": 0.36396162705888246,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.5801,
+      "step": 49
+    },
+    {
+      "epoch": 0.111358574610245,
+      "grad_norm": 0.3485481215402162,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.5802,
+      "step": 50
+    },
+    {
+      "epoch": 0.11358574610244988,
+      "grad_norm": 0.2901852671276199,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.572,
+      "step": 51
+    },
+    {
+      "epoch": 0.11581291759465479,
+      "grad_norm": 0.33239940938527635,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.5871,
+      "step": 52
+    },
+    {
+      "epoch": 0.11804008908685969,
+      "grad_norm": 0.32272368620931075,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.5893,
+      "step": 53
+    },
+    {
+      "epoch": 0.12026726057906459,
+      "grad_norm": 0.310668770698013,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.5813,
+      "step": 54
+    },
+    {
+      "epoch": 0.12249443207126949,
+      "grad_norm": 0.2862388442261126,
+      "learning_rate": 2e-05,
+      "loss": 0.5555,
+      "step": 55
+    },
+    {
+      "epoch": 0.12472160356347439,
+      "grad_norm": 0.28836336316118216,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.5573,
+      "step": 56
+    },
+    {
+      "epoch": 0.12694877505567928,
+      "grad_norm": 0.3343688470290457,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.5607,
+      "step": 57
+    },
+    {
+      "epoch": 0.1291759465478842,
+      "grad_norm": 0.27729103785852094,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.5479,
+      "step": 58
+    },
+    {
+      "epoch": 0.13140311804008908,
+      "grad_norm": 0.3278121714571836,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.5583,
+      "step": 59
+    },
+    {
+      "epoch": 0.133630289532294,
+      "grad_norm": 0.25748897821127487,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 0.5662,
+      "step": 60
+    },
+    {
+      "epoch": 0.1358574610244989,
+      "grad_norm": 0.26680458322672634,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.531,
+      "step": 61
+    },
+    {
+      "epoch": 0.13808463251670378,
+      "grad_norm": 0.2712193652049242,
+      "learning_rate": 2.2592592592592594e-05,
+      "loss": 0.5297,
+      "step": 62
+    },
+    {
+      "epoch": 0.1403118040089087,
+      "grad_norm": 0.2549335734827084,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.5468,
+      "step": 63
+    },
+    {
+      "epoch": 0.14253897550111358,
+      "grad_norm": 0.28875698353844087,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.5219,
+      "step": 64
+    },
+    {
+      "epoch": 0.1447661469933185,
+      "grad_norm": 0.27391864481987854,
+      "learning_rate": 2.3703703703703707e-05,
+      "loss": 0.5204,
+      "step": 65
+    },
+    {
+      "epoch": 0.14699331848552338,
+      "grad_norm": 0.2692725207257318,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.548,
+      "step": 66
+    },
+    {
+      "epoch": 0.1492204899777283,
+      "grad_norm": 0.26312996576101033,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.5306,
+      "step": 67
+    },
+    {
+      "epoch": 0.1514476614699332,
+      "grad_norm": 0.24673062381809363,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 0.5293,
+      "step": 68
+    },
+    {
+      "epoch": 0.15367483296213807,
+      "grad_norm": 0.2544114168493051,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.5334,
+      "step": 69
+    },
+    {
+      "epoch": 0.155902004454343,
+      "grad_norm": 0.268517709442818,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.5563,
+      "step": 70
+    },
+    {
+      "epoch": 0.15812917594654788,
+      "grad_norm": 0.2489177187052713,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.5362,
+      "step": 71
+    },
+    {
+      "epoch": 0.1603563474387528,
+      "grad_norm": 0.2615863452691099,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 0.5451,
+      "step": 72
+    },
+    {
+      "epoch": 0.16258351893095768,
+      "grad_norm": 0.26501318885913844,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5306,
+      "step": 73
+    },
+    {
+      "epoch": 0.16481069042316257,
+      "grad_norm": 0.24947289257521604,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 0.5208,
+      "step": 74
+    },
+    {
+      "epoch": 0.16703786191536749,
+      "grad_norm": 0.28127140174684223,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 0.5338,
+      "step": 75
+    },
+    {
+      "epoch": 0.16926503340757237,
+      "grad_norm": 0.26849164318476576,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.5386,
+      "step": 76
+    },
+    {
+      "epoch": 0.1714922048997773,
+      "grad_norm": 0.263834539821763,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.5436,
+      "step": 77
+    },
+    {
+      "epoch": 0.17371937639198218,
+      "grad_norm": 0.2566975308908174,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 0.5352,
+      "step": 78
+    },
+    {
+      "epoch": 0.1759465478841871,
+      "grad_norm": 0.2993215207257922,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.5334,
+      "step": 79
+    },
+    {
+      "epoch": 0.17817371937639198,
+      "grad_norm": 0.2649930701553602,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 0.529,
+      "step": 80
+    },
+    {
+      "epoch": 0.18040089086859687,
+      "grad_norm": 0.2499755228378194,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.5129,
+      "step": 81
+    },
+    {
+      "epoch": 0.18262806236080179,
+      "grad_norm": 0.25585388447729907,
+      "learning_rate": 3e-05,
+      "loss": 0.5031,
+      "step": 82
+    },
+    {
+      "epoch": 0.18485523385300667,
+      "grad_norm": 0.2804971309732976,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 0.52,
+      "step": 83
+    },
+    {
+      "epoch": 0.1870824053452116,
+      "grad_norm": 0.30532069489713315,
+      "learning_rate": 3.074074074074074e-05,
+      "loss": 0.5113,
+      "step": 84
+    },
+    {
+      "epoch": 0.18930957683741648,
+      "grad_norm": 0.31239584228708184,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.5046,
+      "step": 85
+    },
+    {
+      "epoch": 0.1915367483296214,
+      "grad_norm": 0.29923465866612675,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.5171,
+      "step": 86
+    },
+    {
+      "epoch": 0.19376391982182628,
+      "grad_norm": 0.3220264210471262,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.5306,
+      "step": 87
+    },
+    {
+      "epoch": 0.19599109131403117,
+      "grad_norm": 0.3015679068134566,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 0.5029,
+      "step": 88
+    },
+    {
+      "epoch": 0.19821826280623608,
+      "grad_norm": 0.2791608421023344,
+      "learning_rate": 3.25925925925926e-05,
+      "loss": 0.4814,
+      "step": 89
+    },
+    {
+      "epoch": 0.20044543429844097,
+      "grad_norm": 0.2659375754124709,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 0.5111,
+      "step": 90
+    },
+    {
+      "epoch": 0.2026726057906459,
+      "grad_norm": 0.2997121598755476,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4969,
+      "step": 91
+    },
+    {
+      "epoch": 0.20489977728285078,
+      "grad_norm": 0.27277673860961665,
+      "learning_rate": 3.3703703703703706e-05,
+      "loss": 0.525,
+      "step": 92
+    },
+    {
+      "epoch": 0.2071269487750557,
+      "grad_norm": 0.2694873726831575,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.5277,
+      "step": 93
+    },
+    {
+      "epoch": 0.20935412026726058,
+      "grad_norm": 0.3357690547348528,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.4768,
+      "step": 94
+    },
+    {
+      "epoch": 0.21158129175946547,
+      "grad_norm": 0.3097028888192043,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 0.5327,
+      "step": 95
+    },
+    {
+      "epoch": 0.21380846325167038,
+      "grad_norm": 0.274982135970598,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.5133,
+      "step": 96
+    },
+    {
+      "epoch": 0.21603563474387527,
+      "grad_norm": 0.2689619091248766,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.5061,
+      "step": 97
+    },
+    {
+      "epoch": 0.2182628062360802,
+      "grad_norm": 0.276310674407535,
+      "learning_rate": 3.592592592592593e-05,
+      "loss": 0.5023,
+      "step": 98
+    },
+    {
+      "epoch": 0.22048997772828507,
+      "grad_norm": 0.2925539174112595,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.4902,
+      "step": 99
+    },
+    {
+      "epoch": 0.22271714922049,
+      "grad_norm": 0.2764873755412426,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.4928,
+      "step": 100
+    },
+    {
+      "epoch": 0.22494432071269488,
+      "grad_norm": 0.30596253382561084,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.4947,
+      "step": 101
+    },
+    {
+      "epoch": 0.22717149220489977,
+      "grad_norm": 0.2507900196157514,
+      "learning_rate": 3.740740740740741e-05,
+      "loss": 0.4979,
+      "step": 102
+    },
+    {
+      "epoch": 0.22939866369710468,
+      "grad_norm": 0.41155747976418094,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.5044,
+      "step": 103
+    },
+    {
+      "epoch": 0.23162583518930957,
+      "grad_norm": 0.40004605557757594,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 0.4914,
+      "step": 104
+    },
+    {
+      "epoch": 0.23385300668151449,
+      "grad_norm": 0.3158821021617771,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.4949,
+      "step": 105
+    },
+    {
+      "epoch": 0.23608017817371937,
+      "grad_norm": 0.43094171027762884,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.5046,
+      "step": 106
+    },
+    {
+      "epoch": 0.2383073496659243,
+      "grad_norm": 0.3682504918929769,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.4602,
+      "step": 107
+    },
+    {
+      "epoch": 0.24053452115812918,
+      "grad_norm": 0.33297596471036733,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 0.5138,
+      "step": 108
+    },
+    {
+      "epoch": 0.24276169265033407,
+      "grad_norm": 0.3721638469385512,
+      "learning_rate": 4e-05,
+      "loss": 0.4816,
+      "step": 109
+    },
+    {
+      "epoch": 0.24498886414253898,
+      "grad_norm": 0.3525132455331721,
+      "learning_rate": 4.0370370370370374e-05,
+      "loss": 0.491,
+      "step": 110
+    },
+    {
+      "epoch": 0.24721603563474387,
+      "grad_norm": 0.3284661170025923,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.5086,
+      "step": 111
+    },
+    {
+      "epoch": 0.24944320712694878,
+      "grad_norm": 0.4553024801346518,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 0.4967,
+      "step": 112
+    },
+    {
+      "epoch": 0.2516703786191537,
+      "grad_norm": 0.36062775373958933,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.5107,
+      "step": 113
+    },
+    {
+      "epoch": 0.25389755011135856,
+      "grad_norm": 0.3139139462586485,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 0.4931,
+      "step": 114
+    },
+    {
+      "epoch": 0.2561247216035635,
+      "grad_norm": 0.43508729251982736,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.5057,
+      "step": 115
+    },
+    {
+      "epoch": 0.2583518930957684,
+      "grad_norm": 0.37579492262839576,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.4827,
+      "step": 116
+    },
+    {
+      "epoch": 0.26057906458797325,
+      "grad_norm": 0.3135964755412191,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.4951,
+      "step": 117
+    },
+    {
+      "epoch": 0.26280623608017817,
+      "grad_norm": 0.5103971124839464,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.4883,
+      "step": 118
+    },
+    {
+      "epoch": 0.2650334075723831,
+      "grad_norm": 0.3886694527089134,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.459,
+      "step": 119
+    },
+    {
+      "epoch": 0.267260579064588,
+      "grad_norm": 0.34873133836486525,
+      "learning_rate": 4.4074074074074076e-05,
+      "loss": 0.4935,
+      "step": 120
+    },
+    {
+      "epoch": 0.26948775055679286,
+      "grad_norm": 0.6019576624584538,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.4868,
+      "step": 121
+    },
+    {
+      "epoch": 0.2717149220489978,
+      "grad_norm": 0.36932194797825146,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.4784,
+      "step": 122
+    },
+    {
+      "epoch": 0.2739420935412027,
+      "grad_norm": 0.3804288507510714,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.4774,
+      "step": 123
+    },
+    {
+      "epoch": 0.27616926503340755,
+      "grad_norm": 0.450739644920063,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.4834,
+      "step": 124
+    },
+    {
+      "epoch": 0.27839643652561247,
+      "grad_norm": 0.3166655802453127,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.5027,
+      "step": 125
+    },
+    {
+      "epoch": 0.2806236080178174,
+      "grad_norm": 0.3568553819877089,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.4719,
+      "step": 126
+    },
+    {
+      "epoch": 0.2828507795100223,
+      "grad_norm": 0.2879304124523538,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.4575,
+      "step": 127
+    },
+    {
+      "epoch": 0.28507795100222716,
+      "grad_norm": 0.3859580233017386,
+      "learning_rate": 4.703703703703704e-05,
+      "loss": 0.4751,
+      "step": 128
+    },
+    {
+      "epoch": 0.2873051224944321,
+      "grad_norm": 0.32691499386259953,
+      "learning_rate": 4.740740740740741e-05,
+      "loss": 0.456,
+      "step": 129
+    },
+    {
+      "epoch": 0.289532293986637,
+      "grad_norm": 0.32665326828281926,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.4691,
+      "step": 130
+    },
+    {
+      "epoch": 0.29175946547884185,
+      "grad_norm": 0.4003858496357345,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.5027,
+      "step": 131
+    },
+    {
+      "epoch": 0.29398663697104677,
+      "grad_norm": 0.3287864746571255,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 0.4568,
+      "step": 132
+    },
+    {
+      "epoch": 0.2962138084632517,
+      "grad_norm": 0.34638666355485653,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.4548,
+      "step": 133
+    },
+    {
+      "epoch": 0.2984409799554566,
+      "grad_norm": 0.3736338755007743,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 0.4608,
+      "step": 134
+    },
+    {
+      "epoch": 0.30066815144766146,
+      "grad_norm": 0.3606530799746982,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.4967,
+      "step": 135
+    },
+    {
+      "epoch": 0.3028953229398664,
+      "grad_norm": 0.35828978658937044,
+      "learning_rate": 5e-05,
+      "loss": 0.4663,
+      "step": 136
+    },
+    {
+      "epoch": 0.3051224944320713,
+      "grad_norm": 0.38886007461404093,
+      "learning_rate": 4.995874587458746e-05,
+      "loss": 0.4942,
+      "step": 137
+    },
+    {
+      "epoch": 0.30734966592427615,
+      "grad_norm": 0.3492765115474014,
+      "learning_rate": 4.991749174917492e-05,
+      "loss": 0.502,
+      "step": 138
+    },
+    {
+      "epoch": 0.30957683741648107,
+      "grad_norm": 0.3260537303402738,
+      "learning_rate": 4.987623762376238e-05,
+      "loss": 0.4809,
+      "step": 139
+    },
+    {
+      "epoch": 0.311804008908686,
+      "grad_norm": 0.27678766098703955,
+      "learning_rate": 4.9834983498349835e-05,
+      "loss": 0.4662,
+      "step": 140
+    },
+    {
+      "epoch": 0.31403118040089084,
+      "grad_norm": 0.32583906322889694,
+      "learning_rate": 4.97937293729373e-05,
+      "loss": 0.4744,
+      "step": 141
+    },
+    {
+      "epoch": 0.31625835189309576,
+      "grad_norm": 0.30968501966537665,
+      "learning_rate": 4.975247524752475e-05,
+      "loss": 0.4693,
+      "step": 142
+    },
+    {
+      "epoch": 0.3184855233853007,
+      "grad_norm": 0.300558723706453,
+      "learning_rate": 4.971122112211221e-05,
+      "loss": 0.4543,
+      "step": 143
+    },
+    {
+      "epoch": 0.3207126948775056,
+      "grad_norm": 0.3590756342843173,
+      "learning_rate": 4.9669966996699675e-05,
+      "loss": 0.4848,
+      "step": 144
+    },
+    {
+      "epoch": 0.32293986636971045,
+      "grad_norm": 0.3103656322051761,
+      "learning_rate": 4.9628712871287133e-05,
+      "loss": 0.481,
+      "step": 145
+    },
+    {
+      "epoch": 0.32516703786191536,
+      "grad_norm": 0.36889845947563116,
+      "learning_rate": 4.9587458745874585e-05,
+      "loss": 0.4765,
+      "step": 146
+    },
+    {
+      "epoch": 0.3273942093541203,
+      "grad_norm": 0.34252947893290014,
+      "learning_rate": 4.954620462046205e-05,
+      "loss": 0.4936,
+      "step": 147
+    },
+    {
+      "epoch": 0.32962138084632514,
+      "grad_norm": 0.3848672223009835,
+      "learning_rate": 4.950495049504951e-05,
+      "loss": 0.4498,
+      "step": 148
+    },
+    {
+      "epoch": 0.33184855233853006,
+      "grad_norm": 0.29661062901684304,
+      "learning_rate": 4.9463696369636967e-05,
+      "loss": 0.4878,
+      "step": 149
+    },
+    {
+      "epoch": 0.33407572383073497,
+      "grad_norm": 0.3117169655533877,
+      "learning_rate": 4.9422442244224425e-05,
+      "loss": 0.4786,
+      "step": 150
+    },
+    {
+      "epoch": 0.3363028953229399,
+      "grad_norm": 0.29938186320927185,
+      "learning_rate": 4.938118811881188e-05,
+      "loss": 0.484,
+      "step": 151
+    },
+    {
+      "epoch": 0.33853006681514475,
+      "grad_norm": 0.3677147215204186,
+      "learning_rate": 4.933993399339934e-05,
+      "loss": 0.4943,
+      "step": 152
+    },
+    {
+      "epoch": 0.34075723830734966,
+      "grad_norm": 0.31870028405815753,
+      "learning_rate": 4.92986798679868e-05,
+      "loss": 0.466,
+      "step": 153
+    },
+    {
+      "epoch": 0.3429844097995546,
+      "grad_norm": 0.30485295371098536,
+      "learning_rate": 4.925742574257426e-05,
+      "loss": 0.4716,
+      "step": 154
+    },
+    {
+      "epoch": 0.34521158129175944,
+      "grad_norm": 0.3204799751546297,
+      "learning_rate": 4.9216171617161716e-05,
+      "loss": 0.482,
+      "step": 155
+    },
+    {
+      "epoch": 0.34743875278396436,
+      "grad_norm": 0.3495660839808905,
+      "learning_rate": 4.917491749174918e-05,
+      "loss": 0.4746,
+      "step": 156
+    },
+    {
+      "epoch": 0.34966592427616927,
+      "grad_norm": 0.3758329591880699,
+      "learning_rate": 4.913366336633663e-05,
+      "loss": 0.5131,
+      "step": 157
+    },
+    {
+      "epoch": 0.3518930957683742,
+      "grad_norm": 0.32055625709553975,
+      "learning_rate": 4.909240924092409e-05,
+      "loss": 0.4833,
+      "step": 158
+    },
+    {
+      "epoch": 0.35412026726057905,
+      "grad_norm": 0.32097261525837784,
+      "learning_rate": 4.9051155115511556e-05,
+      "loss": 0.48,
+      "step": 159
+    },
+    {
+      "epoch": 0.35634743875278396,
+      "grad_norm": 0.33480318109614066,
+      "learning_rate": 4.9009900990099014e-05,
+      "loss": 0.4551,
+      "step": 160
+    },
+    {
+      "epoch": 0.3585746102449889,
+      "grad_norm": 0.2781227223952856,
+      "learning_rate": 4.8968646864686466e-05,
+      "loss": 0.47,
+      "step": 161
+    },
+    {
+      "epoch": 0.36080178173719374,
+      "grad_norm": 0.38773722729695975,
+      "learning_rate": 4.892739273927393e-05,
+      "loss": 0.4873,
+      "step": 162
+    },
+    {
+      "epoch": 0.36302895322939865,
+      "grad_norm": 0.27943792922418004,
+      "learning_rate": 4.888613861386139e-05,
+      "loss": 0.4584,
+      "step": 163
+    },
+    {
+      "epoch": 0.36525612472160357,
+      "grad_norm": 0.30003973698376857,
+      "learning_rate": 4.884488448844885e-05,
+      "loss": 0.4636,
+      "step": 164
+    },
+    {
+      "epoch": 0.3674832962138085,
+      "grad_norm": 0.3065918295334727,
+      "learning_rate": 4.8803630363036305e-05,
+      "loss": 0.4776,
+      "step": 165
+    },
+    {
+      "epoch": 0.36971046770601335,
+      "grad_norm": 0.3201763919174801,
+      "learning_rate": 4.8762376237623764e-05,
+      "loss": 0.4961,
+      "step": 166
+    },
+    {
+      "epoch": 0.37193763919821826,
+      "grad_norm": 0.3178300180539279,
+      "learning_rate": 4.872112211221123e-05,
+      "loss": 0.4759,
+      "step": 167
+    },
+    {
+      "epoch": 0.3741648106904232,
+      "grad_norm": 0.29842427156720247,
+      "learning_rate": 4.867986798679868e-05,
+      "loss": 0.4743,
+      "step": 168
+    },
+    {
+      "epoch": 0.37639198218262804,
+      "grad_norm": 0.32502426510303545,
+      "learning_rate": 4.863861386138614e-05,
+      "loss": 0.4669,
+      "step": 169
+    },
+    {
+      "epoch": 0.37861915367483295,
+      "grad_norm": 0.29588998621776147,
+      "learning_rate": 4.8597359735973603e-05,
+      "loss": 0.4669,
+      "step": 170
+    },
+    {
+      "epoch": 0.38084632516703787,
+      "grad_norm": 0.2899635288815167,
+      "learning_rate": 4.855610561056106e-05,
+      "loss": 0.4855,
+      "step": 171
+    },
+    {
+      "epoch": 0.3830734966592428,
+      "grad_norm": 0.3064392168661253,
+      "learning_rate": 4.851485148514851e-05,
+      "loss": 0.4636,
+      "step": 172
+    },
+    {
+      "epoch": 0.38530066815144765,
+      "grad_norm": 0.2986024540889945,
+      "learning_rate": 4.847359735973598e-05,
+      "loss": 0.4876,
+      "step": 173
+    },
+    {
+      "epoch": 0.38752783964365256,
+      "grad_norm": 0.30233642625624657,
+      "learning_rate": 4.8432343234323437e-05,
+      "loss": 0.4731,
+      "step": 174
+    },
+    {
+      "epoch": 0.3897550111358575,
+      "grad_norm": 0.27146933961382946,
+      "learning_rate": 4.8391089108910895e-05,
+      "loss": 0.4724,
+      "step": 175
+    },
+    {
+      "epoch": 0.39198218262806234,
+      "grad_norm": 0.38926742917871604,
+      "learning_rate": 4.834983498349835e-05,
+      "loss": 0.4592,
+      "step": 176
+    },
+    {
+      "epoch": 0.39420935412026725,
+      "grad_norm": 0.2918121435091572,
+      "learning_rate": 4.830858085808581e-05,
+      "loss": 0.478,
+      "step": 177
+    },
+    {
+      "epoch": 0.39643652561247217,
+      "grad_norm": 0.2733312948862096,
+      "learning_rate": 4.826732673267327e-05,
+      "loss": 0.4326,
+      "step": 178
+    },
+    {
+      "epoch": 0.3986636971046771,
+      "grad_norm": 0.28519888504000973,
+      "learning_rate": 4.822607260726073e-05,
+      "loss": 0.472,
+      "step": 179
+    },
+    {
+      "epoch": 0.40089086859688194,
+      "grad_norm": 0.3046631075865034,
+      "learning_rate": 4.8184818481848186e-05,
+      "loss": 0.4724,
+      "step": 180
+    },
+    {
+      "epoch": 0.40311804008908686,
+      "grad_norm": 0.28826960158466797,
+      "learning_rate": 4.8143564356435644e-05,
+      "loss": 0.4741,
+      "step": 181
+    },
+    {
+      "epoch": 0.4053452115812918,
+      "grad_norm": 0.2701880021074098,
+      "learning_rate": 4.810231023102311e-05,
+      "loss": 0.4616,
+      "step": 182
+    },
+    {
+      "epoch": 0.40757238307349664,
+      "grad_norm": 0.27599826258023064,
+      "learning_rate": 4.806105610561056e-05,
+      "loss": 0.4548,
+      "step": 183
+    },
+    {
+      "epoch": 0.40979955456570155,
+      "grad_norm": 0.30476562089568104,
+      "learning_rate": 4.801980198019802e-05,
+      "loss": 0.4479,
+      "step": 184
+    },
+    {
+      "epoch": 0.41202672605790647,
+      "grad_norm": 0.2790459906115023,
+      "learning_rate": 4.7978547854785484e-05,
+      "loss": 0.475,
+      "step": 185
+    },
+    {
+      "epoch": 0.4142538975501114,
+      "grad_norm": 0.27612981777114276,
+      "learning_rate": 4.793729372937294e-05,
+      "loss": 0.4789,
+      "step": 186
+    },
+    {
+      "epoch": 0.41648106904231624,
+      "grad_norm": 0.32980387723538307,
+      "learning_rate": 4.7896039603960394e-05,
+      "loss": 0.4677,
+      "step": 187
+    },
+    {
+      "epoch": 0.41870824053452116,
+      "grad_norm": 0.29955756261599203,
+      "learning_rate": 4.785478547854786e-05,
+      "loss": 0.4815,
+      "step": 188
+    },
+    {
+      "epoch": 0.4209354120267261,
+      "grad_norm": 0.35755084539372534,
+      "learning_rate": 4.781353135313532e-05,
+      "loss": 0.4582,
+      "step": 189
+    },
+    {
+      "epoch": 0.42316258351893093,
+      "grad_norm": 0.31455989081013724,
+      "learning_rate": 4.7772277227722775e-05,
+      "loss": 0.4811,
+      "step": 190
+    },
+    {
+      "epoch": 0.42538975501113585,
+      "grad_norm": 0.29333274730527464,
+      "learning_rate": 4.7731023102310234e-05,
+      "loss": 0.467,
+      "step": 191
+    },
+    {
+      "epoch": 0.42761692650334077,
+      "grad_norm": 0.2877037349403691,
+      "learning_rate": 4.768976897689769e-05,
+      "loss": 0.459,
+      "step": 192
+    },
+    {
+      "epoch": 0.4298440979955457,
+      "grad_norm": 0.2958273465058758,
+      "learning_rate": 4.764851485148515e-05,
+      "loss": 0.4998,
+      "step": 193
+    },
+    {
+      "epoch": 0.43207126948775054,
+      "grad_norm": 0.2781440019522109,
+      "learning_rate": 4.760726072607261e-05,
+      "loss": 0.4439,
+      "step": 194
+    },
+    {
+      "epoch": 0.43429844097995546,
+      "grad_norm": 0.283439750955963,
+      "learning_rate": 4.756600660066007e-05,
+      "loss": 0.4626,
+      "step": 195
+    },
+    {
+      "epoch": 0.4365256124721604,
+      "grad_norm": 0.2911005370573716,
+      "learning_rate": 4.7524752475247525e-05,
+      "loss": 0.4625,
+      "step": 196
+    },
+    {
+      "epoch": 0.43875278396436523,
+      "grad_norm": 0.25120199923795783,
+      "learning_rate": 4.748349834983499e-05,
+      "loss": 0.4488,
+      "step": 197
+    },
+    {
+      "epoch": 0.44097995545657015,
+      "grad_norm": 0.2757619493421524,
+      "learning_rate": 4.744224422442244e-05,
+      "loss": 0.4513,
+      "step": 198
+    },
+    {
+      "epoch": 0.44320712694877507,
+      "grad_norm": 0.3009536575674645,
+      "learning_rate": 4.74009900990099e-05,
+      "loss": 0.4663,
+      "step": 199
+    },
+    {
+      "epoch": 0.44543429844098,
+      "grad_norm": 0.2672399691893589,
+      "learning_rate": 4.7359735973597365e-05,
+      "loss": 0.4563,
+      "step": 200
+    },
+    {
+      "epoch": 0.44766146993318484,
+      "grad_norm": 0.31954353984499995,
+      "learning_rate": 4.731848184818482e-05,
+      "loss": 0.4651,
+      "step": 201
+    },
+    {
+      "epoch": 0.44988864142538976,
+      "grad_norm": 0.2773874407552882,
+      "learning_rate": 4.7277227722772274e-05,
+      "loss": 0.4807,
+      "step": 202
+    },
+    {
+      "epoch": 0.4521158129175947,
+      "grad_norm": 0.26492332012667835,
+      "learning_rate": 4.723597359735974e-05,
+      "loss": 0.4543,
+      "step": 203
+    },
+    {
+      "epoch": 0.45434298440979953,
+      "grad_norm": 0.31184436207960836,
+      "learning_rate": 4.71947194719472e-05,
+      "loss": 0.4694,
+      "step": 204
+    },
+    {
+      "epoch": 0.45657015590200445,
+      "grad_norm": 0.2731081790924368,
+      "learning_rate": 4.7153465346534656e-05,
+      "loss": 0.4572,
+      "step": 205
+    },
+    {
+      "epoch": 0.45879732739420936,
+      "grad_norm": 0.2869232775138865,
+      "learning_rate": 4.7112211221122114e-05,
+      "loss": 0.4681,
+      "step": 206
+    },
+    {
+      "epoch": 0.4610244988864143,
+      "grad_norm": 0.2639755248576992,
+      "learning_rate": 4.707095709570957e-05,
+      "loss": 0.4602,
+      "step": 207
+    },
+    {
+      "epoch": 0.46325167037861914,
+      "grad_norm": 0.29189504155707585,
+      "learning_rate": 4.702970297029703e-05,
+      "loss": 0.4664,
+      "step": 208
+    },
+    {
+      "epoch": 0.46547884187082406,
+      "grad_norm": 0.2601644937321229,
+      "learning_rate": 4.698844884488449e-05,
+      "loss": 0.4483,
+      "step": 209
+    },
+    {
+      "epoch": 0.46770601336302897,
+      "grad_norm": 0.27106301913748243,
+      "learning_rate": 4.694719471947195e-05,
+      "loss": 0.4716,
+      "step": 210
+    },
+    {
+      "epoch": 0.46993318485523383,
+      "grad_norm": 0.31324919458454953,
+      "learning_rate": 4.6905940594059406e-05,
+      "loss": 0.4886,
+      "step": 211
+    },
+    {
+      "epoch": 0.47216035634743875,
+      "grad_norm": 0.28748118504498216,
+      "learning_rate": 4.686468646864687e-05,
+      "loss": 0.4462,
+      "step": 212
+    },
+    {
+      "epoch": 0.47438752783964366,
+      "grad_norm": 0.26192959125223536,
+      "learning_rate": 4.682343234323432e-05,
+      "loss": 0.466,
+      "step": 213
+    },
+    {
+      "epoch": 0.4766146993318486,
+      "grad_norm": 0.2868158200559901,
+      "learning_rate": 4.678217821782179e-05,
+      "loss": 0.4393,
+      "step": 214
+    },
+    {
+      "epoch": 0.47884187082405344,
+      "grad_norm": 0.3082515377114821,
+      "learning_rate": 4.6740924092409245e-05,
+      "loss": 0.4449,
+      "step": 215
+    },
+    {
+      "epoch": 0.48106904231625836,
+      "grad_norm": 0.34124651164996833,
+      "learning_rate": 4.6699669966996704e-05,
+      "loss": 0.4772,
+      "step": 216
+    },
+    {
+      "epoch": 0.48329621380846327,
+      "grad_norm": 0.2944775803261396,
+      "learning_rate": 4.665841584158416e-05,
+      "loss": 0.449,
+      "step": 217
+    },
+    {
+      "epoch": 0.48552338530066813,
+      "grad_norm": 0.3238597167556311,
+      "learning_rate": 4.661716171617162e-05,
+      "loss": 0.4704,
+      "step": 218
+    },
+    {
+      "epoch": 0.48775055679287305,
+      "grad_norm": 0.2931824943444532,
+      "learning_rate": 4.657590759075908e-05,
+      "loss": 0.457,
+      "step": 219
+    },
+    {
+      "epoch": 0.48997772828507796,
+      "grad_norm": 0.3429250020790647,
+      "learning_rate": 4.653465346534654e-05,
+      "loss": 0.458,
+      "step": 220
+    },
+    {
+      "epoch": 0.4922048997772829,
+      "grad_norm": 0.26116764626929456,
+      "learning_rate": 4.6493399339933995e-05,
+      "loss": 0.4531,
+      "step": 221
+    },
+    {
+      "epoch": 0.49443207126948774,
+      "grad_norm": 0.33209861786607725,
+      "learning_rate": 4.645214521452145e-05,
+      "loss": 0.4684,
+      "step": 222
+    },
+    {
+      "epoch": 0.49665924276169265,
+      "grad_norm": 0.29594074682650495,
+      "learning_rate": 4.641089108910892e-05,
+      "loss": 0.4394,
+      "step": 223
+    },
+    {
+      "epoch": 0.49888641425389757,
+      "grad_norm": 0.31788744523283635,
+      "learning_rate": 4.636963696369637e-05,
+      "loss": 0.455,
+      "step": 224
+    },
+    {
+      "epoch": 0.5011135857461024,
+      "grad_norm": 0.29403815622831014,
+      "learning_rate": 4.632838283828383e-05,
+      "loss": 0.4617,
+      "step": 225
+    },
+    {
+      "epoch": 0.5033407572383074,
+      "grad_norm": 0.3246649957564367,
+      "learning_rate": 4.628712871287129e-05,
+      "loss": 0.4486,
+      "step": 226
+    },
+    {
+      "epoch": 0.5055679287305123,
+      "grad_norm": 0.2947521869215355,
+      "learning_rate": 4.624587458745875e-05,
+      "loss": 0.4522,
+      "step": 227
+    },
+    {
+      "epoch": 0.5077951002227171,
+      "grad_norm": 0.35319717032500775,
+      "learning_rate": 4.62046204620462e-05,
+      "loss": 0.4267,
+      "step": 228
+    },
+    {
+      "epoch": 0.5100222717149221,
+      "grad_norm": 0.26486969329235016,
+      "learning_rate": 4.616336633663367e-05,
+      "loss": 0.4454,
+      "step": 229
+    },
+    {
+      "epoch": 0.512249443207127,
+      "grad_norm": 0.28307917021023216,
+      "learning_rate": 4.6122112211221126e-05,
+      "loss": 0.4545,
+      "step": 230
+    },
+    {
+      "epoch": 0.5144766146993318,
+      "grad_norm": 0.29150478292360005,
+      "learning_rate": 4.6080858085808584e-05,
+      "loss": 0.4606,
+      "step": 231
+    },
+    {
+      "epoch": 0.5167037861915368,
+      "grad_norm": 0.29775322707947643,
+      "learning_rate": 4.603960396039604e-05,
+      "loss": 0.4486,
+      "step": 232
+    },
+    {
+      "epoch": 0.5189309576837416,
+      "grad_norm": 0.277863972612543,
+      "learning_rate": 4.59983498349835e-05,
+      "loss": 0.4572,
+      "step": 233
+    },
+    {
+      "epoch": 0.5211581291759465,
+      "grad_norm": 0.28925476269836853,
+      "learning_rate": 4.595709570957096e-05,
+      "loss": 0.4514,
+      "step": 234
+    },
+    {
+      "epoch": 0.5233853006681515,
+      "grad_norm": 0.2892961181023234,
+      "learning_rate": 4.591584158415842e-05,
+      "loss": 0.4419,
+      "step": 235
+    },
+    {
+      "epoch": 0.5256124721603563,
+      "grad_norm": 0.3156166066481885,
+      "learning_rate": 4.5874587458745876e-05,
+      "loss": 0.4886,
+      "step": 236
+    },
+    {
+      "epoch": 0.5278396436525612,
+      "grad_norm": 0.3083820360287261,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.4399,
+      "step": 237
+    },
+    {
+      "epoch": 0.5300668151447662,
+      "grad_norm": 0.2999984154669874,
+      "learning_rate": 4.57920792079208e-05,
+      "loss": 0.4343,
+      "step": 238
+    },
+    {
+      "epoch": 0.532293986636971,
+      "grad_norm": 0.2948137814745657,
+      "learning_rate": 4.575082508250825e-05,
+      "loss": 0.4396,
+      "step": 239
+    },
+    {
+      "epoch": 0.534521158129176,
+      "grad_norm": 0.2867836716684221,
+      "learning_rate": 4.570957095709571e-05,
+      "loss": 0.4444,
+      "step": 240
+    },
+    {
+      "epoch": 0.5367483296213809,
+      "grad_norm": 0.2921803786051076,
+      "learning_rate": 4.5668316831683174e-05,
+      "loss": 0.4569,
+      "step": 241
+    },
+    {
+      "epoch": 0.5389755011135857,
+      "grad_norm": 0.2667402553399932,
+      "learning_rate": 4.562706270627063e-05,
+      "loss": 0.4478,
+      "step": 242
+    },
+    {
+      "epoch": 0.5412026726057907,
+      "grad_norm": 0.2714986926455618,
+      "learning_rate": 4.558580858085808e-05,
+      "loss": 0.4201,
+      "step": 243
+    },
+    {
+      "epoch": 0.5434298440979956,
+      "grad_norm": 0.3010768527952275,
+      "learning_rate": 4.554455445544555e-05,
+      "loss": 0.4422,
+      "step": 244
+    },
+    {
+      "epoch": 0.5456570155902004,
+      "grad_norm": 0.24424723334447315,
+      "learning_rate": 4.550330033003301e-05,
+      "loss": 0.4689,
+      "step": 245
+    },
+    {
+      "epoch": 0.5478841870824054,
+      "grad_norm": 0.2824641561727399,
+      "learning_rate": 4.5462046204620465e-05,
+      "loss": 0.4577,
+      "step": 246
+    },
+    {
+      "epoch": 0.5501113585746102,
+      "grad_norm": 0.262448982344736,
+      "learning_rate": 4.542079207920792e-05,
+      "loss": 0.4411,
+      "step": 247
+    },
+    {
+      "epoch": 0.5523385300668151,
+      "grad_norm": 0.260954523917445,
+      "learning_rate": 4.537953795379538e-05,
+      "loss": 0.4461,
+      "step": 248
+    },
+    {
+      "epoch": 0.5545657015590201,
+      "grad_norm": 0.25493119956939947,
+      "learning_rate": 4.533828382838284e-05,
+      "loss": 0.4457,
+      "step": 249
+    },
+    {
+      "epoch": 0.5567928730512249,
+      "grad_norm": 0.24654376052164162,
+      "learning_rate": 4.52970297029703e-05,
+      "loss": 0.4644,
+      "step": 250
+    },
+    {
+      "epoch": 0.5590200445434298,
+      "grad_norm": 0.2839190109435608,
+      "learning_rate": 4.5255775577557756e-05,
+      "loss": 0.4706,
+      "step": 251
+    },
+    {
+      "epoch": 0.5612472160356348,
+      "grad_norm": 0.26041721433507015,
+      "learning_rate": 4.5214521452145214e-05,
+      "loss": 0.4462,
+      "step": 252
+    },
+    {
+      "epoch": 0.5634743875278396,
+      "grad_norm": 0.2720837819840524,
+      "learning_rate": 4.517326732673268e-05,
+      "loss": 0.4256,
+      "step": 253
+    },
+    {
+      "epoch": 0.5657015590200446,
+      "grad_norm": 0.31459213297941446,
+      "learning_rate": 4.513201320132013e-05,
+      "loss": 0.4423,
+      "step": 254
+    },
+    {
+      "epoch": 0.5679287305122495,
+      "grad_norm": 0.26370544862803846,
+      "learning_rate": 4.509075907590759e-05,
+      "loss": 0.4478,
+      "step": 255
+    },
+    {
+      "epoch": 0.5701559020044543,
+      "grad_norm": 0.2954224532182178,
+      "learning_rate": 4.5049504950495054e-05,
+      "loss": 0.4482,
+      "step": 256
+    },
+    {
+      "epoch": 0.5723830734966593,
+      "grad_norm": 0.24297839932464907,
+      "learning_rate": 4.500825082508251e-05,
+      "loss": 0.4467,
+      "step": 257
+    },
+    {
+      "epoch": 0.5746102449888641,
+      "grad_norm": 0.30732074611692045,
+      "learning_rate": 4.4966996699669964e-05,
+      "loss": 0.4569,
+      "step": 258
+    },
+    {
+      "epoch": 0.576837416481069,
+      "grad_norm": 0.2561720549211904,
+      "learning_rate": 4.492574257425743e-05,
+      "loss": 0.4473,
+      "step": 259
+    },
+    {
+      "epoch": 0.579064587973274,
+      "grad_norm": 0.2538371668472489,
+      "learning_rate": 4.488448844884489e-05,
+      "loss": 0.4526,
+      "step": 260
+    },
+    {
+      "epoch": 0.5812917594654788,
+      "grad_norm": 0.3092696193325629,
+      "learning_rate": 4.4843234323432346e-05,
+      "loss": 0.4764,
+      "step": 261
+    },
+    {
+      "epoch": 0.5835189309576837,
+      "grad_norm": 0.27859232038495874,
+      "learning_rate": 4.4801980198019804e-05,
+      "loss": 0.452,
+      "step": 262
+    },
+    {
+      "epoch": 0.5857461024498887,
+      "grad_norm": 0.3200460894107544,
+      "learning_rate": 4.476072607260726e-05,
+      "loss": 0.4291,
+      "step": 263
+    },
+    {
+      "epoch": 0.5879732739420935,
+      "grad_norm": 0.32931430711122667,
+      "learning_rate": 4.471947194719473e-05,
+      "loss": 0.4582,
+      "step": 264
+    },
+    {
+      "epoch": 0.5902004454342984,
+      "grad_norm": 0.2731605321453466,
+      "learning_rate": 4.467821782178218e-05,
+      "loss": 0.4363,
+      "step": 265
+    },
+    {
+      "epoch": 0.5924276169265034,
+      "grad_norm": 0.2703465656719884,
+      "learning_rate": 4.463696369636964e-05,
+      "loss": 0.4347,
+      "step": 266
+    },
+    {
+      "epoch": 0.5946547884187082,
+      "grad_norm": 0.3626930341938144,
+      "learning_rate": 4.45957095709571e-05,
+      "loss": 0.4425,
+      "step": 267
+    },
+    {
+      "epoch": 0.5968819599109132,
+      "grad_norm": 0.2646653041944948,
+      "learning_rate": 4.455445544554456e-05,
+      "loss": 0.4289,
+      "step": 268
+    },
+    {
+      "epoch": 0.5991091314031181,
+      "grad_norm": 0.29847165588799424,
+      "learning_rate": 4.451320132013201e-05,
+      "loss": 0.4366,
+      "step": 269
+    },
+    {
+      "epoch": 0.6013363028953229,
+      "grad_norm": 0.26851259081432843,
+      "learning_rate": 4.447194719471948e-05,
+      "loss": 0.4703,
+      "step": 270
+    },
+    {
+      "epoch": 0.6035634743875279,
+      "grad_norm": 0.2987098198875876,
+      "learning_rate": 4.4430693069306935e-05,
+      "loss": 0.445,
+      "step": 271
+    },
+    {
+      "epoch": 0.6057906458797327,
+      "grad_norm": 0.2605905093946033,
+      "learning_rate": 4.438943894389439e-05,
+      "loss": 0.4357,
+      "step": 272
+    },
+    {
+      "epoch": 0.6080178173719376,
+      "grad_norm": 0.24844971421080053,
+      "learning_rate": 4.434818481848185e-05,
+      "loss": 0.419,
+      "step": 273
+    },
+    {
+      "epoch": 0.6102449888641426,
+      "grad_norm": 0.3290220226694587,
+      "learning_rate": 4.430693069306931e-05,
+      "loss": 0.4465,
+      "step": 274
+    },
+    {
+      "epoch": 0.6124721603563474,
+      "grad_norm": 0.28480796997833574,
+      "learning_rate": 4.426567656765677e-05,
+      "loss": 0.4441,
+      "step": 275
+    },
+    {
+      "epoch": 0.6146993318485523,
+      "grad_norm": 0.2976483417452742,
+      "learning_rate": 4.4224422442244226e-05,
+      "loss": 0.438,
+      "step": 276
+    },
+    {
+      "epoch": 0.6169265033407573,
+      "grad_norm": 0.32736870022235975,
+      "learning_rate": 4.4183168316831684e-05,
+      "loss": 0.4614,
+      "step": 277
+    },
+    {
+      "epoch": 0.6191536748329621,
+      "grad_norm": 0.26160786488834764,
+      "learning_rate": 4.414191419141914e-05,
+      "loss": 0.4366,
+      "step": 278
+    },
+    {
+      "epoch": 0.621380846325167,
+      "grad_norm": 0.33852355498895254,
+      "learning_rate": 4.410066006600661e-05,
+      "loss": 0.4484,
+      "step": 279
+    },
+    {
+      "epoch": 0.623608017817372,
+      "grad_norm": 0.28362171544376097,
+      "learning_rate": 4.405940594059406e-05,
+      "loss": 0.4381,
+      "step": 280
+    },
+    {
+      "epoch": 0.6258351893095768,
+      "grad_norm": 0.2811074334233704,
+      "learning_rate": 4.401815181518152e-05,
+      "loss": 0.4346,
+      "step": 281
+    },
+    {
+      "epoch": 0.6280623608017817,
+      "grad_norm": 0.3184745078431333,
+      "learning_rate": 4.397689768976898e-05,
+      "loss": 0.4457,
+      "step": 282
+    },
+    {
+      "epoch": 0.6302895322939867,
+      "grad_norm": 0.3008010945284472,
+      "learning_rate": 4.393564356435644e-05,
+      "loss": 0.4511,
+      "step": 283
+    },
+    {
+      "epoch": 0.6325167037861915,
+      "grad_norm": 0.32595315041730505,
+      "learning_rate": 4.389438943894389e-05,
+      "loss": 0.4565,
+      "step": 284
+    },
+    {
+      "epoch": 0.6347438752783965,
+      "grad_norm": 0.242270649578824,
+      "learning_rate": 4.385313531353136e-05,
+      "loss": 0.4247,
+      "step": 285
+    },
+    {
+      "epoch": 0.6369710467706013,
+      "grad_norm": 0.3304882530212611,
+      "learning_rate": 4.3811881188118816e-05,
+      "loss": 0.4459,
+      "step": 286
+    },
+    {
+      "epoch": 0.6391982182628062,
+      "grad_norm": 0.2749998460242237,
+      "learning_rate": 4.3770627062706274e-05,
+      "loss": 0.4725,
+      "step": 287
+    },
+    {
+      "epoch": 0.6414253897550112,
+      "grad_norm": 0.3051711298096592,
+      "learning_rate": 4.372937293729373e-05,
+      "loss": 0.4822,
+      "step": 288
+    },
+    {
+      "epoch": 0.643652561247216,
+      "grad_norm": 0.2742029836047166,
+      "learning_rate": 4.368811881188119e-05,
+      "loss": 0.4497,
+      "step": 289
+    },
+    {
+      "epoch": 0.6458797327394209,
+      "grad_norm": 0.2617353582124388,
+      "learning_rate": 4.364686468646865e-05,
+      "loss": 0.4467,
+      "step": 290
+    },
+    {
+      "epoch": 0.6481069042316259,
+      "grad_norm": 0.2737738804183133,
+      "learning_rate": 4.360561056105611e-05,
+      "loss": 0.4561,
+      "step": 291
+    },
+    {
+      "epoch": 0.6503340757238307,
+      "grad_norm": 0.3091759816395053,
+      "learning_rate": 4.3564356435643565e-05,
+      "loss": 0.4736,
+      "step": 292
+    },
+    {
+      "epoch": 0.6525612472160356,
+      "grad_norm": 0.25626886901062795,
+      "learning_rate": 4.352310231023102e-05,
+      "loss": 0.4311,
+      "step": 293
+    },
+    {
+      "epoch": 0.6547884187082406,
+      "grad_norm": 0.26935906361510203,
+      "learning_rate": 4.348184818481849e-05,
+      "loss": 0.4391,
+      "step": 294
+    },
+    {
+      "epoch": 0.6570155902004454,
+      "grad_norm": 0.29389907195879933,
+      "learning_rate": 4.344059405940594e-05,
+      "loss": 0.4355,
+      "step": 295
+    },
+    {
+      "epoch": 0.6592427616926503,
+      "grad_norm": 0.27782045565159375,
+      "learning_rate": 4.33993399339934e-05,
+      "loss": 0.4432,
+      "step": 296
+    },
+    {
+      "epoch": 0.6614699331848553,
+      "grad_norm": 0.2889974411301799,
+      "learning_rate": 4.335808580858086e-05,
+      "loss": 0.4474,
+      "step": 297
+    },
+    {
+      "epoch": 0.6636971046770601,
+      "grad_norm": 0.271845863873609,
+      "learning_rate": 4.331683168316832e-05,
+      "loss": 0.439,
+      "step": 298
+    },
+    {
+      "epoch": 0.6659242761692651,
+      "grad_norm": 0.29771387683903633,
+      "learning_rate": 4.327557755775577e-05,
+      "loss": 0.4506,
+      "step": 299
+    },
+    {
+      "epoch": 0.6681514476614699,
+      "grad_norm": 0.27637974584922304,
+      "learning_rate": 4.323432343234324e-05,
+      "loss": 0.4532,
+      "step": 300
+    },
+    {
+      "epoch": 0.6703786191536748,
+      "grad_norm": 0.26229222544704045,
+      "learning_rate": 4.3193069306930696e-05,
+      "loss": 0.4366,
+      "step": 301
+    },
+    {
+      "epoch": 0.6726057906458798,
+      "grad_norm": 0.2735519792708174,
+      "learning_rate": 4.3151815181518154e-05,
+      "loss": 0.4473,
+      "step": 302
+    },
+    {
+      "epoch": 0.6748329621380846,
+      "grad_norm": 0.2791681791157702,
+      "learning_rate": 4.311056105610561e-05,
+      "loss": 0.4618,
+      "step": 303
+    },
+    {
+      "epoch": 0.6770601336302895,
+      "grad_norm": 0.2552442658275953,
+      "learning_rate": 4.306930693069307e-05,
+      "loss": 0.454,
+      "step": 304
+    },
+    {
+      "epoch": 0.6792873051224945,
+      "grad_norm": 0.2911430189507063,
+      "learning_rate": 4.302805280528053e-05,
+      "loss": 0.4266,
+      "step": 305
+    },
+    {
+      "epoch": 0.6815144766146993,
+      "grad_norm": 0.27188666629471603,
+      "learning_rate": 4.298679867986799e-05,
+      "loss": 0.4451,
+      "step": 306
+    },
+    {
+      "epoch": 0.6837416481069042,
+      "grad_norm": 0.28050623589469015,
+      "learning_rate": 4.2945544554455446e-05,
+      "loss": 0.4365,
+      "step": 307
+    },
+    {
+      "epoch": 0.6859688195991092,
+      "grad_norm": 0.2935564501685569,
+      "learning_rate": 4.2904290429042904e-05,
+      "loss": 0.4591,
+      "step": 308
+    },
+    {
+      "epoch": 0.688195991091314,
+      "grad_norm": 0.24920342109951324,
+      "learning_rate": 4.286303630363037e-05,
+      "loss": 0.4188,
+      "step": 309
+    },
+    {
+      "epoch": 0.6904231625835189,
+      "grad_norm": 0.25293406559071957,
+      "learning_rate": 4.282178217821782e-05,
+      "loss": 0.4243,
+      "step": 310
+    },
+    {
+      "epoch": 0.6926503340757239,
+      "grad_norm": 0.2614798712738581,
+      "learning_rate": 4.278052805280528e-05,
+      "loss": 0.446,
+      "step": 311
+    },
+    {
+      "epoch": 0.6948775055679287,
+      "grad_norm": 0.30606992019945795,
+      "learning_rate": 4.2739273927392744e-05,
+      "loss": 0.4422,
+      "step": 312
+    },
+    {
+      "epoch": 0.6971046770601337,
+      "grad_norm": 0.24582221104090354,
+      "learning_rate": 4.26980198019802e-05,
+      "loss": 0.4727,
+      "step": 313
+    },
+    {
+      "epoch": 0.6993318485523385,
+      "grad_norm": 0.2706816252510679,
+      "learning_rate": 4.265676567656766e-05,
+      "loss": 0.4467,
+      "step": 314
+    },
+    {
+      "epoch": 0.7015590200445434,
+      "grad_norm": 0.25757445350913805,
+      "learning_rate": 4.261551155115512e-05,
+      "loss": 0.4312,
+      "step": 315
+    },
+    {
+      "epoch": 0.7037861915367484,
+      "grad_norm": 0.24832914827025765,
+      "learning_rate": 4.257425742574258e-05,
+      "loss": 0.4227,
+      "step": 316
+    },
+    {
+      "epoch": 0.7060133630289532,
+      "grad_norm": 0.26987925485376546,
+      "learning_rate": 4.2533003300330035e-05,
+      "loss": 0.4598,
+      "step": 317
+    },
+    {
+      "epoch": 0.7082405345211581,
+      "grad_norm": 0.27766968317320195,
+      "learning_rate": 4.249174917491749e-05,
+      "loss": 0.4523,
+      "step": 318
+    },
+    {
+      "epoch": 0.7104677060133631,
+      "grad_norm": 0.26095987732230747,
+      "learning_rate": 4.245049504950495e-05,
+      "loss": 0.4319,
+      "step": 319
+    },
+    {
+      "epoch": 0.7126948775055679,
+      "grad_norm": 0.26160766384381995,
+      "learning_rate": 4.240924092409242e-05,
+      "loss": 0.4412,
+      "step": 320
+    },
+    {
+      "epoch": 0.7149220489977728,
+      "grad_norm": 0.2584570874357079,
+      "learning_rate": 4.236798679867987e-05,
+      "loss": 0.4551,
+      "step": 321
+    },
+    {
+      "epoch": 0.7171492204899778,
+      "grad_norm": 0.2834159008077785,
+      "learning_rate": 4.2326732673267326e-05,
+      "loss": 0.4431,
+      "step": 322
+    },
+    {
+      "epoch": 0.7193763919821826,
+      "grad_norm": 0.24898946543574163,
+      "learning_rate": 4.228547854785479e-05,
+      "loss": 0.4302,
+      "step": 323
+    },
+    {
+      "epoch": 0.7216035634743875,
+      "grad_norm": 0.28860867984847915,
+      "learning_rate": 4.224422442244225e-05,
+      "loss": 0.4475,
+      "step": 324
+    },
+    {
+      "epoch": 0.7238307349665924,
+      "grad_norm": 0.23833559702503349,
+      "learning_rate": 4.22029702970297e-05,
+      "loss": 0.4716,
+      "step": 325
+    },
+    {
+      "epoch": 0.7260579064587973,
+      "grad_norm": 0.2763178033720794,
+      "learning_rate": 4.2161716171617166e-05,
+      "loss": 0.424,
+      "step": 326
+    },
+    {
+      "epoch": 0.7282850779510023,
+      "grad_norm": 0.2402663207112546,
+      "learning_rate": 4.2120462046204624e-05,
+      "loss": 0.4373,
+      "step": 327
+    },
+    {
+      "epoch": 0.7305122494432071,
+      "grad_norm": 0.2714560047859496,
+      "learning_rate": 4.207920792079208e-05,
+      "loss": 0.434,
+      "step": 328
+    },
+    {
+      "epoch": 0.732739420935412,
+      "grad_norm": 0.22960996560195104,
+      "learning_rate": 4.203795379537954e-05,
+      "loss": 0.4173,
+      "step": 329
+    },
+    {
+      "epoch": 0.734966592427617,
+      "grad_norm": 0.30360671082013174,
+      "learning_rate": 4.1996699669967e-05,
+      "loss": 0.4484,
+      "step": 330
+    },
+    {
+      "epoch": 0.7371937639198218,
+      "grad_norm": 0.24679972364853342,
+      "learning_rate": 4.195544554455446e-05,
+      "loss": 0.4522,
+      "step": 331
+    },
+    {
+      "epoch": 0.7394209354120267,
+      "grad_norm": 0.3047224480559927,
+      "learning_rate": 4.1914191419141916e-05,
+      "loss": 0.4326,
+      "step": 332
+    },
+    {
+      "epoch": 0.7416481069042317,
+      "grad_norm": 0.2546066095218112,
+      "learning_rate": 4.1872937293729374e-05,
+      "loss": 0.4498,
+      "step": 333
+    },
+    {
+      "epoch": 0.7438752783964365,
+      "grad_norm": 0.25721504775664134,
+      "learning_rate": 4.183168316831683e-05,
+      "loss": 0.431,
+      "step": 334
+    },
+    {
+      "epoch": 0.7461024498886414,
+      "grad_norm": 0.2691894404395012,
+      "learning_rate": 4.17904290429043e-05,
+      "loss": 0.4387,
+      "step": 335
+    },
+    {
+      "epoch": 0.7483296213808464,
+      "grad_norm": 0.23271965303359804,
+      "learning_rate": 4.174917491749175e-05,
+      "loss": 0.4399,
+      "step": 336
+    },
+    {
+      "epoch": 0.7505567928730512,
+      "grad_norm": 0.26493570894435625,
+      "learning_rate": 4.170792079207921e-05,
+      "loss": 0.4459,
+      "step": 337
+    },
+    {
+      "epoch": 0.7527839643652561,
+      "grad_norm": 2.0248894289396824,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.474,
+      "step": 338
+    },
+    {
+      "epoch": 0.755011135857461,
+      "grad_norm": 0.3062099581698654,
+      "learning_rate": 4.162541254125413e-05,
+      "loss": 0.4358,
+      "step": 339
+    },
+    {
+      "epoch": 0.7572383073496659,
+      "grad_norm": 0.25946848706627174,
+      "learning_rate": 4.158415841584158e-05,
+      "loss": 0.4233,
+      "step": 340
+    },
+    {
+      "epoch": 0.7594654788418709,
+      "grad_norm": 0.25713280085130474,
+      "learning_rate": 4.154290429042905e-05,
+      "loss": 0.4439,
+      "step": 341
+    },
+    {
+      "epoch": 0.7616926503340757,
+      "grad_norm": 0.3167572006113162,
+      "learning_rate": 4.1501650165016505e-05,
+      "loss": 0.4436,
+      "step": 342
+    },
+    {
+      "epoch": 0.7639198218262806,
+      "grad_norm": 0.26048288751304155,
+      "learning_rate": 4.146039603960396e-05,
+      "loss": 0.417,
+      "step": 343
+    },
+    {
+      "epoch": 0.7661469933184856,
+      "grad_norm": 0.25497473878005894,
+      "learning_rate": 4.141914191419142e-05,
+      "loss": 0.4557,
+      "step": 344
+    },
+    {
+      "epoch": 0.7683741648106904,
+      "grad_norm": 0.2911566218466426,
+      "learning_rate": 4.137788778877888e-05,
+      "loss": 0.4404,
+      "step": 345
+    },
+    {
+      "epoch": 0.7706013363028953,
+      "grad_norm": 0.25971342414899073,
+      "learning_rate": 4.133663366336634e-05,
+      "loss": 0.4318,
+      "step": 346
+    },
+    {
+      "epoch": 0.7728285077951003,
+      "grad_norm": 0.3151547768170685,
+      "learning_rate": 4.1295379537953796e-05,
+      "loss": 0.4105,
+      "step": 347
+    },
+    {
+      "epoch": 0.7750556792873051,
+      "grad_norm": 0.2738918273042278,
+      "learning_rate": 4.1254125412541255e-05,
+      "loss": 0.4436,
+      "step": 348
+    },
+    {
+      "epoch": 0.77728285077951,
+      "grad_norm": 0.296938490357881,
+      "learning_rate": 4.121287128712871e-05,
+      "loss": 0.4448,
+      "step": 349
+    },
+    {
+      "epoch": 0.779510022271715,
+      "grad_norm": 0.27038716684654934,
+      "learning_rate": 4.117161716171618e-05,
+      "loss": 0.4454,
+      "step": 350
+    },
+    {
+      "epoch": 0.7817371937639198,
+      "grad_norm": 0.3227956568011175,
+      "learning_rate": 4.113036303630363e-05,
+      "loss": 0.4406,
+      "step": 351
+    },
+    {
+      "epoch": 0.7839643652561247,
+      "grad_norm": 0.26633298229761343,
+      "learning_rate": 4.108910891089109e-05,
+      "loss": 0.4376,
+      "step": 352
+    },
+    {
+      "epoch": 0.7861915367483296,
+      "grad_norm": 0.28347495442610243,
+      "learning_rate": 4.104785478547855e-05,
+      "loss": 0.4337,
+      "step": 353
+    },
+    {
+      "epoch": 0.7884187082405345,
+      "grad_norm": 0.30182784687239844,
+      "learning_rate": 4.100660066006601e-05,
+      "loss": 0.4606,
+      "step": 354
+    },
+    {
+      "epoch": 0.7906458797327395,
+      "grad_norm": 0.2708234094108942,
+      "learning_rate": 4.096534653465346e-05,
+      "loss": 0.4503,
+      "step": 355
+    },
+    {
+      "epoch": 0.7928730512249443,
+      "grad_norm": 0.2606877113787521,
+      "learning_rate": 4.092409240924093e-05,
+      "loss": 0.4627,
+      "step": 356
+    },
+    {
+      "epoch": 0.7951002227171492,
+      "grad_norm": 0.255498912261423,
+      "learning_rate": 4.0882838283828386e-05,
+      "loss": 0.4494,
+      "step": 357
+    },
+    {
+      "epoch": 0.7973273942093542,
+      "grad_norm": 0.2634192898727719,
+      "learning_rate": 4.0841584158415844e-05,
+      "loss": 0.4378,
+      "step": 358
+    },
+    {
+      "epoch": 0.799554565701559,
+      "grad_norm": 0.2455180865698325,
+      "learning_rate": 4.08003300330033e-05,
+      "loss": 0.4425,
+      "step": 359
+    },
+    {
+      "epoch": 0.8017817371937639,
+      "grad_norm": 0.29057066501650736,
+      "learning_rate": 4.075907590759076e-05,
+      "loss": 0.4504,
+      "step": 360
+    },
+    {
+      "epoch": 0.8040089086859689,
+      "grad_norm": 0.24841510932266841,
+      "learning_rate": 4.071782178217822e-05,
+      "loss": 0.4308,
+      "step": 361
+    },
+    {
+      "epoch": 0.8062360801781737,
+      "grad_norm": 0.30437614999300794,
+      "learning_rate": 4.067656765676568e-05,
+      "loss": 0.4225,
+      "step": 362
+    },
+    {
+      "epoch": 0.8084632516703786,
+      "grad_norm": 0.24654821833407398,
+      "learning_rate": 4.0635313531353135e-05,
+      "loss": 0.4468,
+      "step": 363
+    },
+    {
+      "epoch": 0.8106904231625836,
+      "grad_norm": 0.34206382856361406,
+      "learning_rate": 4.05940594059406e-05,
+      "loss": 0.4238,
+      "step": 364
+    },
+    {
+      "epoch": 0.8129175946547884,
+      "grad_norm": 0.23051720605393852,
+      "learning_rate": 4.055280528052806e-05,
+      "loss": 0.4302,
+      "step": 365
+    },
+    {
+      "epoch": 0.8151447661469933,
+      "grad_norm": 0.30090350911289054,
+      "learning_rate": 4.051155115511551e-05,
+      "loss": 0.4422,
+      "step": 366
+    },
+    {
+      "epoch": 0.8173719376391982,
+      "grad_norm": 0.24573950009262582,
+      "learning_rate": 4.0470297029702975e-05,
+      "loss": 0.4306,
+      "step": 367
+    },
+    {
+      "epoch": 0.8195991091314031,
+      "grad_norm": 0.2882877236903447,
+      "learning_rate": 4.042904290429043e-05,
+      "loss": 0.4652,
+      "step": 368
+    },
+    {
+      "epoch": 0.821826280623608,
+      "grad_norm": 0.24988067797973895,
+      "learning_rate": 4.038778877887789e-05,
+      "loss": 0.4362,
+      "step": 369
+    },
+    {
+      "epoch": 0.8240534521158129,
+      "grad_norm": 0.29408476035049613,
+      "learning_rate": 4.034653465346535e-05,
+      "loss": 0.4435,
+      "step": 370
+    },
+    {
+      "epoch": 0.8262806236080178,
+      "grad_norm": 0.2644363301471362,
+      "learning_rate": 4.030528052805281e-05,
+      "loss": 0.4344,
+      "step": 371
+    },
+    {
+      "epoch": 0.8285077951002228,
+      "grad_norm": 0.29493993411578473,
+      "learning_rate": 4.0264026402640266e-05,
+      "loss": 0.434,
+      "step": 372
+    },
+    {
+      "epoch": 0.8307349665924276,
+      "grad_norm": 0.3468561021222359,
+      "learning_rate": 4.0222772277227725e-05,
+      "loss": 0.4384,
+      "step": 373
+    },
+    {
+      "epoch": 0.8329621380846325,
+      "grad_norm": 0.29607221940526673,
+      "learning_rate": 4.018151815181518e-05,
+      "loss": 0.4628,
+      "step": 374
+    },
+    {
+      "epoch": 0.8351893095768375,
+      "grad_norm": 0.3637525910818386,
+      "learning_rate": 4.014026402640264e-05,
+      "loss": 0.4415,
+      "step": 375
+    },
+    {
+      "epoch": 0.8374164810690423,
+      "grad_norm": 0.28714265367455805,
+      "learning_rate": 4.0099009900990106e-05,
+      "loss": 0.4376,
+      "step": 376
+    },
+    {
+      "epoch": 0.8396436525612472,
+      "grad_norm": 0.32629775286679424,
+      "learning_rate": 4.005775577557756e-05,
+      "loss": 0.4235,
+      "step": 377
+    },
+    {
+      "epoch": 0.8418708240534521,
+      "grad_norm": 0.2802612112542595,
+      "learning_rate": 4.0016501650165016e-05,
+      "loss": 0.4317,
+      "step": 378
+    },
+    {
+      "epoch": 0.844097995545657,
+      "grad_norm": 0.2918531735740327,
+      "learning_rate": 3.997524752475248e-05,
+      "loss": 0.452,
+      "step": 379
+    },
+    {
+      "epoch": 0.8463251670378619,
+      "grad_norm": 0.3352571650408221,
+      "learning_rate": 3.993399339933994e-05,
+      "loss": 0.4386,
+      "step": 380
+    },
+    {
+      "epoch": 0.8485523385300668,
+      "grad_norm": 0.2573244398349893,
+      "learning_rate": 3.989273927392739e-05,
+      "loss": 0.4574,
+      "step": 381
+    },
+    {
+      "epoch": 0.8507795100222717,
+      "grad_norm": 0.3027471516866268,
+      "learning_rate": 3.9851485148514856e-05,
+      "loss": 0.4419,
+      "step": 382
+    },
+    {
+      "epoch": 0.8530066815144766,
+      "grad_norm": 0.2854937531018722,
+      "learning_rate": 3.9810231023102314e-05,
+      "loss": 0.4584,
+      "step": 383
+    },
+    {
+      "epoch": 0.8552338530066815,
+      "grad_norm": 0.24589201590337514,
+      "learning_rate": 3.976897689768977e-05,
+      "loss": 0.4193,
+      "step": 384
+    },
+    {
+      "epoch": 0.8574610244988864,
+      "grad_norm": 0.2778137032428908,
+      "learning_rate": 3.972772277227723e-05,
+      "loss": 0.4343,
+      "step": 385
+    },
+    {
+      "epoch": 0.8596881959910914,
+      "grad_norm": 0.3342562442904411,
+      "learning_rate": 3.968646864686469e-05,
+      "loss": 0.4326,
+      "step": 386
+    },
+    {
+      "epoch": 0.8619153674832962,
+      "grad_norm": 0.24896713048447036,
+      "learning_rate": 3.964521452145215e-05,
+      "loss": 0.432,
+      "step": 387
+    },
+    {
+      "epoch": 0.8641425389755011,
+      "grad_norm": 0.29405006865231625,
+      "learning_rate": 3.9603960396039605e-05,
+      "loss": 0.433,
+      "step": 388
+    },
+    {
+      "epoch": 0.8663697104677061,
+      "grad_norm": 0.23757029182896722,
+      "learning_rate": 3.9562706270627064e-05,
+      "loss": 0.4418,
+      "step": 389
+    },
+    {
+      "epoch": 0.8685968819599109,
+      "grad_norm": 0.2724752861847499,
+      "learning_rate": 3.952145214521452e-05,
+      "loss": 0.422,
+      "step": 390
+    },
+    {
+      "epoch": 0.8708240534521158,
+      "grad_norm": 0.264389356209867,
+      "learning_rate": 3.948019801980199e-05,
+      "loss": 0.4247,
+      "step": 391
+    },
+    {
+      "epoch": 0.8730512249443207,
+      "grad_norm": 0.27027284504631205,
+      "learning_rate": 3.943894389438944e-05,
+      "loss": 0.4307,
+      "step": 392
+    },
+    {
+      "epoch": 0.8752783964365256,
+      "grad_norm": 0.266772697377517,
+      "learning_rate": 3.9397689768976897e-05,
+      "loss": 0.4309,
+      "step": 393
+    },
+    {
+      "epoch": 0.8775055679287305,
+      "grad_norm": 0.2614972736062481,
+      "learning_rate": 3.935643564356436e-05,
+      "loss": 0.4187,
+      "step": 394
+    },
+    {
+      "epoch": 0.8797327394209354,
+      "grad_norm": 0.23777796416631325,
+      "learning_rate": 3.931518151815182e-05,
+      "loss": 0.4303,
+      "step": 395
+    },
+    {
+      "epoch": 0.8819599109131403,
+      "grad_norm": 0.25876129593586256,
+      "learning_rate": 3.927392739273927e-05,
+      "loss": 0.4277,
+      "step": 396
+    },
+    {
+      "epoch": 0.8841870824053452,
+      "grad_norm": 0.24533319198985845,
+      "learning_rate": 3.9232673267326736e-05,
+      "loss": 0.4316,
+      "step": 397
+    },
+    {
+      "epoch": 0.8864142538975501,
+      "grad_norm": 0.2827282102508032,
+      "learning_rate": 3.9191419141914195e-05,
+      "loss": 0.4374,
+      "step": 398
+    },
+    {
+      "epoch": 0.888641425389755,
+      "grad_norm": 0.25672028235670863,
+      "learning_rate": 3.915016501650165e-05,
+      "loss": 0.4334,
+      "step": 399
+    },
+    {
+      "epoch": 0.89086859688196,
+      "grad_norm": 0.2964434942824687,
+      "learning_rate": 3.910891089108911e-05,
+      "loss": 0.4382,
+      "step": 400
+    },
+    {
+      "epoch": 0.8930957683741648,
+      "grad_norm": 0.2992038368053277,
+      "learning_rate": 3.906765676567657e-05,
+      "loss": 0.4506,
+      "step": 401
+    },
+    {
+      "epoch": 0.8953229398663697,
+      "grad_norm": 0.2824229554017287,
+      "learning_rate": 3.902640264026403e-05,
+      "loss": 0.4292,
+      "step": 402
+    },
+    {
+      "epoch": 0.8975501113585747,
+      "grad_norm": 0.3021114688794221,
+      "learning_rate": 3.8985148514851486e-05,
+      "loss": 0.4159,
+      "step": 403
+    },
+    {
+      "epoch": 0.8997772828507795,
+      "grad_norm": 0.2670836799841371,
+      "learning_rate": 3.8943894389438944e-05,
+      "loss": 0.4167,
+      "step": 404
+    },
+    {
+      "epoch": 0.9020044543429844,
+      "grad_norm": 0.2751414532574633,
+      "learning_rate": 3.89026402640264e-05,
+      "loss": 0.4216,
+      "step": 405
+    },
+    {
+      "epoch": 0.9042316258351893,
+      "grad_norm": 0.33093586819706966,
+      "learning_rate": 3.886138613861387e-05,
+      "loss": 0.4221,
+      "step": 406
+    },
+    {
+      "epoch": 0.9064587973273942,
+      "grad_norm": 0.28660771966605225,
+      "learning_rate": 3.882013201320132e-05,
+      "loss": 0.4328,
+      "step": 407
+    },
+    {
+      "epoch": 0.9086859688195991,
+      "grad_norm": 0.31729231099616456,
+      "learning_rate": 3.877887788778878e-05,
+      "loss": 0.42,
+      "step": 408
+    },
+    {
+      "epoch": 0.910913140311804,
+      "grad_norm": 0.28276830996460967,
+      "learning_rate": 3.873762376237624e-05,
+      "loss": 0.4315,
+      "step": 409
+    },
+    {
+      "epoch": 0.9131403118040089,
+      "grad_norm": 0.2637643372082265,
+      "learning_rate": 3.86963696369637e-05,
+      "loss": 0.4464,
+      "step": 410
+    },
+    {
+      "epoch": 0.9153674832962138,
+      "grad_norm": 0.29760756967577906,
+      "learning_rate": 3.865511551155115e-05,
+      "loss": 0.4309,
+      "step": 411
+    },
+    {
+      "epoch": 0.9175946547884187,
+      "grad_norm": 0.27094511138723293,
+      "learning_rate": 3.861386138613862e-05,
+      "loss": 0.4321,
+      "step": 412
+    },
+    {
+      "epoch": 0.9198218262806236,
+      "grad_norm": 0.2583422821283627,
+      "learning_rate": 3.8572607260726075e-05,
+      "loss": 0.4096,
+      "step": 413
+    },
+    {
+      "epoch": 0.9220489977728286,
+      "grad_norm": 0.29483389234975,
+      "learning_rate": 3.8531353135313534e-05,
+      "loss": 0.4352,
+      "step": 414
+    },
+    {
+      "epoch": 0.9242761692650334,
+      "grad_norm": 0.2760297601343316,
+      "learning_rate": 3.849009900990099e-05,
+      "loss": 0.4433,
+      "step": 415
+    },
+    {
+      "epoch": 0.9265033407572383,
+      "grad_norm": 0.2632096095730053,
+      "learning_rate": 3.844884488448845e-05,
+      "loss": 0.444,
+      "step": 416
+    },
+    {
+      "epoch": 0.9287305122494433,
+      "grad_norm": 0.26571838725109725,
+      "learning_rate": 3.8407590759075915e-05,
+      "loss": 0.4479,
+      "step": 417
+    },
+    {
+      "epoch": 0.9309576837416481,
+      "grad_norm": 0.2612223809833972,
+      "learning_rate": 3.8366336633663367e-05,
+      "loss": 0.43,
+      "step": 418
+    },
+    {
+      "epoch": 0.933184855233853,
+      "grad_norm": 0.20729020473362156,
+      "learning_rate": 3.8325082508250825e-05,
+      "loss": 0.4269,
+      "step": 419
+    },
+    {
+      "epoch": 0.9354120267260579,
+      "grad_norm": 0.30206416988432955,
+      "learning_rate": 3.828382838283829e-05,
+      "loss": 0.4401,
+      "step": 420
+    },
+    {
+      "epoch": 0.9376391982182628,
+      "grad_norm": 0.23463191817841658,
+      "learning_rate": 3.824257425742575e-05,
+      "loss": 0.4382,
+      "step": 421
+    },
+    {
+      "epoch": 0.9398663697104677,
+      "grad_norm": 0.2830697542352734,
+      "learning_rate": 3.82013201320132e-05,
+      "loss": 0.4599,
+      "step": 422
+    },
+    {
+      "epoch": 0.9420935412026726,
+      "grad_norm": 0.24602603533970543,
+      "learning_rate": 3.8160066006600665e-05,
+      "loss": 0.4176,
+      "step": 423
+    },
+    {
+      "epoch": 0.9443207126948775,
+      "grad_norm": 0.3337694316665918,
+      "learning_rate": 3.811881188118812e-05,
+      "loss": 0.4435,
+      "step": 424
+    },
+    {
+      "epoch": 0.9465478841870824,
+      "grad_norm": 0.23677648895454642,
+      "learning_rate": 3.807755775577558e-05,
+      "loss": 0.4368,
+      "step": 425
+    },
+    {
+      "epoch": 0.9487750556792873,
+      "grad_norm": 0.29638771606416087,
+      "learning_rate": 3.803630363036304e-05,
+      "loss": 0.4317,
+      "step": 426
+    },
+    {
+      "epoch": 0.9510022271714922,
+      "grad_norm": 0.24211776195738213,
+      "learning_rate": 3.79950495049505e-05,
+      "loss": 0.4274,
+      "step": 427
+    },
+    {
+      "epoch": 0.9532293986636972,
+      "grad_norm": 0.2246529620028757,
+      "learning_rate": 3.7953795379537956e-05,
+      "loss": 0.4191,
+      "step": 428
+    },
+    {
+      "epoch": 0.955456570155902,
+      "grad_norm": 0.25492756281023654,
+      "learning_rate": 3.7912541254125414e-05,
+      "loss": 0.4279,
+      "step": 429
+    },
+    {
+      "epoch": 0.9576837416481069,
+      "grad_norm": 0.26189081420987503,
+      "learning_rate": 3.787128712871287e-05,
+      "loss": 0.4436,
+      "step": 430
+    },
+    {
+      "epoch": 0.9599109131403119,
+      "grad_norm": 0.243975014278495,
+      "learning_rate": 3.783003300330033e-05,
+      "loss": 0.4292,
+      "step": 431
+    },
+    {
+      "epoch": 0.9621380846325167,
+      "grad_norm": 0.242157774083566,
+      "learning_rate": 3.7788778877887796e-05,
+      "loss": 0.4066,
+      "step": 432
+    },
+    {
+      "epoch": 0.9643652561247216,
+      "grad_norm": 0.23787781526211782,
+      "learning_rate": 3.774752475247525e-05,
+      "loss": 0.4162,
+      "step": 433
+    },
+    {
+      "epoch": 0.9665924276169265,
+      "grad_norm": 0.265386727407684,
+      "learning_rate": 3.7706270627062705e-05,
+      "loss": 0.4317,
+      "step": 434
+    },
+    {
+      "epoch": 0.9688195991091314,
+      "grad_norm": 0.2757387320364763,
+      "learning_rate": 3.766501650165017e-05,
+      "loss": 0.429,
+      "step": 435
+    },
+    {
+      "epoch": 0.9710467706013363,
+      "grad_norm": 0.23589298838942482,
+      "learning_rate": 3.762376237623763e-05,
+      "loss": 0.427,
+      "step": 436
+    },
+    {
+      "epoch": 0.9732739420935412,
+      "grad_norm": 0.3019480813401635,
+      "learning_rate": 3.758250825082508e-05,
+      "loss": 0.4105,
+      "step": 437
+    },
+    {
+      "epoch": 0.9755011135857461,
+      "grad_norm": 0.2322268885125827,
+      "learning_rate": 3.7541254125412545e-05,
+      "loss": 0.4292,
+      "step": 438
+    },
+    {
+      "epoch": 0.977728285077951,
+      "grad_norm": 0.29553504935288555,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4117,
+      "step": 439
+    },
+    {
+      "epoch": 0.9799554565701559,
+      "grad_norm": 0.25987416783203476,
+      "learning_rate": 3.745874587458746e-05,
+      "loss": 0.4152,
+      "step": 440
+    },
+    {
+      "epoch": 0.9821826280623608,
+      "grad_norm": 0.23372024946025963,
+      "learning_rate": 3.741749174917492e-05,
+      "loss": 0.4064,
+      "step": 441
+    },
+    {
+      "epoch": 0.9844097995545658,
+      "grad_norm": 0.3761401693903718,
+      "learning_rate": 3.737623762376238e-05,
+      "loss": 0.4434,
+      "step": 442
+    },
+    {
+      "epoch": 0.9866369710467706,
+      "grad_norm": 0.240221654416868,
+      "learning_rate": 3.7334983498349837e-05,
+      "loss": 0.4329,
+      "step": 443
+    },
+    {
+      "epoch": 0.9888641425389755,
+      "grad_norm": 0.22546093040269893,
+      "learning_rate": 3.7293729372937295e-05,
+      "loss": 0.4308,
+      "step": 444
+    },
+    {
+      "epoch": 0.9910913140311804,
+      "grad_norm": 0.27239165721549624,
+      "learning_rate": 3.725247524752475e-05,
+      "loss": 0.4505,
+      "step": 445
+    },
+    {
+      "epoch": 0.9933184855233853,
+      "grad_norm": 0.2611268555670104,
+      "learning_rate": 3.721122112211221e-05,
+      "loss": 0.4296,
+      "step": 446
+    },
+    {
+      "epoch": 0.9955456570155902,
+      "grad_norm": 0.22475728637651804,
+      "learning_rate": 3.7169966996699676e-05,
+      "loss": 0.4284,
+      "step": 447
+    },
+    {
+      "epoch": 0.9977728285077951,
+      "grad_norm": 0.28347572303148216,
+      "learning_rate": 3.712871287128713e-05,
+      "loss": 0.4475,
+      "step": 448
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2508993557749204,
+      "learning_rate": 3.7087458745874586e-05,
+      "loss": 0.4384,
+      "step": 449
+    },
+    {
+      "epoch": 1.0022271714922049,
+      "grad_norm": 0.29315360782372374,
+      "learning_rate": 3.704620462046205e-05,
+      "loss": 0.3846,
+      "step": 450
+    },
+    {
+      "epoch": 1.0044543429844097,
+      "grad_norm": 0.27683401693327114,
+      "learning_rate": 3.700495049504951e-05,
+      "loss": 0.3782,
+      "step": 451
+    },
+    {
+      "epoch": 1.0066815144766148,
+      "grad_norm": 0.23634840578661528,
+      "learning_rate": 3.696369636963696e-05,
+      "loss": 0.3674,
+      "step": 452
+    },
+    {
+      "epoch": 1.0089086859688197,
+      "grad_norm": 0.3002065151882711,
+      "learning_rate": 3.6922442244224426e-05,
+      "loss": 0.3759,
+      "step": 453
+    },
+    {
+      "epoch": 1.0111358574610245,
+      "grad_norm": 0.25303557292326284,
+      "learning_rate": 3.6881188118811884e-05,
+      "loss": 0.3301,
+      "step": 454
+    },
+    {
+      "epoch": 1.0133630289532294,
+      "grad_norm": 0.3119179255994917,
+      "learning_rate": 3.683993399339934e-05,
+      "loss": 0.3601,
+      "step": 455
+    },
+    {
+      "epoch": 1.0155902004454342,
+      "grad_norm": 0.2915555001895583,
+      "learning_rate": 3.67986798679868e-05,
+      "loss": 0.3628,
+      "step": 456
+    },
+    {
+      "epoch": 1.017817371937639,
+      "grad_norm": 0.25225185820529017,
+      "learning_rate": 3.675742574257426e-05,
+      "loss": 0.3549,
+      "step": 457
+    },
+    {
+      "epoch": 1.0200445434298442,
+      "grad_norm": 0.29883651297742836,
+      "learning_rate": 3.671617161716172e-05,
+      "loss": 0.3542,
+      "step": 458
+    },
+    {
+      "epoch": 1.022271714922049,
+      "grad_norm": 0.2744252590186249,
+      "learning_rate": 3.6674917491749175e-05,
+      "loss": 0.3829,
+      "step": 459
+    },
+    {
+      "epoch": 1.024498886414254,
+      "grad_norm": 0.25716474537280526,
+      "learning_rate": 3.6633663366336634e-05,
+      "loss": 0.368,
+      "step": 460
+    },
+    {
+      "epoch": 1.0267260579064588,
+      "grad_norm": 0.2609820678616415,
+      "learning_rate": 3.659240924092409e-05,
+      "loss": 0.3721,
+      "step": 461
+    },
+    {
+      "epoch": 1.0289532293986636,
+      "grad_norm": 0.27931915304683225,
+      "learning_rate": 3.655115511551156e-05,
+      "loss": 0.3585,
+      "step": 462
+    },
+    {
+      "epoch": 1.0311804008908685,
+      "grad_norm": 0.25961656509726105,
+      "learning_rate": 3.650990099009901e-05,
+      "loss": 0.3645,
+      "step": 463
+    },
+    {
+      "epoch": 1.0334075723830736,
+      "grad_norm": 0.2585612306648301,
+      "learning_rate": 3.646864686468647e-05,
+      "loss": 0.3547,
+      "step": 464
+    },
+    {
+      "epoch": 1.0356347438752784,
+      "grad_norm": 0.2989697101748151,
+      "learning_rate": 3.642739273927393e-05,
+      "loss": 0.3891,
+      "step": 465
+    },
+    {
+      "epoch": 1.0378619153674833,
+      "grad_norm": 0.2557096993002286,
+      "learning_rate": 3.638613861386139e-05,
+      "loss": 0.3586,
+      "step": 466
+    },
+    {
+      "epoch": 1.0400890868596881,
+      "grad_norm": 0.2711858210459584,
+      "learning_rate": 3.634488448844885e-05,
+      "loss": 0.3475,
+      "step": 467
+    },
+    {
+      "epoch": 1.042316258351893,
+      "grad_norm": 0.251733504110324,
+      "learning_rate": 3.6303630363036307e-05,
+      "loss": 0.3378,
+      "step": 468
+    },
+    {
+      "epoch": 1.044543429844098,
+      "grad_norm": 0.29452366944429953,
+      "learning_rate": 3.6262376237623765e-05,
+      "loss": 0.3551,
+      "step": 469
+    },
+    {
+      "epoch": 1.046770601336303,
+      "grad_norm": 0.26713159756280325,
+      "learning_rate": 3.622112211221122e-05,
+      "loss": 0.3603,
+      "step": 470
+    },
+    {
+      "epoch": 1.0489977728285078,
+      "grad_norm": 0.26982792989271304,
+      "learning_rate": 3.617986798679868e-05,
+      "loss": 0.3445,
+      "step": 471
+    },
+    {
+      "epoch": 1.0512249443207127,
+      "grad_norm": 0.29846303257877244,
+      "learning_rate": 3.613861386138614e-05,
+      "loss": 0.3242,
+      "step": 472
+    },
+    {
+      "epoch": 1.0534521158129175,
+      "grad_norm": 0.24183731075742984,
+      "learning_rate": 3.6097359735973605e-05,
+      "loss": 0.3721,
+      "step": 473
+    },
+    {
+      "epoch": 1.0556792873051224,
+      "grad_norm": 0.29520023901045445,
+      "learning_rate": 3.6056105610561056e-05,
+      "loss": 0.355,
+      "step": 474
+    },
+    {
+      "epoch": 1.0579064587973275,
+      "grad_norm": 0.3047522734685861,
+      "learning_rate": 3.6014851485148514e-05,
+      "loss": 0.351,
+      "step": 475
+    },
+    {
+      "epoch": 1.0601336302895323,
+      "grad_norm": 0.26831780421495066,
+      "learning_rate": 3.597359735973598e-05,
+      "loss": 0.3585,
+      "step": 476
+    },
+    {
+      "epoch": 1.0623608017817372,
+      "grad_norm": 0.27060903468927394,
+      "learning_rate": 3.593234323432344e-05,
+      "loss": 0.3552,
+      "step": 477
+    },
+    {
+      "epoch": 1.064587973273942,
+      "grad_norm": 0.2620962919692702,
+      "learning_rate": 3.589108910891089e-05,
+      "loss": 0.358,
+      "step": 478
+    },
+    {
+      "epoch": 1.066815144766147,
+      "grad_norm": 0.28827255446142563,
+      "learning_rate": 3.5849834983498354e-05,
+      "loss": 0.3548,
+      "step": 479
+    },
+    {
+      "epoch": 1.069042316258352,
+      "grad_norm": 0.2730218486644132,
+      "learning_rate": 3.580858085808581e-05,
+      "loss": 0.3525,
+      "step": 480
+    },
+    {
+      "epoch": 1.0712694877505569,
+      "grad_norm": 0.22229426724187393,
+      "learning_rate": 3.576732673267327e-05,
+      "loss": 0.3519,
+      "step": 481
+    },
+    {
+      "epoch": 1.0734966592427617,
+      "grad_norm": 0.25096650404869353,
+      "learning_rate": 3.572607260726073e-05,
+      "loss": 0.3677,
+      "step": 482
+    },
+    {
+      "epoch": 1.0757238307349666,
+      "grad_norm": 0.21704229516633425,
+      "learning_rate": 3.568481848184819e-05,
+      "loss": 0.3434,
+      "step": 483
+    },
+    {
+      "epoch": 1.0779510022271714,
+      "grad_norm": 0.2510064984692589,
+      "learning_rate": 3.5643564356435645e-05,
+      "loss": 0.3579,
+      "step": 484
+    },
+    {
+      "epoch": 1.0801781737193763,
+      "grad_norm": 0.2444418100304988,
+      "learning_rate": 3.5602310231023104e-05,
+      "loss": 0.3668,
+      "step": 485
+    },
+    {
+      "epoch": 1.0824053452115814,
+      "grad_norm": 0.22997335647464462,
+      "learning_rate": 3.556105610561056e-05,
+      "loss": 0.3517,
+      "step": 486
+    },
+    {
+      "epoch": 1.0846325167037862,
+      "grad_norm": 0.24920261498930532,
+      "learning_rate": 3.551980198019802e-05,
+      "loss": 0.3681,
+      "step": 487
+    },
+    {
+      "epoch": 1.086859688195991,
+      "grad_norm": 0.22014098096769694,
+      "learning_rate": 3.5478547854785485e-05,
+      "loss": 0.3482,
+      "step": 488
+    },
+    {
+      "epoch": 1.089086859688196,
+      "grad_norm": 0.2065658510344769,
+      "learning_rate": 3.543729372937294e-05,
+      "loss": 0.3287,
+      "step": 489
+    },
+    {
+      "epoch": 1.0913140311804008,
+      "grad_norm": 0.23758692598217404,
+      "learning_rate": 3.5396039603960395e-05,
+      "loss": 0.3635,
+      "step": 490
+    },
+    {
+      "epoch": 1.093541202672606,
+      "grad_norm": 0.2571919107965832,
+      "learning_rate": 3.535478547854786e-05,
+      "loss": 0.382,
+      "step": 491
+    },
+    {
+      "epoch": 1.0957683741648108,
+      "grad_norm": 0.24202112635122558,
+      "learning_rate": 3.531353135313532e-05,
+      "loss": 0.3551,
+      "step": 492
+    },
+    {
+      "epoch": 1.0979955456570156,
+      "grad_norm": 0.26045245117952176,
+      "learning_rate": 3.527227722772277e-05,
+      "loss": 0.339,
+      "step": 493
+    },
+    {
+      "epoch": 1.1002227171492205,
+      "grad_norm": 0.2473663708499315,
+      "learning_rate": 3.5231023102310235e-05,
+      "loss": 0.361,
+      "step": 494
+    },
+    {
+      "epoch": 1.1024498886414253,
+      "grad_norm": 0.2681151131548373,
+      "learning_rate": 3.518976897689769e-05,
+      "loss": 0.3609,
+      "step": 495
+    },
+    {
+      "epoch": 1.1046770601336302,
+      "grad_norm": 0.2630111804406823,
+      "learning_rate": 3.514851485148515e-05,
+      "loss": 0.3575,
+      "step": 496
+    },
+    {
+      "epoch": 1.1069042316258353,
+      "grad_norm": 0.2625195461023254,
+      "learning_rate": 3.510726072607261e-05,
+      "loss": 0.3679,
+      "step": 497
+    },
+    {
+      "epoch": 1.1091314031180401,
+      "grad_norm": 0.2666448934183908,
+      "learning_rate": 3.506600660066007e-05,
+      "loss": 0.3709,
+      "step": 498
+    },
+    {
+      "epoch": 1.111358574610245,
+      "grad_norm": 0.28347710182340724,
+      "learning_rate": 3.5024752475247526e-05,
+      "loss": 0.3615,
+      "step": 499
+    },
+    {
+      "epoch": 1.1135857461024499,
+      "grad_norm": 0.24355474473276895,
+      "learning_rate": 3.4983498349834984e-05,
+      "loss": 0.3363,
+      "step": 500
+    },
+    {
+      "epoch": 1.1158129175946547,
+      "grad_norm": 0.2729082414959115,
+      "learning_rate": 3.494224422442244e-05,
+      "loss": 0.3594,
+      "step": 501
+    },
+    {
+      "epoch": 1.1180400890868596,
+      "grad_norm": 0.29469449597071923,
+      "learning_rate": 3.49009900990099e-05,
+      "loss": 0.3639,
+      "step": 502
+    },
+    {
+      "epoch": 1.1202672605790647,
+      "grad_norm": 0.25125408230757457,
+      "learning_rate": 3.4859735973597366e-05,
+      "loss": 0.3545,
+      "step": 503
+    },
+    {
+      "epoch": 1.1224944320712695,
+      "grad_norm": 0.24147558861664065,
+      "learning_rate": 3.481848184818482e-05,
+      "loss": 0.3662,
+      "step": 504
+    },
+    {
+      "epoch": 1.1247216035634744,
+      "grad_norm": 0.29415126030943894,
+      "learning_rate": 3.4777227722772276e-05,
+      "loss": 0.3454,
+      "step": 505
+    },
+    {
+      "epoch": 1.1269487750556793,
+      "grad_norm": 0.24720611055347713,
+      "learning_rate": 3.473597359735974e-05,
+      "loss": 0.3497,
+      "step": 506
+    },
+    {
+      "epoch": 1.1291759465478841,
+      "grad_norm": 0.25957656510507626,
+      "learning_rate": 3.46947194719472e-05,
+      "loss": 0.347,
+      "step": 507
+    },
+    {
+      "epoch": 1.131403118040089,
+      "grad_norm": 0.2450716784138782,
+      "learning_rate": 3.465346534653465e-05,
+      "loss": 0.366,
+      "step": 508
+    },
+    {
+      "epoch": 1.133630289532294,
+      "grad_norm": 0.30031981997460955,
+      "learning_rate": 3.4612211221122115e-05,
+      "loss": 0.3576,
+      "step": 509
+    },
+    {
+      "epoch": 1.135857461024499,
+      "grad_norm": 0.24446990508884073,
+      "learning_rate": 3.4570957095709574e-05,
+      "loss": 0.3573,
+      "step": 510
+    },
+    {
+      "epoch": 1.1380846325167038,
+      "grad_norm": 0.22244869982263982,
+      "learning_rate": 3.452970297029703e-05,
+      "loss": 0.354,
+      "step": 511
+    },
+    {
+      "epoch": 1.1403118040089086,
+      "grad_norm": 0.26352213431786436,
+      "learning_rate": 3.448844884488449e-05,
+      "loss": 0.3693,
+      "step": 512
+    },
+    {
+      "epoch": 1.1425389755011135,
+      "grad_norm": 0.2142385019175425,
+      "learning_rate": 3.444719471947195e-05,
+      "loss": 0.339,
+      "step": 513
+    },
+    {
+      "epoch": 1.1447661469933186,
+      "grad_norm": 0.277588850926534,
+      "learning_rate": 3.440594059405941e-05,
+      "loss": 0.3579,
+      "step": 514
+    },
+    {
+      "epoch": 1.1469933184855234,
+      "grad_norm": 0.23632968130731732,
+      "learning_rate": 3.4364686468646865e-05,
+      "loss": 0.3569,
+      "step": 515
+    },
+    {
+      "epoch": 1.1492204899777283,
+      "grad_norm": 0.2788228005066622,
+      "learning_rate": 3.432343234323432e-05,
+      "loss": 0.3613,
+      "step": 516
+    },
+    {
+      "epoch": 1.1514476614699332,
+      "grad_norm": 0.24060550221371907,
+      "learning_rate": 3.428217821782179e-05,
+      "loss": 0.3477,
+      "step": 517
+    },
+    {
+      "epoch": 1.153674832962138,
+      "grad_norm": 0.258548604271602,
+      "learning_rate": 3.4240924092409246e-05,
+      "loss": 0.3357,
+      "step": 518
+    },
+    {
+      "epoch": 1.1559020044543429,
+      "grad_norm": 0.2193227357341931,
+      "learning_rate": 3.41996699669967e-05,
+      "loss": 0.3387,
+      "step": 519
+    },
+    {
+      "epoch": 1.158129175946548,
+      "grad_norm": 0.23695591929388757,
+      "learning_rate": 3.415841584158416e-05,
+      "loss": 0.35,
+      "step": 520
+    },
+    {
+      "epoch": 1.1603563474387528,
+      "grad_norm": 0.24085283197224489,
+      "learning_rate": 3.411716171617162e-05,
+      "loss": 0.3518,
+      "step": 521
+    },
+    {
+      "epoch": 1.1625835189309577,
+      "grad_norm": 0.26698388090393044,
+      "learning_rate": 3.407590759075908e-05,
+      "loss": 0.3728,
+      "step": 522
+    },
+    {
+      "epoch": 1.1648106904231625,
+      "grad_norm": 0.2514334608054634,
+      "learning_rate": 3.403465346534654e-05,
+      "loss": 0.368,
+      "step": 523
+    },
+    {
+      "epoch": 1.1670378619153674,
+      "grad_norm": 0.2706263407063563,
+      "learning_rate": 3.3993399339933996e-05,
+      "loss": 0.3625,
+      "step": 524
+    },
+    {
+      "epoch": 1.1692650334075725,
+      "grad_norm": 0.23649808379024595,
+      "learning_rate": 3.3952145214521454e-05,
+      "loss": 0.3473,
+      "step": 525
+    },
+    {
+      "epoch": 1.1714922048997773,
+      "grad_norm": 0.22320907181621547,
+      "learning_rate": 3.391089108910891e-05,
+      "loss": 0.334,
+      "step": 526
+    },
+    {
+      "epoch": 1.1737193763919822,
+      "grad_norm": 0.2395789753574448,
+      "learning_rate": 3.386963696369637e-05,
+      "loss": 0.3577,
+      "step": 527
+    },
+    {
+      "epoch": 1.175946547884187,
+      "grad_norm": 0.22982302309839056,
+      "learning_rate": 3.382838283828383e-05,
+      "loss": 0.336,
+      "step": 528
+    },
+    {
+      "epoch": 1.178173719376392,
+      "grad_norm": 0.26592300199211427,
+      "learning_rate": 3.3787128712871294e-05,
+      "loss": 0.3737,
+      "step": 529
+    },
+    {
+      "epoch": 1.1804008908685968,
+      "grad_norm": 0.2379664507554433,
+      "learning_rate": 3.3745874587458746e-05,
+      "loss": 0.3508,
+      "step": 530
+    },
+    {
+      "epoch": 1.1826280623608019,
+      "grad_norm": 0.27284638634221087,
+      "learning_rate": 3.3704620462046204e-05,
+      "loss": 0.3788,
+      "step": 531
+    },
+    {
+      "epoch": 1.1848552338530067,
+      "grad_norm": 0.22298690979985145,
+      "learning_rate": 3.366336633663367e-05,
+      "loss": 0.361,
+      "step": 532
+    },
+    {
+      "epoch": 1.1870824053452116,
+      "grad_norm": 0.25342901703627585,
+      "learning_rate": 3.362211221122113e-05,
+      "loss": 0.3461,
+      "step": 533
+    },
+    {
+      "epoch": 1.1893095768374164,
+      "grad_norm": 0.2462918835412843,
+      "learning_rate": 3.358085808580858e-05,
+      "loss": 0.3673,
+      "step": 534
+    },
+    {
+      "epoch": 1.1915367483296213,
+      "grad_norm": 0.23480418831778518,
+      "learning_rate": 3.3539603960396044e-05,
+      "loss": 0.3305,
+      "step": 535
+    },
+    {
+      "epoch": 1.1937639198218264,
+      "grad_norm": 0.22579058718977288,
+      "learning_rate": 3.34983498349835e-05,
+      "loss": 0.3484,
+      "step": 536
+    },
+    {
+      "epoch": 1.1959910913140313,
+      "grad_norm": 0.22351739536686402,
+      "learning_rate": 3.345709570957096e-05,
+      "loss": 0.3605,
+      "step": 537
+    },
+    {
+      "epoch": 1.1982182628062361,
+      "grad_norm": 0.23250187182794887,
+      "learning_rate": 3.341584158415842e-05,
+      "loss": 0.3525,
+      "step": 538
+    },
+    {
+      "epoch": 1.200445434298441,
+      "grad_norm": 0.22930925968735663,
+      "learning_rate": 3.337458745874588e-05,
+      "loss": 0.3667,
+      "step": 539
+    },
+    {
+      "epoch": 1.2026726057906458,
+      "grad_norm": 0.23045423822229336,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3495,
+      "step": 540
+    },
+    {
+      "epoch": 1.2048997772828507,
+      "grad_norm": 0.23457659707681205,
+      "learning_rate": 3.329207920792079e-05,
+      "loss": 0.358,
+      "step": 541
+    },
+    {
+      "epoch": 1.2071269487750558,
+      "grad_norm": 0.26247810496966095,
+      "learning_rate": 3.325082508250825e-05,
+      "loss": 0.3507,
+      "step": 542
+    },
+    {
+      "epoch": 1.2093541202672606,
+      "grad_norm": 0.21223867220084797,
+      "learning_rate": 3.320957095709571e-05,
+      "loss": 0.3527,
+      "step": 543
+    },
+    {
+      "epoch": 1.2115812917594655,
+      "grad_norm": 0.235166791081928,
+      "learning_rate": 3.3168316831683175e-05,
+      "loss": 0.357,
+      "step": 544
+    },
+    {
+      "epoch": 1.2138084632516704,
+      "grad_norm": 0.2541299314311305,
+      "learning_rate": 3.3127062706270626e-05,
+      "loss": 0.3559,
+      "step": 545
+    },
+    {
+      "epoch": 1.2160356347438752,
+      "grad_norm": 0.23336231564934568,
+      "learning_rate": 3.3085808580858084e-05,
+      "loss": 0.3723,
+      "step": 546
+    },
+    {
+      "epoch": 1.2182628062360803,
+      "grad_norm": 0.24580602513394764,
+      "learning_rate": 3.304455445544555e-05,
+      "loss": 0.3684,
+      "step": 547
+    },
+    {
+      "epoch": 1.2204899777282852,
+      "grad_norm": 0.20485755646576795,
+      "learning_rate": 3.300330033003301e-05,
+      "loss": 0.3531,
+      "step": 548
+    },
+    {
+      "epoch": 1.22271714922049,
+      "grad_norm": 0.24371288647694048,
+      "learning_rate": 3.296204620462046e-05,
+      "loss": 0.3585,
+      "step": 549
+    },
+    {
+      "epoch": 1.2249443207126949,
+      "grad_norm": 0.19774021220722549,
+      "learning_rate": 3.2920792079207924e-05,
+      "loss": 0.339,
+      "step": 550
+    },
+    {
+      "epoch": 1.2271714922048997,
+      "grad_norm": 0.2600235842796674,
+      "learning_rate": 3.287953795379538e-05,
+      "loss": 0.36,
+      "step": 551
+    },
+    {
+      "epoch": 1.2293986636971046,
+      "grad_norm": 0.22024280724938608,
+      "learning_rate": 3.283828382838284e-05,
+      "loss": 0.3513,
+      "step": 552
+    },
+    {
+      "epoch": 1.2316258351893095,
+      "grad_norm": 0.23903906420041407,
+      "learning_rate": 3.27970297029703e-05,
+      "loss": 0.3552,
+      "step": 553
+    },
+    {
+      "epoch": 1.2338530066815145,
+      "grad_norm": 0.270801620970715,
+      "learning_rate": 3.275577557755776e-05,
+      "loss": 0.3688,
+      "step": 554
+    },
+    {
+      "epoch": 1.2360801781737194,
+      "grad_norm": 0.21905876480622277,
+      "learning_rate": 3.2714521452145216e-05,
+      "loss": 0.3445,
+      "step": 555
+    },
+    {
+      "epoch": 1.2383073496659243,
+      "grad_norm": 0.2403828223658685,
+      "learning_rate": 3.2673267326732674e-05,
+      "loss": 0.3511,
+      "step": 556
+    },
+    {
+      "epoch": 1.2405345211581291,
+      "grad_norm": 0.2456723575282447,
+      "learning_rate": 3.263201320132013e-05,
+      "loss": 0.3544,
+      "step": 557
+    },
+    {
+      "epoch": 1.242761692650334,
+      "grad_norm": 0.26448553284073184,
+      "learning_rate": 3.259075907590759e-05,
+      "loss": 0.3522,
+      "step": 558
+    },
+    {
+      "epoch": 1.244988864142539,
+      "grad_norm": 0.24073145486725794,
+      "learning_rate": 3.2549504950495055e-05,
+      "loss": 0.3508,
+      "step": 559
+    },
+    {
+      "epoch": 1.247216035634744,
+      "grad_norm": 0.27736400148415535,
+      "learning_rate": 3.250825082508251e-05,
+      "loss": 0.3648,
+      "step": 560
+    },
+    {
+      "epoch": 1.2494432071269488,
+      "grad_norm": 0.2728738915660453,
+      "learning_rate": 3.2466996699669965e-05,
+      "loss": 0.3704,
+      "step": 561
+    },
+    {
+      "epoch": 1.2516703786191536,
+      "grad_norm": 0.26300898311896,
+      "learning_rate": 3.242574257425743e-05,
+      "loss": 0.3668,
+      "step": 562
+    },
+    {
+      "epoch": 1.2538975501113585,
+      "grad_norm": 0.2548127745894571,
+      "learning_rate": 3.238448844884489e-05,
+      "loss": 0.3729,
+      "step": 563
+    },
+    {
+      "epoch": 1.2561247216035634,
+      "grad_norm": 0.1961809070235041,
+      "learning_rate": 3.234323432343234e-05,
+      "loss": 0.3527,
+      "step": 564
+    },
+    {
+      "epoch": 1.2583518930957684,
+      "grad_norm": 0.22948976382101388,
+      "learning_rate": 3.2301980198019805e-05,
+      "loss": 0.3367,
+      "step": 565
+    },
+    {
+      "epoch": 1.2605790645879733,
+      "grad_norm": 0.2716897968846777,
+      "learning_rate": 3.226072607260726e-05,
+      "loss": 0.3791,
+      "step": 566
+    },
+    {
+      "epoch": 1.2628062360801782,
+      "grad_norm": 0.23513234834544378,
+      "learning_rate": 3.221947194719472e-05,
+      "loss": 0.3693,
+      "step": 567
+    },
+    {
+      "epoch": 1.265033407572383,
+      "grad_norm": 0.24787013974811967,
+      "learning_rate": 3.217821782178218e-05,
+      "loss": 0.3384,
+      "step": 568
+    },
+    {
+      "epoch": 1.267260579064588,
+      "grad_norm": 0.25107053798110607,
+      "learning_rate": 3.213696369636964e-05,
+      "loss": 0.3677,
+      "step": 569
+    },
+    {
+      "epoch": 1.269487750556793,
+      "grad_norm": 0.20914988529919357,
+      "learning_rate": 3.20957095709571e-05,
+      "loss": 0.3499,
+      "step": 570
+    },
+    {
+      "epoch": 1.2717149220489978,
+      "grad_norm": 0.25215966314258914,
+      "learning_rate": 3.2054455445544554e-05,
+      "loss": 0.3491,
+      "step": 571
+    },
+    {
+      "epoch": 1.2739420935412027,
+      "grad_norm": 0.25325080210028006,
+      "learning_rate": 3.201320132013201e-05,
+      "loss": 0.3582,
+      "step": 572
+    },
+    {
+      "epoch": 1.2761692650334076,
+      "grad_norm": 0.21780047792534563,
+      "learning_rate": 3.197194719471948e-05,
+      "loss": 0.3474,
+      "step": 573
+    },
+    {
+      "epoch": 1.2783964365256124,
+      "grad_norm": 0.2214051811188461,
+      "learning_rate": 3.1930693069306936e-05,
+      "loss": 0.3552,
+      "step": 574
+    },
+    {
+      "epoch": 1.2806236080178173,
+      "grad_norm": 0.24543082087845347,
+      "learning_rate": 3.188943894389439e-05,
+      "loss": 0.3529,
+      "step": 575
+    },
+    {
+      "epoch": 1.2828507795100224,
+      "grad_norm": 0.20158614638718708,
+      "learning_rate": 3.184818481848185e-05,
+      "loss": 0.3434,
+      "step": 576
+    },
+    {
+      "epoch": 1.2850779510022272,
+      "grad_norm": 0.2296252359307742,
+      "learning_rate": 3.180693069306931e-05,
+      "loss": 0.3751,
+      "step": 577
+    },
+    {
+      "epoch": 1.287305122494432,
+      "grad_norm": 0.2185964200183311,
+      "learning_rate": 3.176567656765677e-05,
+      "loss": 0.367,
+      "step": 578
+    },
+    {
+      "epoch": 1.289532293986637,
+      "grad_norm": 0.2398417205034457,
+      "learning_rate": 3.172442244224423e-05,
+      "loss": 0.3573,
+      "step": 579
+    },
+    {
+      "epoch": 1.2917594654788418,
+      "grad_norm": 0.20606951067584606,
+      "learning_rate": 3.1683168316831686e-05,
+      "loss": 0.3605,
+      "step": 580
+    },
+    {
+      "epoch": 1.2939866369710469,
+      "grad_norm": 0.253402154414871,
+      "learning_rate": 3.1641914191419144e-05,
+      "loss": 0.385,
+      "step": 581
+    },
+    {
+      "epoch": 1.2962138084632517,
+      "grad_norm": 0.21569327617012055,
+      "learning_rate": 3.16006600660066e-05,
+      "loss": 0.3615,
+      "step": 582
+    },
+    {
+      "epoch": 1.2984409799554566,
+      "grad_norm": 0.20993771911712605,
+      "learning_rate": 3.155940594059406e-05,
+      "loss": 0.3467,
+      "step": 583
+    },
+    {
+      "epoch": 1.3006681514476615,
+      "grad_norm": 0.25707612252706463,
+      "learning_rate": 3.151815181518152e-05,
+      "loss": 0.3698,
+      "step": 584
+    },
+    {
+      "epoch": 1.3028953229398663,
+      "grad_norm": 0.23729147871242315,
+      "learning_rate": 3.1476897689768984e-05,
+      "loss": 0.3571,
+      "step": 585
+    },
+    {
+      "epoch": 1.3051224944320712,
+      "grad_norm": 0.2058740175214167,
+      "learning_rate": 3.1435643564356435e-05,
+      "loss": 0.3491,
+      "step": 586
+    },
+    {
+      "epoch": 1.307349665924276,
+      "grad_norm": 0.25787314593557403,
+      "learning_rate": 3.139438943894389e-05,
+      "loss": 0.3729,
+      "step": 587
+    },
+    {
+      "epoch": 1.3095768374164811,
+      "grad_norm": 0.23567481362489415,
+      "learning_rate": 3.135313531353136e-05,
+      "loss": 0.3674,
+      "step": 588
+    },
+    {
+      "epoch": 1.311804008908686,
+      "grad_norm": 0.2103686872656417,
+      "learning_rate": 3.131188118811882e-05,
+      "loss": 0.3546,
+      "step": 589
+    },
+    {
+      "epoch": 1.3140311804008908,
+      "grad_norm": 0.2246018723719588,
+      "learning_rate": 3.127062706270627e-05,
+      "loss": 0.3342,
+      "step": 590
+    },
+    {
+      "epoch": 1.3162583518930957,
+      "grad_norm": 0.2550281829523844,
+      "learning_rate": 3.122937293729373e-05,
+      "loss": 0.3604,
+      "step": 591
+    },
+    {
+      "epoch": 1.3184855233853008,
+      "grad_norm": 0.22197952276250127,
+      "learning_rate": 3.118811881188119e-05,
+      "loss": 0.3429,
+      "step": 592
+    },
+    {
+      "epoch": 1.3207126948775056,
+      "grad_norm": 0.25880191441627104,
+      "learning_rate": 3.114686468646865e-05,
+      "loss": 0.3461,
+      "step": 593
+    },
+    {
+      "epoch": 1.3229398663697105,
+      "grad_norm": 0.22702691621590193,
+      "learning_rate": 3.110561056105611e-05,
+      "loss": 0.3584,
+      "step": 594
+    },
+    {
+      "epoch": 1.3251670378619154,
+      "grad_norm": 0.23337868994919198,
+      "learning_rate": 3.1064356435643566e-05,
+      "loss": 0.3791,
+      "step": 595
+    },
+    {
+      "epoch": 1.3273942093541202,
+      "grad_norm": 0.24931669973599038,
+      "learning_rate": 3.1023102310231024e-05,
+      "loss": 0.358,
+      "step": 596
+    },
+    {
+      "epoch": 1.329621380846325,
+      "grad_norm": 0.21248967636665253,
+      "learning_rate": 3.098184818481848e-05,
+      "loss": 0.3333,
+      "step": 597
+    },
+    {
+      "epoch": 1.33184855233853,
+      "grad_norm": 0.22273911325503096,
+      "learning_rate": 3.094059405940594e-05,
+      "loss": 0.3763,
+      "step": 598
+    },
+    {
+      "epoch": 1.334075723830735,
+      "grad_norm": 0.2314939505997935,
+      "learning_rate": 3.08993399339934e-05,
+      "loss": 0.3507,
+      "step": 599
+    },
+    {
+      "epoch": 1.3363028953229399,
+      "grad_norm": 0.22611256534493768,
+      "learning_rate": 3.0858085808580864e-05,
+      "loss": 0.3657,
+      "step": 600
+    },
+    {
+      "epoch": 1.3385300668151447,
+      "grad_norm": 0.21856148705704515,
+      "learning_rate": 3.0816831683168316e-05,
+      "loss": 0.3602,
+      "step": 601
+    },
+    {
+      "epoch": 1.3407572383073496,
+      "grad_norm": 0.2418827295907198,
+      "learning_rate": 3.0775577557755774e-05,
+      "loss": 0.3382,
+      "step": 602
+    },
+    {
+      "epoch": 1.3429844097995547,
+      "grad_norm": 0.23647807359827352,
+      "learning_rate": 3.073432343234324e-05,
+      "loss": 0.3602,
+      "step": 603
+    },
+    {
+      "epoch": 1.3452115812917596,
+      "grad_norm": 0.22924025988751517,
+      "learning_rate": 3.06930693069307e-05,
+      "loss": 0.3572,
+      "step": 604
+    },
+    {
+      "epoch": 1.3474387527839644,
+      "grad_norm": 0.22831615741384578,
+      "learning_rate": 3.065181518151815e-05,
+      "loss": 0.36,
+      "step": 605
+    },
+    {
+      "epoch": 1.3496659242761693,
+      "grad_norm": 0.24207308088696777,
+      "learning_rate": 3.0610561056105614e-05,
+      "loss": 0.3591,
+      "step": 606
+    },
+    {
+      "epoch": 1.3518930957683741,
+      "grad_norm": 0.22533322573657474,
+      "learning_rate": 3.056930693069307e-05,
+      "loss": 0.3783,
+      "step": 607
+    },
+    {
+      "epoch": 1.354120267260579,
+      "grad_norm": 0.22354237123525514,
+      "learning_rate": 3.052805280528053e-05,
+      "loss": 0.3646,
+      "step": 608
+    },
+    {
+      "epoch": 1.3563474387527839,
+      "grad_norm": 0.21082532615883712,
+      "learning_rate": 3.048679867986799e-05,
+      "loss": 0.3685,
+      "step": 609
+    },
+    {
+      "epoch": 1.358574610244989,
+      "grad_norm": 0.21724487552837043,
+      "learning_rate": 3.0445544554455447e-05,
+      "loss": 0.3637,
+      "step": 610
+    },
+    {
+      "epoch": 1.3608017817371938,
+      "grad_norm": 0.2282169646639709,
+      "learning_rate": 3.0404290429042902e-05,
+      "loss": 0.361,
+      "step": 611
+    },
+    {
+      "epoch": 1.3630289532293987,
+      "grad_norm": 0.20533332057422468,
+      "learning_rate": 3.0363036303630367e-05,
+      "loss": 0.3335,
+      "step": 612
+    },
+    {
+      "epoch": 1.3652561247216035,
+      "grad_norm": 0.21314989490113376,
+      "learning_rate": 3.032178217821782e-05,
+      "loss": 0.3592,
+      "step": 613
+    },
+    {
+      "epoch": 1.3674832962138086,
+      "grad_norm": 0.2141447141940089,
+      "learning_rate": 3.028052805280528e-05,
+      "loss": 0.3359,
+      "step": 614
+    },
+    {
+      "epoch": 1.3697104677060135,
+      "grad_norm": 0.24581117751522166,
+      "learning_rate": 3.023927392739274e-05,
+      "loss": 0.3716,
+      "step": 615
+    },
+    {
+      "epoch": 1.3719376391982183,
+      "grad_norm": 0.2275094549967322,
+      "learning_rate": 3.01980198019802e-05,
+      "loss": 0.3553,
+      "step": 616
+    },
+    {
+      "epoch": 1.3741648106904232,
+      "grad_norm": 0.23356205976864639,
+      "learning_rate": 3.015676567656766e-05,
+      "loss": 0.3482,
+      "step": 617
+    },
+    {
+      "epoch": 1.376391982182628,
+      "grad_norm": 0.23882460698057495,
+      "learning_rate": 3.0115511551155116e-05,
+      "loss": 0.3579,
+      "step": 618
+    },
+    {
+      "epoch": 1.378619153674833,
+      "grad_norm": 0.23482925204649976,
+      "learning_rate": 3.0074257425742575e-05,
+      "loss": 0.3467,
+      "step": 619
+    },
+    {
+      "epoch": 1.3808463251670378,
+      "grad_norm": 0.22272122322754917,
+      "learning_rate": 3.0033003300330036e-05,
+      "loss": 0.3558,
+      "step": 620
+    },
+    {
+      "epoch": 1.3830734966592428,
+      "grad_norm": 0.2417717293854326,
+      "learning_rate": 2.9991749174917494e-05,
+      "loss": 0.3399,
+      "step": 621
+    },
+    {
+      "epoch": 1.3853006681514477,
+      "grad_norm": 0.21560732990542106,
+      "learning_rate": 2.995049504950495e-05,
+      "loss": 0.3263,
+      "step": 622
+    },
+    {
+      "epoch": 1.3875278396436526,
+      "grad_norm": 0.23511847998843455,
+      "learning_rate": 2.9909240924092414e-05,
+      "loss": 0.357,
+      "step": 623
+    },
+    {
+      "epoch": 1.3897550111358574,
+      "grad_norm": 0.2419549890007216,
+      "learning_rate": 2.986798679867987e-05,
+      "loss": 0.3376,
+      "step": 624
+    },
+    {
+      "epoch": 1.3919821826280623,
+      "grad_norm": 0.2418915221005408,
+      "learning_rate": 2.9826732673267327e-05,
+      "loss": 0.3559,
+      "step": 625
+    },
+    {
+      "epoch": 1.3942093541202674,
+      "grad_norm": 0.2219366354806474,
+      "learning_rate": 2.978547854785479e-05,
+      "loss": 0.3607,
+      "step": 626
+    },
+    {
+      "epoch": 1.3964365256124722,
+      "grad_norm": 0.22151027055415365,
+      "learning_rate": 2.9744224422442247e-05,
+      "loss": 0.3421,
+      "step": 627
+    },
+    {
+      "epoch": 1.398663697104677,
+      "grad_norm": 0.2818296636352545,
+      "learning_rate": 2.9702970297029702e-05,
+      "loss": 0.3598,
+      "step": 628
+    },
+    {
+      "epoch": 1.400890868596882,
+      "grad_norm": 0.22608544158053545,
+      "learning_rate": 2.9661716171617164e-05,
+      "loss": 0.3757,
+      "step": 629
+    },
+    {
+      "epoch": 1.4031180400890868,
+      "grad_norm": 0.2542515757235494,
+      "learning_rate": 2.9620462046204622e-05,
+      "loss": 0.3566,
+      "step": 630
+    },
+    {
+      "epoch": 1.4053452115812917,
+      "grad_norm": 0.22082951585632452,
+      "learning_rate": 2.957920792079208e-05,
+      "loss": 0.3532,
+      "step": 631
+    },
+    {
+      "epoch": 1.4075723830734965,
+      "grad_norm": 0.2152525241114565,
+      "learning_rate": 2.9537953795379542e-05,
+      "loss": 0.3476,
+      "step": 632
+    },
+    {
+      "epoch": 1.4097995545657016,
+      "grad_norm": 0.22504727209571587,
+      "learning_rate": 2.9496699669966997e-05,
+      "loss": 0.3506,
+      "step": 633
+    },
+    {
+      "epoch": 1.4120267260579065,
+      "grad_norm": 0.23833476284032978,
+      "learning_rate": 2.9455445544554455e-05,
+      "loss": 0.3698,
+      "step": 634
+    },
+    {
+      "epoch": 1.4142538975501113,
+      "grad_norm": 0.21086039461299577,
+      "learning_rate": 2.9414191419141917e-05,
+      "loss": 0.366,
+      "step": 635
+    },
+    {
+      "epoch": 1.4164810690423162,
+      "grad_norm": 0.209689633966253,
+      "learning_rate": 2.9372937293729375e-05,
+      "loss": 0.3464,
+      "step": 636
+    },
+    {
+      "epoch": 1.4187082405345213,
+      "grad_norm": 0.24474710567495397,
+      "learning_rate": 2.933168316831683e-05,
+      "loss": 0.3532,
+      "step": 637
+    },
+    {
+      "epoch": 1.4209354120267261,
+      "grad_norm": 0.22400192345008954,
+      "learning_rate": 2.9290429042904295e-05,
+      "loss": 0.3655,
+      "step": 638
+    },
+    {
+      "epoch": 1.423162583518931,
+      "grad_norm": 0.2058797826542391,
+      "learning_rate": 2.924917491749175e-05,
+      "loss": 0.3604,
+      "step": 639
+    },
+    {
+      "epoch": 1.4253897550111359,
+      "grad_norm": 0.2063600373971181,
+      "learning_rate": 2.9207920792079208e-05,
+      "loss": 0.3544,
+      "step": 640
+    },
+    {
+      "epoch": 1.4276169265033407,
+      "grad_norm": 0.19789395719309963,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.3443,
+      "step": 641
+    },
+    {
+      "epoch": 1.4298440979955456,
+      "grad_norm": 0.19534241146623657,
+      "learning_rate": 2.9125412541254128e-05,
+      "loss": 0.3558,
+      "step": 642
+    },
+    {
+      "epoch": 1.4320712694877504,
+      "grad_norm": 0.20204042866647118,
+      "learning_rate": 2.9084158415841583e-05,
+      "loss": 0.3696,
+      "step": 643
+    },
+    {
+      "epoch": 1.4342984409799555,
+      "grad_norm": 0.22097812615526374,
+      "learning_rate": 2.9042904290429045e-05,
+      "loss": 0.3601,
+      "step": 644
+    },
+    {
+      "epoch": 1.4365256124721604,
+      "grad_norm": 0.19680712562974112,
+      "learning_rate": 2.9001650165016503e-05,
+      "loss": 0.3398,
+      "step": 645
+    },
+    {
+      "epoch": 1.4387527839643652,
+      "grad_norm": 0.20759775269766909,
+      "learning_rate": 2.896039603960396e-05,
+      "loss": 0.3478,
+      "step": 646
+    },
+    {
+      "epoch": 1.44097995545657,
+      "grad_norm": 0.2345285222857468,
+      "learning_rate": 2.8919141914191423e-05,
+      "loss": 0.3636,
+      "step": 647
+    },
+    {
+      "epoch": 1.4432071269487752,
+      "grad_norm": 0.22986722410151925,
+      "learning_rate": 2.8877887788778878e-05,
+      "loss": 0.3724,
+      "step": 648
+    },
+    {
+      "epoch": 1.44543429844098,
+      "grad_norm": 0.23105094408176388,
+      "learning_rate": 2.8836633663366336e-05,
+      "loss": 0.3548,
+      "step": 649
+    },
+    {
+      "epoch": 1.447661469933185,
+      "grad_norm": 0.25829463050442975,
+      "learning_rate": 2.8795379537953797e-05,
+      "loss": 0.3719,
+      "step": 650
+    },
+    {
+      "epoch": 1.4498886414253898,
+      "grad_norm": 0.424453313230707,
+      "learning_rate": 2.8754125412541256e-05,
+      "loss": 0.3745,
+      "step": 651
+    },
+    {
+      "epoch": 1.4521158129175946,
+      "grad_norm": 0.2546120321330925,
+      "learning_rate": 2.871287128712871e-05,
+      "loss": 0.372,
+      "step": 652
+    },
+    {
+      "epoch": 1.4543429844097995,
+      "grad_norm": 0.23383800944478716,
+      "learning_rate": 2.8671617161716176e-05,
+      "loss": 0.3525,
+      "step": 653
+    },
+    {
+      "epoch": 1.4565701559020043,
+      "grad_norm": 7.318371253265792,
+      "learning_rate": 2.863036303630363e-05,
+      "loss": 0.3722,
+      "step": 654
+    },
+    {
+      "epoch": 1.4587973273942094,
+      "grad_norm": 0.28892038696657446,
+      "learning_rate": 2.858910891089109e-05,
+      "loss": 0.3373,
+      "step": 655
+    },
+    {
+      "epoch": 1.4610244988864143,
+      "grad_norm": 0.278423963067231,
+      "learning_rate": 2.854785478547855e-05,
+      "loss": 0.3519,
+      "step": 656
+    },
+    {
+      "epoch": 1.4632516703786191,
+      "grad_norm": 0.20740580993529,
+      "learning_rate": 2.850660066006601e-05,
+      "loss": 0.3452,
+      "step": 657
+    },
+    {
+      "epoch": 1.465478841870824,
+      "grad_norm": 0.2559023935627639,
+      "learning_rate": 2.8465346534653464e-05,
+      "loss": 0.3542,
+      "step": 658
+    },
+    {
+      "epoch": 1.467706013363029,
+      "grad_norm": 0.2494031935968005,
+      "learning_rate": 2.8424092409240925e-05,
+      "loss": 0.3554,
+      "step": 659
+    },
+    {
+      "epoch": 1.469933184855234,
+      "grad_norm": 0.21827118145506857,
+      "learning_rate": 2.8382838283828383e-05,
+      "loss": 0.3652,
+      "step": 660
+    },
+    {
+      "epoch": 1.4721603563474388,
+      "grad_norm": 0.24533730978200863,
+      "learning_rate": 2.834158415841584e-05,
+      "loss": 0.3598,
+      "step": 661
+    },
+    {
+      "epoch": 1.4743875278396437,
+      "grad_norm": 0.25328154863191904,
+      "learning_rate": 2.8300330033003303e-05,
+      "loss": 0.3544,
+      "step": 662
+    },
+    {
+      "epoch": 1.4766146993318485,
+      "grad_norm": 0.2668011805645278,
+      "learning_rate": 2.8259075907590758e-05,
+      "loss": 0.3596,
+      "step": 663
+    },
+    {
+      "epoch": 1.4788418708240534,
+      "grad_norm": 0.22271391411019326,
+      "learning_rate": 2.8217821782178216e-05,
+      "loss": 0.3441,
+      "step": 664
+    },
+    {
+      "epoch": 1.4810690423162582,
+      "grad_norm": 0.25938717488194835,
+      "learning_rate": 2.8176567656765678e-05,
+      "loss": 0.3658,
+      "step": 665
+    },
+    {
+      "epoch": 1.4832962138084633,
+      "grad_norm": 0.20781850068281127,
+      "learning_rate": 2.8135313531353136e-05,
+      "loss": 0.3721,
+      "step": 666
+    },
+    {
+      "epoch": 1.4855233853006682,
+      "grad_norm": 0.23085833582398785,
+      "learning_rate": 2.8094059405940598e-05,
+      "loss": 0.354,
+      "step": 667
+    },
+    {
+      "epoch": 1.487750556792873,
+      "grad_norm": 0.2656626528844188,
+      "learning_rate": 2.8052805280528056e-05,
+      "loss": 0.3638,
+      "step": 668
+    },
+    {
+      "epoch": 1.489977728285078,
+      "grad_norm": 0.24044869806682156,
+      "learning_rate": 2.801155115511551e-05,
+      "loss": 0.3659,
+      "step": 669
+    },
+    {
+      "epoch": 1.492204899777283,
+      "grad_norm": 0.21472096623862313,
+      "learning_rate": 2.7970297029702973e-05,
+      "loss": 0.3676,
+      "step": 670
+    },
+    {
+      "epoch": 1.4944320712694878,
+      "grad_norm": 0.24686435809063587,
+      "learning_rate": 2.792904290429043e-05,
+      "loss": 0.3585,
+      "step": 671
+    },
+    {
+      "epoch": 1.4966592427616927,
+      "grad_norm": 0.24840471134095787,
+      "learning_rate": 2.788778877887789e-05,
+      "loss": 0.3664,
+      "step": 672
+    },
+    {
+      "epoch": 1.4988864142538976,
+      "grad_norm": 0.2215556073459033,
+      "learning_rate": 2.784653465346535e-05,
+      "loss": 0.3551,
+      "step": 673
+    },
+    {
+      "epoch": 1.5011135857461024,
+      "grad_norm": 0.2671878307827319,
+      "learning_rate": 2.7805280528052806e-05,
+      "loss": 0.3395,
+      "step": 674
+    },
+    {
+      "epoch": 1.5033407572383073,
+      "grad_norm": 0.22945942117384913,
+      "learning_rate": 2.7764026402640264e-05,
+      "loss": 0.3753,
+      "step": 675
+    },
+    {
+      "epoch": 1.5055679287305122,
+      "grad_norm": 0.3375181873255935,
+      "learning_rate": 2.7722772277227726e-05,
+      "loss": 0.365,
+      "step": 676
+    },
+    {
+      "epoch": 1.507795100222717,
+      "grad_norm": 0.22619237117388163,
+      "learning_rate": 2.7681518151815184e-05,
+      "loss": 0.3488,
+      "step": 677
+    },
+    {
+      "epoch": 1.510022271714922,
+      "grad_norm": 0.34134705084759415,
+      "learning_rate": 2.764026402640264e-05,
+      "loss": 0.3562,
+      "step": 678
+    },
+    {
+      "epoch": 1.512249443207127,
+      "grad_norm": 0.2377850642522043,
+      "learning_rate": 2.7599009900990104e-05,
+      "loss": 0.3564,
+      "step": 679
+    },
+    {
+      "epoch": 1.5144766146993318,
+      "grad_norm": 0.22461681967594943,
+      "learning_rate": 2.755775577557756e-05,
+      "loss": 0.375,
+      "step": 680
+    },
+    {
+      "epoch": 1.516703786191537,
+      "grad_norm": 0.27346902302943626,
+      "learning_rate": 2.7516501650165017e-05,
+      "loss": 0.3499,
+      "step": 681
+    },
+    {
+      "epoch": 1.5189309576837418,
+      "grad_norm": 0.2622303223021103,
+      "learning_rate": 2.747524752475248e-05,
+      "loss": 0.3742,
+      "step": 682
+    },
+    {
+      "epoch": 1.5211581291759466,
+      "grad_norm": 0.23462898998214818,
+      "learning_rate": 2.7433993399339937e-05,
+      "loss": 0.3425,
+      "step": 683
+    },
+    {
+      "epoch": 1.5233853006681515,
+      "grad_norm": 0.2416333741824504,
+      "learning_rate": 2.7392739273927392e-05,
+      "loss": 0.3611,
+      "step": 684
+    },
+    {
+      "epoch": 1.5256124721603563,
+      "grad_norm": 0.24498854592268735,
+      "learning_rate": 2.7351485148514853e-05,
+      "loss": 0.3724,
+      "step": 685
+    },
+    {
+      "epoch": 1.5278396436525612,
+      "grad_norm": 0.21533813780649638,
+      "learning_rate": 2.731023102310231e-05,
+      "loss": 0.357,
+      "step": 686
+    },
+    {
+      "epoch": 1.530066815144766,
+      "grad_norm": 0.2198186097543753,
+      "learning_rate": 2.726897689768977e-05,
+      "loss": 0.3496,
+      "step": 687
+    },
+    {
+      "epoch": 1.532293986636971,
+      "grad_norm": 0.21972547732337955,
+      "learning_rate": 2.722772277227723e-05,
+      "loss": 0.3696,
+      "step": 688
+    },
+    {
+      "epoch": 1.534521158129176,
+      "grad_norm": 0.22960568812910168,
+      "learning_rate": 2.7186468646864686e-05,
+      "loss": 0.3638,
+      "step": 689
+    },
+    {
+      "epoch": 1.5367483296213809,
+      "grad_norm": 0.21587924800281283,
+      "learning_rate": 2.7145214521452145e-05,
+      "loss": 0.3461,
+      "step": 690
+    },
+    {
+      "epoch": 1.5389755011135857,
+      "grad_norm": 0.21450364367405003,
+      "learning_rate": 2.7103960396039606e-05,
+      "loss": 0.3505,
+      "step": 691
+    },
+    {
+      "epoch": 1.5412026726057908,
+      "grad_norm": 0.2272515582570843,
+      "learning_rate": 2.7062706270627065e-05,
+      "loss": 0.3566,
+      "step": 692
+    },
+    {
+      "epoch": 1.5434298440979957,
+      "grad_norm": 0.21731442522252906,
+      "learning_rate": 2.702145214521452e-05,
+      "loss": 0.3718,
+      "step": 693
+    },
+    {
+      "epoch": 1.5456570155902005,
+      "grad_norm": 0.23203118652825902,
+      "learning_rate": 2.6980198019801985e-05,
+      "loss": 0.3835,
+      "step": 694
+    },
+    {
+      "epoch": 1.5478841870824054,
+      "grad_norm": 0.24072617248442002,
+      "learning_rate": 2.693894389438944e-05,
+      "loss": 0.3567,
+      "step": 695
+    },
+    {
+      "epoch": 1.5501113585746102,
+      "grad_norm": 0.21211113741060783,
+      "learning_rate": 2.6897689768976898e-05,
+      "loss": 0.3621,
+      "step": 696
+    },
+    {
+      "epoch": 1.552338530066815,
+      "grad_norm": 0.22442534721044624,
+      "learning_rate": 2.685643564356436e-05,
+      "loss": 0.3552,
+      "step": 697
+    },
+    {
+      "epoch": 1.55456570155902,
+      "grad_norm": 0.22886218544287978,
+      "learning_rate": 2.6815181518151818e-05,
+      "loss": 0.3441,
+      "step": 698
+    },
+    {
+      "epoch": 1.5567928730512248,
+      "grad_norm": 0.21190605211030045,
+      "learning_rate": 2.6773927392739272e-05,
+      "loss": 0.3415,
+      "step": 699
+    },
+    {
+      "epoch": 1.5590200445434297,
+      "grad_norm": 0.21397403019649194,
+      "learning_rate": 2.6732673267326734e-05,
+      "loss": 0.3577,
+      "step": 700
+    },
+    {
+      "epoch": 1.5612472160356348,
+      "grad_norm": 0.20798407959268334,
+      "learning_rate": 2.6691419141914192e-05,
+      "loss": 0.3246,
+      "step": 701
+    },
+    {
+      "epoch": 1.5634743875278396,
+      "grad_norm": 0.22674241290780542,
+      "learning_rate": 2.665016501650165e-05,
+      "loss": 0.3308,
+      "step": 702
+    },
+    {
+      "epoch": 1.5657015590200447,
+      "grad_norm": 0.21971871603168083,
+      "learning_rate": 2.6608910891089112e-05,
+      "loss": 0.3386,
+      "step": 703
+    },
+    {
+      "epoch": 1.5679287305122496,
+      "grad_norm": 0.2371853987642058,
+      "learning_rate": 2.6567656765676567e-05,
+      "loss": 0.3496,
+      "step": 704
+    },
+    {
+      "epoch": 1.5701559020044544,
+      "grad_norm": 0.21379269248974136,
+      "learning_rate": 2.6526402640264025e-05,
+      "loss": 0.3481,
+      "step": 705
+    },
+    {
+      "epoch": 1.5723830734966593,
+      "grad_norm": 0.19585786330703275,
+      "learning_rate": 2.6485148514851487e-05,
+      "loss": 0.351,
+      "step": 706
+    },
+    {
+      "epoch": 1.5746102449888641,
+      "grad_norm": 0.2587421310900264,
+      "learning_rate": 2.6443894389438945e-05,
+      "loss": 0.3676,
+      "step": 707
+    },
+    {
+      "epoch": 1.576837416481069,
+      "grad_norm": 0.2585479760875973,
+      "learning_rate": 2.64026402640264e-05,
+      "loss": 0.3861,
+      "step": 708
+    },
+    {
+      "epoch": 1.5790645879732739,
+      "grad_norm": 0.2516216375907583,
+      "learning_rate": 2.6361386138613865e-05,
+      "loss": 0.3582,
+      "step": 709
+    },
+    {
+      "epoch": 1.5812917594654787,
+      "grad_norm": 0.20405466633047462,
+      "learning_rate": 2.632013201320132e-05,
+      "loss": 0.3483,
+      "step": 710
+    },
+    {
+      "epoch": 1.5835189309576836,
+      "grad_norm": 0.2475547863332457,
+      "learning_rate": 2.6278877887788778e-05,
+      "loss": 0.3615,
+      "step": 711
+    },
+    {
+      "epoch": 1.5857461024498887,
+      "grad_norm": 0.2081150380819684,
+      "learning_rate": 2.623762376237624e-05,
+      "loss": 0.3479,
+      "step": 712
+    },
+    {
+      "epoch": 1.5879732739420935,
+      "grad_norm": 0.24448725611838398,
+      "learning_rate": 2.6196369636963698e-05,
+      "loss": 0.3646,
+      "step": 713
+    },
+    {
+      "epoch": 1.5902004454342984,
+      "grad_norm": 0.26812892610271716,
+      "learning_rate": 2.6155115511551153e-05,
+      "loss": 0.3666,
+      "step": 714
+    },
+    {
+      "epoch": 1.5924276169265035,
+      "grad_norm": 0.21579759719455277,
+      "learning_rate": 2.6113861386138615e-05,
+      "loss": 0.3615,
+      "step": 715
+    },
+    {
+      "epoch": 1.5946547884187083,
+      "grad_norm": 0.25509286016118415,
+      "learning_rate": 2.6072607260726073e-05,
+      "loss": 0.3669,
+      "step": 716
+    },
+    {
+      "epoch": 1.5968819599109132,
+      "grad_norm": 0.24970731310233607,
+      "learning_rate": 2.6031353135313535e-05,
+      "loss": 0.3384,
+      "step": 717
+    },
+    {
+      "epoch": 1.599109131403118,
+      "grad_norm": 0.23967809915859692,
+      "learning_rate": 2.5990099009900993e-05,
+      "loss": 0.3748,
+      "step": 718
+    },
+    {
+      "epoch": 1.601336302895323,
+      "grad_norm": 0.2390244120163411,
+      "learning_rate": 2.5948844884488448e-05,
+      "loss": 0.3534,
+      "step": 719
+    },
+    {
+      "epoch": 1.6035634743875278,
+      "grad_norm": 0.2227378080348747,
+      "learning_rate": 2.5907590759075913e-05,
+      "loss": 0.3503,
+      "step": 720
+    },
+    {
+      "epoch": 1.6057906458797326,
+      "grad_norm": 0.22344303559182338,
+      "learning_rate": 2.5866336633663368e-05,
+      "loss": 0.3494,
+      "step": 721
+    },
+    {
+      "epoch": 1.6080178173719375,
+      "grad_norm": 0.2096719404959539,
+      "learning_rate": 2.5825082508250826e-05,
+      "loss": 0.3516,
+      "step": 722
+    },
+    {
+      "epoch": 1.6102449888641426,
+      "grad_norm": 0.23541364148816837,
+      "learning_rate": 2.5783828382838288e-05,
+      "loss": 0.3583,
+      "step": 723
+    },
+    {
+      "epoch": 1.6124721603563474,
+      "grad_norm": 0.24240489602243556,
+      "learning_rate": 2.5742574257425746e-05,
+      "loss": 0.3628,
+      "step": 724
+    },
+    {
+      "epoch": 1.6146993318485523,
+      "grad_norm": 0.23277176159424937,
+      "learning_rate": 2.57013201320132e-05,
+      "loss": 0.3619,
+      "step": 725
+    },
+    {
+      "epoch": 1.6169265033407574,
+      "grad_norm": 0.2261700410987079,
+      "learning_rate": 2.5660066006600662e-05,
+      "loss": 0.3539,
+      "step": 726
+    },
+    {
+      "epoch": 1.6191536748329622,
+      "grad_norm": 0.2475352723888899,
+      "learning_rate": 2.561881188118812e-05,
+      "loss": 0.3638,
+      "step": 727
+    },
+    {
+      "epoch": 1.621380846325167,
+      "grad_norm": 0.2637043817157399,
+      "learning_rate": 2.557755775577558e-05,
+      "loss": 0.3687,
+      "step": 728
+    },
+    {
+      "epoch": 1.623608017817372,
+      "grad_norm": 0.22209926097728006,
+      "learning_rate": 2.553630363036304e-05,
+      "loss": 0.3729,
+      "step": 729
+    },
+    {
+      "epoch": 1.6258351893095768,
+      "grad_norm": 0.22107752502682854,
+      "learning_rate": 2.5495049504950495e-05,
+      "loss": 0.3585,
+      "step": 730
+    },
+    {
+      "epoch": 1.6280623608017817,
+      "grad_norm": 0.23862184087559932,
+      "learning_rate": 2.5453795379537954e-05,
+      "loss": 0.3737,
+      "step": 731
+    },
+    {
+      "epoch": 1.6302895322939865,
+      "grad_norm": 0.22649801341135953,
+      "learning_rate": 2.5412541254125415e-05,
+      "loss": 0.3731,
+      "step": 732
+    },
+    {
+      "epoch": 1.6325167037861914,
+      "grad_norm": 0.22755105603708417,
+      "learning_rate": 2.5371287128712873e-05,
+      "loss": 0.3516,
+      "step": 733
+    },
+    {
+      "epoch": 1.6347438752783965,
+      "grad_norm": 0.19858185617627222,
+      "learning_rate": 2.533003300330033e-05,
+      "loss": 0.3607,
+      "step": 734
+    },
+    {
+      "epoch": 1.6369710467706013,
+      "grad_norm": 0.2297540445090836,
+      "learning_rate": 2.5288778877887793e-05,
+      "loss": 0.3424,
+      "step": 735
+    },
+    {
+      "epoch": 1.6391982182628062,
+      "grad_norm": 0.24870602060429842,
+      "learning_rate": 2.5247524752475248e-05,
+      "loss": 0.3715,
+      "step": 736
+    },
+    {
+      "epoch": 1.6414253897550113,
+      "grad_norm": 0.2024465654655418,
+      "learning_rate": 2.5206270627062707e-05,
+      "loss": 0.3591,
+      "step": 737
+    },
+    {
+      "epoch": 1.6436525612472161,
+      "grad_norm": 0.24668224515250928,
+      "learning_rate": 2.5165016501650168e-05,
+      "loss": 0.355,
+      "step": 738
+    },
+    {
+      "epoch": 1.645879732739421,
+      "grad_norm": 0.19057252993207835,
+      "learning_rate": 2.5123762376237626e-05,
+      "loss": 0.3433,
+      "step": 739
+    },
+    {
+      "epoch": 1.6481069042316259,
+      "grad_norm": 0.21552382920458873,
+      "learning_rate": 2.508250825082508e-05,
+      "loss": 0.3596,
+      "step": 740
+    },
+    {
+      "epoch": 1.6503340757238307,
+      "grad_norm": 0.22755184572320342,
+      "learning_rate": 2.5041254125412543e-05,
+      "loss": 0.3828,
+      "step": 741
+    },
+    {
+      "epoch": 1.6525612472160356,
+      "grad_norm": 0.20334457475049023,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3695,
+      "step": 742
+    },
+    {
+      "epoch": 1.6547884187082404,
+      "grad_norm": 0.22480977110261696,
+      "learning_rate": 2.495874587458746e-05,
+      "loss": 0.384,
+      "step": 743
+    },
+    {
+      "epoch": 1.6570155902004453,
+      "grad_norm": 0.20554152248562454,
+      "learning_rate": 2.4917491749174918e-05,
+      "loss": 0.3502,
+      "step": 744
+    },
+    {
+      "epoch": 1.6592427616926502,
+      "grad_norm": 0.18831042914637963,
+      "learning_rate": 2.4876237623762376e-05,
+      "loss": 0.3398,
+      "step": 745
+    },
+    {
+      "epoch": 1.6614699331848553,
+      "grad_norm": 0.2197300674636681,
+      "learning_rate": 2.4834983498349838e-05,
+      "loss": 0.3325,
+      "step": 746
+    },
+    {
+      "epoch": 1.6636971046770601,
+      "grad_norm": 0.24074656874943445,
+      "learning_rate": 2.4793729372937292e-05,
+      "loss": 0.3802,
+      "step": 747
+    },
+    {
+      "epoch": 1.6659242761692652,
+      "grad_norm": 0.2137682640854006,
+      "learning_rate": 2.4752475247524754e-05,
+      "loss": 0.3511,
+      "step": 748
+    },
+    {
+      "epoch": 1.66815144766147,
+      "grad_norm": 0.1957331739172998,
+      "learning_rate": 2.4711221122112212e-05,
+      "loss": 0.3473,
+      "step": 749
+    },
+    {
+      "epoch": 1.670378619153675,
+      "grad_norm": 0.2164395963159786,
+      "learning_rate": 2.466996699669967e-05,
+      "loss": 0.3789,
+      "step": 750
+    },
+    {
+      "epoch": 1.6726057906458798,
+      "grad_norm": 0.22848414539053694,
+      "learning_rate": 2.462871287128713e-05,
+      "loss": 0.3611,
+      "step": 751
+    },
+    {
+      "epoch": 1.6748329621380846,
+      "grad_norm": 0.19882041874875156,
+      "learning_rate": 2.458745874587459e-05,
+      "loss": 0.3403,
+      "step": 752
+    },
+    {
+      "epoch": 1.6770601336302895,
+      "grad_norm": 0.20825632682323666,
+      "learning_rate": 2.4546204620462045e-05,
+      "loss": 0.3414,
+      "step": 753
+    },
+    {
+      "epoch": 1.6792873051224944,
+      "grad_norm": 0.23873860748628373,
+      "learning_rate": 2.4504950495049507e-05,
+      "loss": 0.3581,
+      "step": 754
+    },
+    {
+      "epoch": 1.6815144766146992,
+      "grad_norm": 0.22543112165583132,
+      "learning_rate": 2.4463696369636965e-05,
+      "loss": 0.3532,
+      "step": 755
+    },
+    {
+      "epoch": 1.683741648106904,
+      "grad_norm": 0.21800601765293173,
+      "learning_rate": 2.4422442244224424e-05,
+      "loss": 0.3676,
+      "step": 756
+    },
+    {
+      "epoch": 1.6859688195991092,
+      "grad_norm": 0.2140720368883988,
+      "learning_rate": 2.4381188118811882e-05,
+      "loss": 0.344,
+      "step": 757
+    },
+    {
+      "epoch": 1.688195991091314,
+      "grad_norm": 0.21486782545441024,
+      "learning_rate": 2.433993399339934e-05,
+      "loss": 0.3472,
+      "step": 758
+    },
+    {
+      "epoch": 1.6904231625835189,
+      "grad_norm": 0.21027644587528624,
+      "learning_rate": 2.4298679867986802e-05,
+      "loss": 0.3577,
+      "step": 759
+    },
+    {
+      "epoch": 1.692650334075724,
+      "grad_norm": 0.21417715876608553,
+      "learning_rate": 2.4257425742574257e-05,
+      "loss": 0.3573,
+      "step": 760
+    },
+    {
+      "epoch": 1.6948775055679288,
+      "grad_norm": 0.21757364057279047,
+      "learning_rate": 2.4216171617161718e-05,
+      "loss": 0.3652,
+      "step": 761
+    },
+    {
+      "epoch": 1.6971046770601337,
+      "grad_norm": 0.22363498217306269,
+      "learning_rate": 2.4174917491749177e-05,
+      "loss": 0.3551,
+      "step": 762
+    },
+    {
+      "epoch": 1.6993318485523385,
+      "grad_norm": 0.2119492697596465,
+      "learning_rate": 2.4133663366336635e-05,
+      "loss": 0.3607,
+      "step": 763
+    },
+    {
+      "epoch": 1.7015590200445434,
+      "grad_norm": 0.2096134105158527,
+      "learning_rate": 2.4092409240924093e-05,
+      "loss": 0.3391,
+      "step": 764
+    },
+    {
+      "epoch": 1.7037861915367483,
+      "grad_norm": 0.20537784757587624,
+      "learning_rate": 2.4051155115511555e-05,
+      "loss": 0.3677,
+      "step": 765
+    },
+    {
+      "epoch": 1.7060133630289531,
+      "grad_norm": 0.23113674389564584,
+      "learning_rate": 2.400990099009901e-05,
+      "loss": 0.3557,
+      "step": 766
+    },
+    {
+      "epoch": 1.708240534521158,
+      "grad_norm": 0.21121564485322036,
+      "learning_rate": 2.396864686468647e-05,
+      "loss": 0.3697,
+      "step": 767
+    },
+    {
+      "epoch": 1.710467706013363,
+      "grad_norm": 0.2258527543992175,
+      "learning_rate": 2.392739273927393e-05,
+      "loss": 0.3503,
+      "step": 768
+    },
+    {
+      "epoch": 1.712694877505568,
+      "grad_norm": 0.22888646361558582,
+      "learning_rate": 2.3886138613861388e-05,
+      "loss": 0.3589,
+      "step": 769
+    },
+    {
+      "epoch": 1.7149220489977728,
+      "grad_norm": 0.20663664144884927,
+      "learning_rate": 2.3844884488448846e-05,
+      "loss": 0.3519,
+      "step": 770
+    },
+    {
+      "epoch": 1.7171492204899779,
+      "grad_norm": 0.19925681560320907,
+      "learning_rate": 2.3803630363036304e-05,
+      "loss": 0.3455,
+      "step": 771
+    },
+    {
+      "epoch": 1.7193763919821827,
+      "grad_norm": 0.20251366706043314,
+      "learning_rate": 2.3762376237623762e-05,
+      "loss": 0.3592,
+      "step": 772
+    },
+    {
+      "epoch": 1.7216035634743876,
+      "grad_norm": 0.22138065506078447,
+      "learning_rate": 2.372112211221122e-05,
+      "loss": 0.3696,
+      "step": 773
+    },
+    {
+      "epoch": 1.7238307349665924,
+      "grad_norm": 0.2165757188295888,
+      "learning_rate": 2.3679867986798682e-05,
+      "loss": 0.3454,
+      "step": 774
+    },
+    {
+      "epoch": 1.7260579064587973,
+      "grad_norm": 0.204436241685765,
+      "learning_rate": 2.3638613861386137e-05,
+      "loss": 0.3504,
+      "step": 775
+    },
+    {
+      "epoch": 1.7282850779510022,
+      "grad_norm": 0.2517471400589238,
+      "learning_rate": 2.35973597359736e-05,
+      "loss": 0.3681,
+      "step": 776
+    },
+    {
+      "epoch": 1.730512249443207,
+      "grad_norm": 0.21082242020594905,
+      "learning_rate": 2.3556105610561057e-05,
+      "loss": 0.3522,
+      "step": 777
+    },
+    {
+      "epoch": 1.732739420935412,
+      "grad_norm": 0.31274971698593285,
+      "learning_rate": 2.3514851485148515e-05,
+      "loss": 0.3392,
+      "step": 778
+    },
+    {
+      "epoch": 1.734966592427617,
+      "grad_norm": 0.21797373750156024,
+      "learning_rate": 2.3473597359735974e-05,
+      "loss": 0.3464,
+      "step": 779
+    },
+    {
+      "epoch": 1.7371937639198218,
+      "grad_norm": 0.22426124926037705,
+      "learning_rate": 2.3432343234323435e-05,
+      "loss": 0.3381,
+      "step": 780
+    },
+    {
+      "epoch": 1.7394209354120267,
+      "grad_norm": 0.20958226019028067,
+      "learning_rate": 2.3391089108910894e-05,
+      "loss": 0.3613,
+      "step": 781
+    },
+    {
+      "epoch": 1.7416481069042318,
+      "grad_norm": 0.21997431509131082,
+      "learning_rate": 2.3349834983498352e-05,
+      "loss": 0.3414,
+      "step": 782
+    },
+    {
+      "epoch": 1.7438752783964366,
+      "grad_norm": 0.2585047283057077,
+      "learning_rate": 2.330858085808581e-05,
+      "loss": 0.3525,
+      "step": 783
+    },
+    {
+      "epoch": 1.7461024498886415,
+      "grad_norm": 0.1973554212874199,
+      "learning_rate": 2.326732673267327e-05,
+      "loss": 0.3453,
+      "step": 784
+    },
+    {
+      "epoch": 1.7483296213808464,
+      "grad_norm": 0.20148204624557856,
+      "learning_rate": 2.3226072607260727e-05,
+      "loss": 0.3506,
+      "step": 785
+    },
+    {
+      "epoch": 1.7505567928730512,
+      "grad_norm": 0.22456688219605642,
+      "learning_rate": 2.3184818481848185e-05,
+      "loss": 0.3706,
+      "step": 786
+    },
+    {
+      "epoch": 1.752783964365256,
+      "grad_norm": 0.23859155562530457,
+      "learning_rate": 2.3143564356435647e-05,
+      "loss": 0.3667,
+      "step": 787
+    },
+    {
+      "epoch": 1.755011135857461,
+      "grad_norm": 0.20273679068782885,
+      "learning_rate": 2.31023102310231e-05,
+      "loss": 0.3559,
+      "step": 788
+    },
+    {
+      "epoch": 1.7572383073496658,
+      "grad_norm": 0.23549903462162344,
+      "learning_rate": 2.3061056105610563e-05,
+      "loss": 0.3572,
+      "step": 789
+    },
+    {
+      "epoch": 1.7594654788418709,
+      "grad_norm": 0.20664546271924084,
+      "learning_rate": 2.301980198019802e-05,
+      "loss": 0.3669,
+      "step": 790
+    },
+    {
+      "epoch": 1.7616926503340757,
+      "grad_norm": 0.18573699358086376,
+      "learning_rate": 2.297854785478548e-05,
+      "loss": 0.3614,
+      "step": 791
+    },
+    {
+      "epoch": 1.7639198218262806,
+      "grad_norm": 0.1903401493341969,
+      "learning_rate": 2.2937293729372938e-05,
+      "loss": 0.3579,
+      "step": 792
+    },
+    {
+      "epoch": 1.7661469933184857,
+      "grad_norm": 0.2294105880619965,
+      "learning_rate": 2.28960396039604e-05,
+      "loss": 0.3821,
+      "step": 793
+    },
+    {
+      "epoch": 1.7683741648106905,
+      "grad_norm": 0.18707069115216823,
+      "learning_rate": 2.2854785478547854e-05,
+      "loss": 0.3464,
+      "step": 794
+    },
+    {
+      "epoch": 1.7706013363028954,
+      "grad_norm": 0.1928806368608763,
+      "learning_rate": 2.2813531353135316e-05,
+      "loss": 0.3471,
+      "step": 795
+    },
+    {
+      "epoch": 1.7728285077951003,
+      "grad_norm": 0.19157137046568395,
+      "learning_rate": 2.2772277227722774e-05,
+      "loss": 0.3484,
+      "step": 796
+    },
+    {
+      "epoch": 1.7750556792873051,
+      "grad_norm": 0.1877096266003065,
+      "learning_rate": 2.2731023102310232e-05,
+      "loss": 0.3427,
+      "step": 797
+    },
+    {
+      "epoch": 1.77728285077951,
+      "grad_norm": 0.23812477838664892,
+      "learning_rate": 2.268976897689769e-05,
+      "loss": 0.3638,
+      "step": 798
+    },
+    {
+      "epoch": 1.7795100222717148,
+      "grad_norm": 0.22681495216210937,
+      "learning_rate": 2.264851485148515e-05,
+      "loss": 0.3379,
+      "step": 799
+    },
+    {
+      "epoch": 1.7817371937639197,
+      "grad_norm": 0.19704186122557504,
+      "learning_rate": 2.2607260726072607e-05,
+      "loss": 0.3397,
+      "step": 800
+    },
+    {
+      "epoch": 1.7839643652561246,
+      "grad_norm": 0.22884211956510753,
+      "learning_rate": 2.2566006600660065e-05,
+      "loss": 0.3714,
+      "step": 801
+    },
+    {
+      "epoch": 1.7861915367483296,
+      "grad_norm": 0.21565251967603208,
+      "learning_rate": 2.2524752475247527e-05,
+      "loss": 0.3491,
+      "step": 802
+    },
+    {
+      "epoch": 1.7884187082405345,
+      "grad_norm": 0.2114423583255959,
+      "learning_rate": 2.2483498349834982e-05,
+      "loss": 0.3499,
+      "step": 803
+    },
+    {
+      "epoch": 1.7906458797327396,
+      "grad_norm": 0.20991719559533167,
+      "learning_rate": 2.2442244224422444e-05,
+      "loss": 0.342,
+      "step": 804
+    },
+    {
+      "epoch": 1.7928730512249444,
+      "grad_norm": 0.20128216471513238,
+      "learning_rate": 2.2400990099009902e-05,
+      "loss": 0.338,
+      "step": 805
+    },
+    {
+      "epoch": 1.7951002227171493,
+      "grad_norm": 0.213830139777128,
+      "learning_rate": 2.2359735973597364e-05,
+      "loss": 0.3618,
+      "step": 806
+    },
+    {
+      "epoch": 1.7973273942093542,
+      "grad_norm": 0.21259676449647513,
+      "learning_rate": 2.231848184818482e-05,
+      "loss": 0.3691,
+      "step": 807
+    },
+    {
+      "epoch": 1.799554565701559,
+      "grad_norm": 0.20495606063114558,
+      "learning_rate": 2.227722772277228e-05,
+      "loss": 0.3513,
+      "step": 808
+    },
+    {
+      "epoch": 1.8017817371937639,
+      "grad_norm": 0.22136308674013255,
+      "learning_rate": 2.223597359735974e-05,
+      "loss": 0.3619,
+      "step": 809
+    },
+    {
+      "epoch": 1.8040089086859687,
+      "grad_norm": 0.20024726503318607,
+      "learning_rate": 2.2194719471947197e-05,
+      "loss": 0.3371,
+      "step": 810
+    },
+    {
+      "epoch": 1.8062360801781736,
+      "grad_norm": 0.21344216526329354,
+      "learning_rate": 2.2153465346534655e-05,
+      "loss": 0.342,
+      "step": 811
+    },
+    {
+      "epoch": 1.8084632516703785,
+      "grad_norm": 0.2088397853784964,
+      "learning_rate": 2.2112211221122113e-05,
+      "loss": 0.3625,
+      "step": 812
+    },
+    {
+      "epoch": 1.8106904231625836,
+      "grad_norm": 0.18187149160175733,
+      "learning_rate": 2.207095709570957e-05,
+      "loss": 0.3346,
+      "step": 813
+    },
+    {
+      "epoch": 1.8129175946547884,
+      "grad_norm": 0.23299910264582777,
+      "learning_rate": 2.202970297029703e-05,
+      "loss": 0.3541,
+      "step": 814
+    },
+    {
+      "epoch": 1.8151447661469933,
+      "grad_norm": 0.20028677312315757,
+      "learning_rate": 2.198844884488449e-05,
+      "loss": 0.3453,
+      "step": 815
+    },
+    {
+      "epoch": 1.8173719376391984,
+      "grad_norm": 0.18263471009850116,
+      "learning_rate": 2.1947194719471946e-05,
+      "loss": 0.3316,
+      "step": 816
+    },
+    {
+      "epoch": 1.8195991091314032,
+      "grad_norm": 0.22062543651918515,
+      "learning_rate": 2.1905940594059408e-05,
+      "loss": 0.3658,
+      "step": 817
+    },
+    {
+      "epoch": 1.821826280623608,
+      "grad_norm": 0.22688623750486514,
+      "learning_rate": 2.1864686468646866e-05,
+      "loss": 0.3511,
+      "step": 818
+    },
+    {
+      "epoch": 1.824053452115813,
+      "grad_norm": 0.1879569517915985,
+      "learning_rate": 2.1823432343234324e-05,
+      "loss": 0.3217,
+      "step": 819
+    },
+    {
+      "epoch": 1.8262806236080178,
+      "grad_norm": 0.1940381066075902,
+      "learning_rate": 2.1782178217821783e-05,
+      "loss": 0.3419,
+      "step": 820
+    },
+    {
+      "epoch": 1.8285077951002227,
+      "grad_norm": 0.2168695271382076,
+      "learning_rate": 2.1740924092409244e-05,
+      "loss": 0.3529,
+      "step": 821
+    },
+    {
+      "epoch": 1.8307349665924275,
+      "grad_norm": 0.20486712169741159,
+      "learning_rate": 2.16996699669967e-05,
+      "loss": 0.3575,
+      "step": 822
+    },
+    {
+      "epoch": 1.8329621380846324,
+      "grad_norm": 0.1883050227368406,
+      "learning_rate": 2.165841584158416e-05,
+      "loss": 0.352,
+      "step": 823
+    },
+    {
+      "epoch": 1.8351893095768375,
+      "grad_norm": 0.23082829161307306,
+      "learning_rate": 2.161716171617162e-05,
+      "loss": 0.3663,
+      "step": 824
+    },
+    {
+      "epoch": 1.8374164810690423,
+      "grad_norm": 0.2307482347765725,
+      "learning_rate": 2.1575907590759077e-05,
+      "loss": 0.3544,
+      "step": 825
+    },
+    {
+      "epoch": 1.8396436525612472,
+      "grad_norm": 0.19843248597663882,
+      "learning_rate": 2.1534653465346535e-05,
+      "loss": 0.3569,
+      "step": 826
+    },
+    {
+      "epoch": 1.8418708240534523,
+      "grad_norm": 0.21195874697924322,
+      "learning_rate": 2.1493399339933994e-05,
+      "loss": 0.3645,
+      "step": 827
+    },
+    {
+      "epoch": 1.8440979955456571,
+      "grad_norm": 0.256220949643074,
+      "learning_rate": 2.1452145214521452e-05,
+      "loss": 0.3742,
+      "step": 828
+    },
+    {
+      "epoch": 1.846325167037862,
+      "grad_norm": 0.22055038762020165,
+      "learning_rate": 2.141089108910891e-05,
+      "loss": 0.348,
+      "step": 829
+    },
+    {
+      "epoch": 1.8485523385300668,
+      "grad_norm": 0.22906582912580156,
+      "learning_rate": 2.1369636963696372e-05,
+      "loss": 0.3366,
+      "step": 830
+    },
+    {
+      "epoch": 1.8507795100222717,
+      "grad_norm": 0.20269610390285273,
+      "learning_rate": 2.132838283828383e-05,
+      "loss": 0.3629,
+      "step": 831
+    },
+    {
+      "epoch": 1.8530066815144766,
+      "grad_norm": 0.20815631614968835,
+      "learning_rate": 2.128712871287129e-05,
+      "loss": 0.3787,
+      "step": 832
+    },
+    {
+      "epoch": 1.8552338530066814,
+      "grad_norm": 0.21411980143547962,
+      "learning_rate": 2.1245874587458747e-05,
+      "loss": 0.3623,
+      "step": 833
+    },
+    {
+      "epoch": 1.8574610244988863,
+      "grad_norm": 0.19020683521295328,
+      "learning_rate": 2.120462046204621e-05,
+      "loss": 0.3499,
+      "step": 834
+    },
+    {
+      "epoch": 1.8596881959910914,
+      "grad_norm": 0.19331331092632276,
+      "learning_rate": 2.1163366336633663e-05,
+      "loss": 0.3631,
+      "step": 835
+    },
+    {
+      "epoch": 1.8619153674832962,
+      "grad_norm": 0.18899933504226676,
+      "learning_rate": 2.1122112211221125e-05,
+      "loss": 0.3448,
+      "step": 836
+    },
+    {
+      "epoch": 1.864142538975501,
+      "grad_norm": 0.21380273552379628,
+      "learning_rate": 2.1080858085808583e-05,
+      "loss": 0.3489,
+      "step": 837
+    },
+    {
+      "epoch": 1.8663697104677062,
+      "grad_norm": 0.18658613921707976,
+      "learning_rate": 2.103960396039604e-05,
+      "loss": 0.3391,
+      "step": 838
+    },
+    {
+      "epoch": 1.868596881959911,
+      "grad_norm": 0.18984197833204985,
+      "learning_rate": 2.09983498349835e-05,
+      "loss": 0.3498,
+      "step": 839
+    },
+    {
+      "epoch": 1.8708240534521159,
+      "grad_norm": 0.19697181465612057,
+      "learning_rate": 2.0957095709570958e-05,
+      "loss": 0.341,
+      "step": 840
+    },
+    {
+      "epoch": 1.8730512249443207,
+      "grad_norm": 0.1931060808637572,
+      "learning_rate": 2.0915841584158416e-05,
+      "loss": 0.3513,
+      "step": 841
+    },
+    {
+      "epoch": 1.8752783964365256,
+      "grad_norm": 0.20233431207039582,
+      "learning_rate": 2.0874587458745874e-05,
+      "loss": 0.3662,
+      "step": 842
+    },
+    {
+      "epoch": 1.8775055679287305,
+      "grad_norm": 0.18786521609001686,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.3614,
+      "step": 843
+    },
+    {
+      "epoch": 1.8797327394209353,
+      "grad_norm": 0.2091422737597329,
+      "learning_rate": 2.079207920792079e-05,
+      "loss": 0.366,
+      "step": 844
+    },
+    {
+      "epoch": 1.8819599109131402,
+      "grad_norm": 0.2150767180245874,
+      "learning_rate": 2.0750825082508253e-05,
+      "loss": 0.3357,
+      "step": 845
+    },
+    {
+      "epoch": 1.884187082405345,
+      "grad_norm": 0.19410541083903396,
+      "learning_rate": 2.070957095709571e-05,
+      "loss": 0.3532,
+      "step": 846
+    },
+    {
+      "epoch": 1.8864142538975501,
+      "grad_norm": 0.19933649104227888,
+      "learning_rate": 2.066831683168317e-05,
+      "loss": 0.3451,
+      "step": 847
+    },
+    {
+      "epoch": 1.888641425389755,
+      "grad_norm": 0.22397787665117305,
+      "learning_rate": 2.0627062706270627e-05,
+      "loss": 0.3613,
+      "step": 848
+    },
+    {
+      "epoch": 1.89086859688196,
+      "grad_norm": 0.5256275726743158,
+      "learning_rate": 2.058580858085809e-05,
+      "loss": 0.3539,
+      "step": 849
+    },
+    {
+      "epoch": 1.893095768374165,
+      "grad_norm": 0.19385487622928163,
+      "learning_rate": 2.0544554455445544e-05,
+      "loss": 0.3422,
+      "step": 850
+    },
+    {
+      "epoch": 1.8953229398663698,
+      "grad_norm": 0.24260306396119488,
+      "learning_rate": 2.0503300330033005e-05,
+      "loss": 0.3573,
+      "step": 851
+    },
+    {
+      "epoch": 1.8975501113585747,
+      "grad_norm": 0.20760454513198467,
+      "learning_rate": 2.0462046204620464e-05,
+      "loss": 0.3542,
+      "step": 852
+    },
+    {
+      "epoch": 1.8997772828507795,
+      "grad_norm": 0.19936257212107614,
+      "learning_rate": 2.0420792079207922e-05,
+      "loss": 0.3383,
+      "step": 853
+    },
+    {
+      "epoch": 1.9020044543429844,
+      "grad_norm": 0.19183978964833118,
+      "learning_rate": 2.037953795379538e-05,
+      "loss": 0.3503,
+      "step": 854
+    },
+    {
+      "epoch": 1.9042316258351892,
+      "grad_norm": 0.21365631815045058,
+      "learning_rate": 2.033828382838284e-05,
+      "loss": 0.3656,
+      "step": 855
+    },
+    {
+      "epoch": 1.906458797327394,
+      "grad_norm": 0.2024022437530225,
+      "learning_rate": 2.02970297029703e-05,
+      "loss": 0.3527,
+      "step": 856
+    },
+    {
+      "epoch": 1.908685968819599,
+      "grad_norm": 0.20226674151760932,
+      "learning_rate": 2.0255775577557755e-05,
+      "loss": 0.3446,
+      "step": 857
+    },
+    {
+      "epoch": 1.910913140311804,
+      "grad_norm": 0.227127328970819,
+      "learning_rate": 2.0214521452145217e-05,
+      "loss": 0.3536,
+      "step": 858
+    },
+    {
+      "epoch": 1.913140311804009,
+      "grad_norm": 0.2197239772687608,
+      "learning_rate": 2.0173267326732675e-05,
+      "loss": 0.3617,
+      "step": 859
+    },
+    {
+      "epoch": 1.9153674832962138,
+      "grad_norm": 0.22087403838276748,
+      "learning_rate": 2.0132013201320133e-05,
+      "loss": 0.3681,
+      "step": 860
+    },
+    {
+      "epoch": 1.9175946547884188,
+      "grad_norm": 0.1995669977450703,
+      "learning_rate": 2.009075907590759e-05,
+      "loss": 0.3358,
+      "step": 861
+    },
+    {
+      "epoch": 1.9198218262806237,
+      "grad_norm": 0.21117604925991532,
+      "learning_rate": 2.0049504950495053e-05,
+      "loss": 0.365,
+      "step": 862
+    },
+    {
+      "epoch": 1.9220489977728286,
+      "grad_norm": 0.21954207043818752,
+      "learning_rate": 2.0008250825082508e-05,
+      "loss": 0.3609,
+      "step": 863
+    },
+    {
+      "epoch": 1.9242761692650334,
+      "grad_norm": 0.22316141925777574,
+      "learning_rate": 1.996699669966997e-05,
+      "loss": 0.3527,
+      "step": 864
+    },
+    {
+      "epoch": 1.9265033407572383,
+      "grad_norm": 0.23643909690979611,
+      "learning_rate": 1.9925742574257428e-05,
+      "loss": 0.3547,
+      "step": 865
+    },
+    {
+      "epoch": 1.9287305122494431,
+      "grad_norm": 0.21579483005725839,
+      "learning_rate": 1.9884488448844886e-05,
+      "loss": 0.3411,
+      "step": 866
+    },
+    {
+      "epoch": 1.930957683741648,
+      "grad_norm": 0.21575285488649312,
+      "learning_rate": 1.9843234323432344e-05,
+      "loss": 0.3427,
+      "step": 867
+    },
+    {
+      "epoch": 1.9331848552338529,
+      "grad_norm": 0.22376247499518298,
+      "learning_rate": 1.9801980198019803e-05,
+      "loss": 0.3519,
+      "step": 868
+    },
+    {
+      "epoch": 1.935412026726058,
+      "grad_norm": 0.21499454175201782,
+      "learning_rate": 1.976072607260726e-05,
+      "loss": 0.3468,
+      "step": 869
+    },
+    {
+      "epoch": 1.9376391982182628,
+      "grad_norm": 0.495687391678954,
+      "learning_rate": 1.971947194719472e-05,
+      "loss": 0.3615,
+      "step": 870
+    },
+    {
+      "epoch": 1.9398663697104677,
+      "grad_norm": 0.20809205707497128,
+      "learning_rate": 1.967821782178218e-05,
+      "loss": 0.3362,
+      "step": 871
+    },
+    {
+      "epoch": 1.9420935412026727,
+      "grad_norm": 0.22896154103126082,
+      "learning_rate": 1.9636963696369636e-05,
+      "loss": 0.3526,
+      "step": 872
+    },
+    {
+      "epoch": 1.9443207126948776,
+      "grad_norm": 0.21096464134793458,
+      "learning_rate": 1.9595709570957097e-05,
+      "loss": 0.3405,
+      "step": 873
+    },
+    {
+      "epoch": 1.9465478841870825,
+      "grad_norm": 0.24066397750177496,
+      "learning_rate": 1.9554455445544556e-05,
+      "loss": 0.3726,
+      "step": 874
+    },
+    {
+      "epoch": 1.9487750556792873,
+      "grad_norm": 0.210675877389854,
+      "learning_rate": 1.9513201320132014e-05,
+      "loss": 0.3632,
+      "step": 875
+    },
+    {
+      "epoch": 1.9510022271714922,
+      "grad_norm": 0.25099975657337337,
+      "learning_rate": 1.9471947194719472e-05,
+      "loss": 0.3477,
+      "step": 876
+    },
+    {
+      "epoch": 1.953229398663697,
+      "grad_norm": 0.21434420701597884,
+      "learning_rate": 1.9430693069306934e-05,
+      "loss": 0.3519,
+      "step": 877
+    },
+    {
+      "epoch": 1.955456570155902,
+      "grad_norm": 0.22837925031767414,
+      "learning_rate": 1.938943894389439e-05,
+      "loss": 0.3518,
+      "step": 878
+    },
+    {
+      "epoch": 1.9576837416481068,
+      "grad_norm": 0.20704495835176406,
+      "learning_rate": 1.934818481848185e-05,
+      "loss": 0.3542,
+      "step": 879
+    },
+    {
+      "epoch": 1.9599109131403119,
+      "grad_norm": 0.2065597872221957,
+      "learning_rate": 1.930693069306931e-05,
+      "loss": 0.3588,
+      "step": 880
+    },
+    {
+      "epoch": 1.9621380846325167,
+      "grad_norm": 0.21160630744172634,
+      "learning_rate": 1.9265676567656767e-05,
+      "loss": 0.3493,
+      "step": 881
+    },
+    {
+      "epoch": 1.9643652561247216,
+      "grad_norm": 0.1982684867710196,
+      "learning_rate": 1.9224422442244225e-05,
+      "loss": 0.3717,
+      "step": 882
+    },
+    {
+      "epoch": 1.9665924276169267,
+      "grad_norm": 0.193658353418164,
+      "learning_rate": 1.9183168316831683e-05,
+      "loss": 0.3363,
+      "step": 883
+    },
+    {
+      "epoch": 1.9688195991091315,
+      "grad_norm": 0.23119980336159962,
+      "learning_rate": 1.9141914191419145e-05,
+      "loss": 0.3672,
+      "step": 884
+    },
+    {
+      "epoch": 1.9710467706013364,
+      "grad_norm": 0.20944985808577066,
+      "learning_rate": 1.91006600660066e-05,
+      "loss": 0.3763,
+      "step": 885
+    },
+    {
+      "epoch": 1.9732739420935412,
+      "grad_norm": 0.21358766726060044,
+      "learning_rate": 1.905940594059406e-05,
+      "loss": 0.3584,
+      "step": 886
+    },
+    {
+      "epoch": 1.975501113585746,
+      "grad_norm": 0.20776733067310027,
+      "learning_rate": 1.901815181518152e-05,
+      "loss": 0.3547,
+      "step": 887
+    },
+    {
+      "epoch": 1.977728285077951,
+      "grad_norm": 0.20491175737408723,
+      "learning_rate": 1.8976897689768978e-05,
+      "loss": 0.3471,
+      "step": 888
+    },
+    {
+      "epoch": 1.9799554565701558,
+      "grad_norm": 0.20282014026788026,
+      "learning_rate": 1.8935643564356436e-05,
+      "loss": 0.3648,
+      "step": 889
+    },
+    {
+      "epoch": 1.9821826280623607,
+      "grad_norm": 0.2366672812165325,
+      "learning_rate": 1.8894389438943898e-05,
+      "loss": 0.3557,
+      "step": 890
+    },
+    {
+      "epoch": 1.9844097995545658,
+      "grad_norm": 0.22040321762718917,
+      "learning_rate": 1.8853135313531353e-05,
+      "loss": 0.3663,
+      "step": 891
+    },
+    {
+      "epoch": 1.9866369710467706,
+      "grad_norm": 0.19386414737029778,
+      "learning_rate": 1.8811881188118814e-05,
+      "loss": 0.3429,
+      "step": 892
+    },
+    {
+      "epoch": 1.9888641425389755,
+      "grad_norm": 0.19057462809948442,
+      "learning_rate": 1.8770627062706273e-05,
+      "loss": 0.3277,
+      "step": 893
+    },
+    {
+      "epoch": 1.9910913140311806,
+      "grad_norm": 0.20415488056283265,
+      "learning_rate": 1.872937293729373e-05,
+      "loss": 0.3436,
+      "step": 894
+    },
+    {
+      "epoch": 1.9933184855233854,
+      "grad_norm": 0.20784568086315625,
+      "learning_rate": 1.868811881188119e-05,
+      "loss": 0.3451,
+      "step": 895
+    },
+    {
+      "epoch": 1.9955456570155903,
+      "grad_norm": 0.20041974234227464,
+      "learning_rate": 1.8646864686468647e-05,
+      "loss": 0.363,
+      "step": 896
+    },
+    {
+      "epoch": 1.9977728285077951,
+      "grad_norm": 0.20635848257127473,
+      "learning_rate": 1.8605610561056106e-05,
+      "loss": 0.3604,
+      "step": 897
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.22254223937545428,
+      "learning_rate": 1.8564356435643564e-05,
+      "loss": 0.3253,
+      "step": 898
+    },
+    {
+      "epoch": 2.002227171492205,
+      "grad_norm": 0.2917879126350186,
+      "learning_rate": 1.8523102310231026e-05,
+      "loss": 0.2855,
+      "step": 899
+    },
+    {
+      "epoch": 2.0044543429844097,
+      "grad_norm": 0.2154422238702394,
+      "learning_rate": 1.848184818481848e-05,
+      "loss": 0.2734,
+      "step": 900
+    },
+    {
+      "epoch": 2.0066815144766146,
+      "grad_norm": 0.3199788710157364,
+      "learning_rate": 1.8440594059405942e-05,
+      "loss": 0.2753,
+      "step": 901
+    },
+    {
+      "epoch": 2.0089086859688194,
+      "grad_norm": 0.2564787157140946,
+      "learning_rate": 1.83993399339934e-05,
+      "loss": 0.2773,
+      "step": 902
+    },
+    {
+      "epoch": 2.0111358574610243,
+      "grad_norm": 0.21607906102561703,
+      "learning_rate": 1.835808580858086e-05,
+      "loss": 0.2644,
+      "step": 903
+    },
+    {
+      "epoch": 2.0133630289532296,
+      "grad_norm": 0.2505134364435312,
+      "learning_rate": 1.8316831683168317e-05,
+      "loss": 0.2655,
+      "step": 904
+    },
+    {
+      "epoch": 2.0155902004454345,
+      "grad_norm": 0.2533151837819191,
+      "learning_rate": 1.827557755775578e-05,
+      "loss": 0.2655,
+      "step": 905
+    },
+    {
+      "epoch": 2.0178173719376393,
+      "grad_norm": 0.21189444791251513,
+      "learning_rate": 1.8234323432343233e-05,
+      "loss": 0.2606,
+      "step": 906
+    },
+    {
+      "epoch": 2.020044543429844,
+      "grad_norm": 0.254765005496945,
+      "learning_rate": 1.8193069306930695e-05,
+      "loss": 0.2748,
+      "step": 907
+    },
+    {
+      "epoch": 2.022271714922049,
+      "grad_norm": 0.2778581181072693,
+      "learning_rate": 1.8151815181518153e-05,
+      "loss": 0.2832,
+      "step": 908
+    },
+    {
+      "epoch": 2.024498886414254,
+      "grad_norm": 0.23046986522656218,
+      "learning_rate": 1.811056105610561e-05,
+      "loss": 0.2866,
+      "step": 909
+    },
+    {
+      "epoch": 2.0267260579064588,
+      "grad_norm": 0.19649792986980408,
+      "learning_rate": 1.806930693069307e-05,
+      "loss": 0.2613,
+      "step": 910
+    },
+    {
+      "epoch": 2.0289532293986636,
+      "grad_norm": 0.2248761596893703,
+      "learning_rate": 1.8028052805280528e-05,
+      "loss": 0.2645,
+      "step": 911
+    },
+    {
+      "epoch": 2.0311804008908685,
+      "grad_norm": 0.22485773246436794,
+      "learning_rate": 1.798679867986799e-05,
+      "loss": 0.2548,
+      "step": 912
+    },
+    {
+      "epoch": 2.0334075723830733,
+      "grad_norm": 0.21140559547488116,
+      "learning_rate": 1.7945544554455445e-05,
+      "loss": 0.2687,
+      "step": 913
+    },
+    {
+      "epoch": 2.035634743875278,
+      "grad_norm": 0.23292737166191993,
+      "learning_rate": 1.7904290429042906e-05,
+      "loss": 0.2822,
+      "step": 914
+    },
+    {
+      "epoch": 2.0378619153674835,
+      "grad_norm": 0.23934152384126353,
+      "learning_rate": 1.7863036303630364e-05,
+      "loss": 0.2916,
+      "step": 915
+    },
+    {
+      "epoch": 2.0400890868596884,
+      "grad_norm": 0.20766845385270943,
+      "learning_rate": 1.7821782178217823e-05,
+      "loss": 0.275,
+      "step": 916
+    },
+    {
+      "epoch": 2.0423162583518932,
+      "grad_norm": 0.24248253295638,
+      "learning_rate": 1.778052805280528e-05,
+      "loss": 0.2693,
+      "step": 917
+    },
+    {
+      "epoch": 2.044543429844098,
+      "grad_norm": 0.23923396143066,
+      "learning_rate": 1.7739273927392743e-05,
+      "loss": 0.2687,
+      "step": 918
+    },
+    {
+      "epoch": 2.046770601336303,
+      "grad_norm": 0.21384650391369592,
+      "learning_rate": 1.7698019801980197e-05,
+      "loss": 0.2594,
+      "step": 919
+    },
+    {
+      "epoch": 2.048997772828508,
+      "grad_norm": 0.21923106830127312,
+      "learning_rate": 1.765676567656766e-05,
+      "loss": 0.2648,
+      "step": 920
+    },
+    {
+      "epoch": 2.0512249443207127,
+      "grad_norm": 0.21920380748549745,
+      "learning_rate": 1.7615511551155117e-05,
+      "loss": 0.2627,
+      "step": 921
+    },
+    {
+      "epoch": 2.0534521158129175,
+      "grad_norm": 0.221655093219312,
+      "learning_rate": 1.7574257425742576e-05,
+      "loss": 0.2895,
+      "step": 922
+    },
+    {
+      "epoch": 2.0556792873051224,
+      "grad_norm": 0.240989269486224,
+      "learning_rate": 1.7533003300330034e-05,
+      "loss": 0.286,
+      "step": 923
+    },
+    {
+      "epoch": 2.0579064587973273,
+      "grad_norm": 0.2265543567040155,
+      "learning_rate": 1.7491749174917492e-05,
+      "loss": 0.2652,
+      "step": 924
+    },
+    {
+      "epoch": 2.060133630289532,
+      "grad_norm": 0.6652483460142316,
+      "learning_rate": 1.745049504950495e-05,
+      "loss": 0.282,
+      "step": 925
+    },
+    {
+      "epoch": 2.062360801781737,
+      "grad_norm": 0.22338210010928283,
+      "learning_rate": 1.740924092409241e-05,
+      "loss": 0.2606,
+      "step": 926
+    },
+    {
+      "epoch": 2.0645879732739423,
+      "grad_norm": 0.2198982670309833,
+      "learning_rate": 1.736798679867987e-05,
+      "loss": 0.2669,
+      "step": 927
+    },
+    {
+      "epoch": 2.066815144766147,
+      "grad_norm": 0.2158641679889132,
+      "learning_rate": 1.7326732673267325e-05,
+      "loss": 0.2836,
+      "step": 928
+    },
+    {
+      "epoch": 2.069042316258352,
+      "grad_norm": 0.23593023008090405,
+      "learning_rate": 1.7285478547854787e-05,
+      "loss": 0.2796,
+      "step": 929
+    },
+    {
+      "epoch": 2.071269487750557,
+      "grad_norm": 0.2411263199249157,
+      "learning_rate": 1.7244224422442245e-05,
+      "loss": 0.2792,
+      "step": 930
+    },
+    {
+      "epoch": 2.0734966592427617,
+      "grad_norm": 0.2245801693901106,
+      "learning_rate": 1.7202970297029703e-05,
+      "loss": 0.2847,
+      "step": 931
+    },
+    {
+      "epoch": 2.0757238307349666,
+      "grad_norm": 0.2196777801290482,
+      "learning_rate": 1.716171617161716e-05,
+      "loss": 0.2721,
+      "step": 932
+    },
+    {
+      "epoch": 2.0779510022271714,
+      "grad_norm": 0.21503454059486582,
+      "learning_rate": 1.7120462046204623e-05,
+      "loss": 0.2577,
+      "step": 933
+    },
+    {
+      "epoch": 2.0801781737193763,
+      "grad_norm": 0.22247608374544567,
+      "learning_rate": 1.707920792079208e-05,
+      "loss": 0.2743,
+      "step": 934
+    },
+    {
+      "epoch": 2.082405345211581,
+      "grad_norm": 0.2181193004044087,
+      "learning_rate": 1.703795379537954e-05,
+      "loss": 0.2724,
+      "step": 935
+    },
+    {
+      "epoch": 2.084632516703786,
+      "grad_norm": 0.22746691098548233,
+      "learning_rate": 1.6996699669966998e-05,
+      "loss": 0.2796,
+      "step": 936
+    },
+    {
+      "epoch": 2.086859688195991,
+      "grad_norm": 0.19996305587329172,
+      "learning_rate": 1.6955445544554456e-05,
+      "loss": 0.2595,
+      "step": 937
+    },
+    {
+      "epoch": 2.089086859688196,
+      "grad_norm": 0.20563128154362556,
+      "learning_rate": 1.6914191419141915e-05,
+      "loss": 0.2652,
+      "step": 938
+    },
+    {
+      "epoch": 2.091314031180401,
+      "grad_norm": 0.2199513127201024,
+      "learning_rate": 1.6872937293729373e-05,
+      "loss": 0.2907,
+      "step": 939
+    },
+    {
+      "epoch": 2.093541202672606,
+      "grad_norm": 0.219015689133294,
+      "learning_rate": 1.6831683168316834e-05,
+      "loss": 0.2844,
+      "step": 940
+    },
+    {
+      "epoch": 2.0957683741648108,
+      "grad_norm": 0.19858942867264778,
+      "learning_rate": 1.679042904290429e-05,
+      "loss": 0.2768,
+      "step": 941
+    },
+    {
+      "epoch": 2.0979955456570156,
+      "grad_norm": 0.19938626187858713,
+      "learning_rate": 1.674917491749175e-05,
+      "loss": 0.2683,
+      "step": 942
+    },
+    {
+      "epoch": 2.1002227171492205,
+      "grad_norm": 0.19250255807127986,
+      "learning_rate": 1.670792079207921e-05,
+      "loss": 0.2515,
+      "step": 943
+    },
+    {
+      "epoch": 2.1024498886414253,
+      "grad_norm": 0.20353899319399785,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2605,
+      "step": 944
+    },
+    {
+      "epoch": 2.10467706013363,
+      "grad_norm": 0.21472333914350644,
+      "learning_rate": 1.6625412541254126e-05,
+      "loss": 0.2594,
+      "step": 945
+    },
+    {
+      "epoch": 2.106904231625835,
+      "grad_norm": 0.19331389223832582,
+      "learning_rate": 1.6584158415841587e-05,
+      "loss": 0.2779,
+      "step": 946
+    },
+    {
+      "epoch": 2.10913140311804,
+      "grad_norm": 0.20336569846190305,
+      "learning_rate": 1.6542904290429042e-05,
+      "loss": 0.279,
+      "step": 947
+    },
+    {
+      "epoch": 2.111358574610245,
+      "grad_norm": 0.21346116075505495,
+      "learning_rate": 1.6501650165016504e-05,
+      "loss": 0.2842,
+      "step": 948
+    },
+    {
+      "epoch": 2.11358574610245,
+      "grad_norm": 0.19001897728741188,
+      "learning_rate": 1.6460396039603962e-05,
+      "loss": 0.2627,
+      "step": 949
+    },
+    {
+      "epoch": 2.115812917594655,
+      "grad_norm": 0.21819996809302278,
+      "learning_rate": 1.641914191419142e-05,
+      "loss": 0.2733,
+      "step": 950
+    },
+    {
+      "epoch": 2.11804008908686,
+      "grad_norm": 0.1927124136398043,
+      "learning_rate": 1.637788778877888e-05,
+      "loss": 0.2734,
+      "step": 951
+    },
+    {
+      "epoch": 2.1202672605790647,
+      "grad_norm": 0.20472312874405912,
+      "learning_rate": 1.6336633663366337e-05,
+      "loss": 0.2704,
+      "step": 952
+    },
+    {
+      "epoch": 2.1224944320712695,
+      "grad_norm": 0.19355755128512284,
+      "learning_rate": 1.6295379537953795e-05,
+      "loss": 0.2652,
+      "step": 953
+    },
+    {
+      "epoch": 2.1247216035634744,
+      "grad_norm": 0.2323231078698614,
+      "learning_rate": 1.6254125412541253e-05,
+      "loss": 0.2987,
+      "step": 954
+    },
+    {
+      "epoch": 2.1269487750556793,
+      "grad_norm": 0.20626477328090762,
+      "learning_rate": 1.6212871287128715e-05,
+      "loss": 0.2719,
+      "step": 955
+    },
+    {
+      "epoch": 2.129175946547884,
+      "grad_norm": 0.21231036913203105,
+      "learning_rate": 1.617161716171617e-05,
+      "loss": 0.279,
+      "step": 956
+    },
+    {
+      "epoch": 2.131403118040089,
+      "grad_norm": 0.20467332985866407,
+      "learning_rate": 1.613036303630363e-05,
+      "loss": 0.2759,
+      "step": 957
+    },
+    {
+      "epoch": 2.133630289532294,
+      "grad_norm": 0.1887795320491789,
+      "learning_rate": 1.608910891089109e-05,
+      "loss": 0.2754,
+      "step": 958
+    },
+    {
+      "epoch": 2.1358574610244987,
+      "grad_norm": 0.1965487030780761,
+      "learning_rate": 1.604785478547855e-05,
+      "loss": 0.2598,
+      "step": 959
+    },
+    {
+      "epoch": 2.138084632516704,
+      "grad_norm": 0.19578496963005818,
+      "learning_rate": 1.6006600660066006e-05,
+      "loss": 0.2726,
+      "step": 960
+    },
+    {
+      "epoch": 2.140311804008909,
+      "grad_norm": 0.1945963650282771,
+      "learning_rate": 1.5965346534653468e-05,
+      "loss": 0.2773,
+      "step": 961
+    },
+    {
+      "epoch": 2.1425389755011137,
+      "grad_norm": 0.19656389161987292,
+      "learning_rate": 1.5924092409240926e-05,
+      "loss": 0.2711,
+      "step": 962
+    },
+    {
+      "epoch": 2.1447661469933186,
+      "grad_norm": 0.2632692695136681,
+      "learning_rate": 1.5882838283828385e-05,
+      "loss": 0.2865,
+      "step": 963
+    },
+    {
+      "epoch": 2.1469933184855234,
+      "grad_norm": 0.20089943924523118,
+      "learning_rate": 1.5841584158415843e-05,
+      "loss": 0.2787,
+      "step": 964
+    },
+    {
+      "epoch": 2.1492204899777283,
+      "grad_norm": 0.2059781904386693,
+      "learning_rate": 1.58003300330033e-05,
+      "loss": 0.2611,
+      "step": 965
+    },
+    {
+      "epoch": 2.151447661469933,
+      "grad_norm": 0.20264777332725245,
+      "learning_rate": 1.575907590759076e-05,
+      "loss": 0.2717,
+      "step": 966
+    },
+    {
+      "epoch": 2.153674832962138,
+      "grad_norm": 0.1899938605683298,
+      "learning_rate": 1.5717821782178218e-05,
+      "loss": 0.245,
+      "step": 967
+    },
+    {
+      "epoch": 2.155902004454343,
+      "grad_norm": 0.22255692718026238,
+      "learning_rate": 1.567656765676568e-05,
+      "loss": 0.2667,
+      "step": 968
+    },
+    {
+      "epoch": 2.1581291759465477,
+      "grad_norm": 0.22124431095213815,
+      "learning_rate": 1.5635313531353134e-05,
+      "loss": 0.2711,
+      "step": 969
+    },
+    {
+      "epoch": 2.1603563474387526,
+      "grad_norm": 0.19003905428878046,
+      "learning_rate": 1.5594059405940596e-05,
+      "loss": 0.258,
+      "step": 970
+    },
+    {
+      "epoch": 2.1625835189309575,
+      "grad_norm": 0.23522626395657878,
+      "learning_rate": 1.5552805280528054e-05,
+      "loss": 0.2757,
+      "step": 971
+    },
+    {
+      "epoch": 2.1648106904231628,
+      "grad_norm": 0.20825230040823375,
+      "learning_rate": 1.5511551155115512e-05,
+      "loss": 0.2798,
+      "step": 972
+    },
+    {
+      "epoch": 2.1670378619153676,
+      "grad_norm": 0.1914440301183656,
+      "learning_rate": 1.547029702970297e-05,
+      "loss": 0.2462,
+      "step": 973
+    },
+    {
+      "epoch": 2.1692650334075725,
+      "grad_norm": 0.19981127087807474,
+      "learning_rate": 1.5429042904290432e-05,
+      "loss": 0.2927,
+      "step": 974
+    },
+    {
+      "epoch": 2.1714922048997773,
+      "grad_norm": 0.1894361460402151,
+      "learning_rate": 1.5387788778877887e-05,
+      "loss": 0.2503,
+      "step": 975
+    },
+    {
+      "epoch": 2.173719376391982,
+      "grad_norm": 0.22245929757102517,
+      "learning_rate": 1.534653465346535e-05,
+      "loss": 0.2685,
+      "step": 976
+    },
+    {
+      "epoch": 2.175946547884187,
+      "grad_norm": 0.19330615950838612,
+      "learning_rate": 1.5305280528052807e-05,
+      "loss": 0.267,
+      "step": 977
+    },
+    {
+      "epoch": 2.178173719376392,
+      "grad_norm": 0.19932638559655444,
+      "learning_rate": 1.5264026402640265e-05,
+      "loss": 0.2598,
+      "step": 978
+    },
+    {
+      "epoch": 2.180400890868597,
+      "grad_norm": 0.21889869568522272,
+      "learning_rate": 1.5222772277227723e-05,
+      "loss": 0.2611,
+      "step": 979
+    },
+    {
+      "epoch": 2.1826280623608016,
+      "grad_norm": 0.19773355190252426,
+      "learning_rate": 1.5181518151815183e-05,
+      "loss": 0.2722,
+      "step": 980
+    },
+    {
+      "epoch": 2.1848552338530065,
+      "grad_norm": 0.22056458426364567,
+      "learning_rate": 1.514026402640264e-05,
+      "loss": 0.2885,
+      "step": 981
+    },
+    {
+      "epoch": 2.187082405345212,
+      "grad_norm": 0.2113264331065395,
+      "learning_rate": 1.50990099009901e-05,
+      "loss": 0.2855,
+      "step": 982
+    },
+    {
+      "epoch": 2.1893095768374167,
+      "grad_norm": 0.19545727618324243,
+      "learning_rate": 1.5057755775577558e-05,
+      "loss": 0.2691,
+      "step": 983
+    },
+    {
+      "epoch": 2.1915367483296215,
+      "grad_norm": 0.20344543295711154,
+      "learning_rate": 1.5016501650165018e-05,
+      "loss": 0.2768,
+      "step": 984
+    },
+    {
+      "epoch": 2.1937639198218264,
+      "grad_norm": 0.2050600328409546,
+      "learning_rate": 1.4975247524752475e-05,
+      "loss": 0.2662,
+      "step": 985
+    },
+    {
+      "epoch": 2.1959910913140313,
+      "grad_norm": 0.1995231599297292,
+      "learning_rate": 1.4933993399339935e-05,
+      "loss": 0.2756,
+      "step": 986
+    },
+    {
+      "epoch": 2.198218262806236,
+      "grad_norm": 0.21638732107895453,
+      "learning_rate": 1.4892739273927395e-05,
+      "loss": 0.2845,
+      "step": 987
+    },
+    {
+      "epoch": 2.200445434298441,
+      "grad_norm": 0.5148923073947419,
+      "learning_rate": 1.4851485148514851e-05,
+      "loss": 0.2676,
+      "step": 988
+    },
+    {
+      "epoch": 2.202672605790646,
+      "grad_norm": 0.1949433992776448,
+      "learning_rate": 1.4810231023102311e-05,
+      "loss": 0.2676,
+      "step": 989
+    },
+    {
+      "epoch": 2.2048997772828507,
+      "grad_norm": 0.21378970538937492,
+      "learning_rate": 1.4768976897689771e-05,
+      "loss": 0.2745,
+      "step": 990
+    },
+    {
+      "epoch": 2.2071269487750556,
+      "grad_norm": 0.21371986128888149,
+      "learning_rate": 1.4727722772277228e-05,
+      "loss": 0.2634,
+      "step": 991
+    },
+    {
+      "epoch": 2.2093541202672604,
+      "grad_norm": 0.20102771902186908,
+      "learning_rate": 1.4686468646864688e-05,
+      "loss": 0.269,
+      "step": 992
+    },
+    {
+      "epoch": 2.2115812917594653,
+      "grad_norm": 0.20157010634064784,
+      "learning_rate": 1.4645214521452147e-05,
+      "loss": 0.2755,
+      "step": 993
+    },
+    {
+      "epoch": 2.2138084632516706,
+      "grad_norm": 0.22829795778649614,
+      "learning_rate": 1.4603960396039604e-05,
+      "loss": 0.2576,
+      "step": 994
+    },
+    {
+      "epoch": 2.2160356347438754,
+      "grad_norm": 0.20055620976746705,
+      "learning_rate": 1.4562706270627064e-05,
+      "loss": 0.2593,
+      "step": 995
+    },
+    {
+      "epoch": 2.2182628062360803,
+      "grad_norm": 0.22009126881477245,
+      "learning_rate": 1.4521452145214522e-05,
+      "loss": 0.2931,
+      "step": 996
+    },
+    {
+      "epoch": 2.220489977728285,
+      "grad_norm": 0.21589312179945735,
+      "learning_rate": 1.448019801980198e-05,
+      "loss": 0.2791,
+      "step": 997
+    },
+    {
+      "epoch": 2.22271714922049,
+      "grad_norm": 0.2107323668966019,
+      "learning_rate": 1.4438943894389439e-05,
+      "loss": 0.266,
+      "step": 998
+    },
+    {
+      "epoch": 2.224944320712695,
+      "grad_norm": 0.21192905027945474,
+      "learning_rate": 1.4397689768976899e-05,
+      "loss": 0.2679,
+      "step": 999
+    },
+    {
+      "epoch": 2.2271714922048997,
+      "grad_norm": 0.20176617353427523,
+      "learning_rate": 1.4356435643564355e-05,
+      "loss": 0.2894,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2293986636971046,
+      "grad_norm": 0.20379631413029126,
+      "learning_rate": 1.4315181518151815e-05,
+      "loss": 0.2545,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2316258351893095,
+      "grad_norm": 0.2042103186102569,
+      "learning_rate": 1.4273927392739275e-05,
+      "loss": 0.2815,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2338530066815143,
+      "grad_norm": 0.21761471948192412,
+      "learning_rate": 1.4232673267326732e-05,
+      "loss": 0.2702,
+      "step": 1003
+    },
+    {
+      "epoch": 2.236080178173719,
+      "grad_norm": 0.20551688021741427,
+      "learning_rate": 1.4191419141914192e-05,
+      "loss": 0.2667,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2383073496659245,
+      "grad_norm": 0.2220623508747115,
+      "learning_rate": 1.4150165016501652e-05,
+      "loss": 0.2953,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2405345211581293,
+      "grad_norm": 0.19576050778193377,
+      "learning_rate": 1.4108910891089108e-05,
+      "loss": 0.2767,
+      "step": 1006
+    },
+    {
+      "epoch": 2.242761692650334,
+      "grad_norm": 0.20687710009830737,
+      "learning_rate": 1.4067656765676568e-05,
+      "loss": 0.2663,
+      "step": 1007
+    },
+    {
+      "epoch": 2.244988864142539,
+      "grad_norm": 0.21049378082401707,
+      "learning_rate": 1.4026402640264028e-05,
+      "loss": 0.2663,
+      "step": 1008
+    },
+    {
+      "epoch": 2.247216035634744,
+      "grad_norm": 0.19273517672626037,
+      "learning_rate": 1.3985148514851486e-05,
+      "loss": 0.2714,
+      "step": 1009
+    },
+    {
+      "epoch": 2.249443207126949,
+      "grad_norm": 0.19519934516938037,
+      "learning_rate": 1.3943894389438945e-05,
+      "loss": 0.2729,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2516703786191536,
+      "grad_norm": 0.2067756058579799,
+      "learning_rate": 1.3902640264026403e-05,
+      "loss": 0.2784,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2538975501113585,
+      "grad_norm": 0.19889505929302467,
+      "learning_rate": 1.3861386138613863e-05,
+      "loss": 0.2604,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2561247216035634,
+      "grad_norm": 0.2022726860047395,
+      "learning_rate": 1.382013201320132e-05,
+      "loss": 0.2866,
+      "step": 1013
+    },
+    {
+      "epoch": 2.2583518930957682,
+      "grad_norm": 0.21628682874250826,
+      "learning_rate": 1.377887788778878e-05,
+      "loss": 0.2783,
+      "step": 1014
+    },
+    {
+      "epoch": 2.260579064587973,
+      "grad_norm": 0.22534223881440016,
+      "learning_rate": 1.373762376237624e-05,
+      "loss": 0.2655,
+      "step": 1015
+    },
+    {
+      "epoch": 2.262806236080178,
+      "grad_norm": 0.2237062248627211,
+      "learning_rate": 1.3696369636963696e-05,
+      "loss": 0.2752,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2650334075723833,
+      "grad_norm": 0.19796277865121822,
+      "learning_rate": 1.3655115511551156e-05,
+      "loss": 0.2841,
+      "step": 1017
+    },
+    {
+      "epoch": 2.267260579064588,
+      "grad_norm": 0.22580292007775335,
+      "learning_rate": 1.3613861386138616e-05,
+      "loss": 0.282,
+      "step": 1018
+    },
+    {
+      "epoch": 2.269487750556793,
+      "grad_norm": 0.23063814482354972,
+      "learning_rate": 1.3572607260726072e-05,
+      "loss": 0.2686,
+      "step": 1019
+    },
+    {
+      "epoch": 2.271714922048998,
+      "grad_norm": 0.22712325781289536,
+      "learning_rate": 1.3531353135313532e-05,
+      "loss": 0.2695,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2739420935412027,
+      "grad_norm": 0.20320569611937253,
+      "learning_rate": 1.3490099009900992e-05,
+      "loss": 0.2627,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2761692650334076,
+      "grad_norm": 0.22639429235101766,
+      "learning_rate": 1.3448844884488449e-05,
+      "loss": 0.2766,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2783964365256124,
+      "grad_norm": 0.2149149957786946,
+      "learning_rate": 1.3407590759075909e-05,
+      "loss": 0.2808,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2806236080178173,
+      "grad_norm": 0.20731196155393639,
+      "learning_rate": 1.3366336633663367e-05,
+      "loss": 0.2559,
+      "step": 1024
+    },
+    {
+      "epoch": 2.282850779510022,
+      "grad_norm": 0.20039508340783346,
+      "learning_rate": 1.3325082508250825e-05,
+      "loss": 0.2634,
+      "step": 1025
+    },
+    {
+      "epoch": 2.285077951002227,
+      "grad_norm": 0.21716535787151417,
+      "learning_rate": 1.3283828382838284e-05,
+      "loss": 0.2649,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2873051224944323,
+      "grad_norm": 0.19008765090691132,
+      "learning_rate": 1.3242574257425743e-05,
+      "loss": 0.2655,
+      "step": 1027
+    },
+    {
+      "epoch": 2.289532293986637,
+      "grad_norm": 0.21202624251215715,
+      "learning_rate": 1.32013201320132e-05,
+      "loss": 0.2817,
+      "step": 1028
+    },
+    {
+      "epoch": 2.291759465478842,
+      "grad_norm": 0.21711447076279528,
+      "learning_rate": 1.316006600660066e-05,
+      "loss": 0.2757,
+      "step": 1029
+    },
+    {
+      "epoch": 2.293986636971047,
+      "grad_norm": 0.2053491511971421,
+      "learning_rate": 1.311881188118812e-05,
+      "loss": 0.2814,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2962138084632517,
+      "grad_norm": 0.17966000787705147,
+      "learning_rate": 1.3077557755775577e-05,
+      "loss": 0.249,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2984409799554566,
+      "grad_norm": 0.2097272539902852,
+      "learning_rate": 1.3036303630363036e-05,
+      "loss": 0.2816,
+      "step": 1032
+    },
+    {
+      "epoch": 2.3006681514476615,
+      "grad_norm": 0.22355137552793322,
+      "learning_rate": 1.2995049504950496e-05,
+      "loss": 0.2949,
+      "step": 1033
+    },
+    {
+      "epoch": 2.3028953229398663,
+      "grad_norm": 0.1973752759660782,
+      "learning_rate": 1.2953795379537956e-05,
+      "loss": 0.2729,
+      "step": 1034
+    },
+    {
+      "epoch": 2.305122494432071,
+      "grad_norm": 0.1938493711007142,
+      "learning_rate": 1.2912541254125413e-05,
+      "loss": 0.2748,
+      "step": 1035
+    },
+    {
+      "epoch": 2.307349665924276,
+      "grad_norm": 0.18013326195522006,
+      "learning_rate": 1.2871287128712873e-05,
+      "loss": 0.2577,
+      "step": 1036
+    },
+    {
+      "epoch": 2.309576837416481,
+      "grad_norm": 0.212280614898667,
+      "learning_rate": 1.2830033003300331e-05,
+      "loss": 0.2782,
+      "step": 1037
+    },
+    {
+      "epoch": 2.3118040089086858,
+      "grad_norm": 0.22611659220287397,
+      "learning_rate": 1.278877887788779e-05,
+      "loss": 0.2817,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3140311804008906,
+      "grad_norm": 0.18816070651441863,
+      "learning_rate": 1.2747524752475248e-05,
+      "loss": 0.272,
+      "step": 1039
+    },
+    {
+      "epoch": 2.316258351893096,
+      "grad_norm": 0.20816294895034484,
+      "learning_rate": 1.2706270627062708e-05,
+      "loss": 0.2707,
+      "step": 1040
+    },
+    {
+      "epoch": 2.318485523385301,
+      "grad_norm": 0.23837725326477052,
+      "learning_rate": 1.2665016501650164e-05,
+      "loss": 0.2921,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3207126948775056,
+      "grad_norm": 0.19486477654916956,
+      "learning_rate": 1.2623762376237624e-05,
+      "loss": 0.2708,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3229398663697105,
+      "grad_norm": 0.20569445864400387,
+      "learning_rate": 1.2582508250825084e-05,
+      "loss": 0.2925,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3251670378619154,
+      "grad_norm": 0.20033809716151726,
+      "learning_rate": 1.254125412541254e-05,
+      "loss": 0.2689,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3273942093541202,
+      "grad_norm": 0.19824467122341974,
+      "learning_rate": 1.25e-05,
+      "loss": 0.2694,
+      "step": 1045
+    },
+    {
+      "epoch": 2.329621380846325,
+      "grad_norm": 0.18892618648996176,
+      "learning_rate": 1.2458745874587459e-05,
+      "loss": 0.2724,
+      "step": 1046
+    },
+    {
+      "epoch": 2.33184855233853,
+      "grad_norm": 0.1850645919338069,
+      "learning_rate": 1.2417491749174919e-05,
+      "loss": 0.2525,
+      "step": 1047
+    },
+    {
+      "epoch": 2.334075723830735,
+      "grad_norm": 0.21133712560743997,
+      "learning_rate": 1.2376237623762377e-05,
+      "loss": 0.2827,
+      "step": 1048
+    },
+    {
+      "epoch": 2.33630289532294,
+      "grad_norm": 0.18596901788858,
+      "learning_rate": 1.2334983498349835e-05,
+      "loss": 0.255,
+      "step": 1049
+    },
+    {
+      "epoch": 2.338530066815145,
+      "grad_norm": 0.18710426212811665,
+      "learning_rate": 1.2293729372937295e-05,
+      "loss": 0.2571,
+      "step": 1050
+    },
+    {
+      "epoch": 2.34075723830735,
+      "grad_norm": 0.2056915328832616,
+      "learning_rate": 1.2252475247524754e-05,
+      "loss": 0.2565,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3429844097995547,
+      "grad_norm": 0.21136713475340663,
+      "learning_rate": 1.2211221122112212e-05,
+      "loss": 0.2625,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3452115812917596,
+      "grad_norm": 0.19210427375018393,
+      "learning_rate": 1.216996699669967e-05,
+      "loss": 0.2613,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3474387527839644,
+      "grad_norm": 0.1980726074792816,
+      "learning_rate": 1.2128712871287128e-05,
+      "loss": 0.2671,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3496659242761693,
+      "grad_norm": 0.19016273244979234,
+      "learning_rate": 1.2087458745874588e-05,
+      "loss": 0.2812,
+      "step": 1055
+    },
+    {
+      "epoch": 2.351893095768374,
+      "grad_norm": 0.19642715940775765,
+      "learning_rate": 1.2046204620462047e-05,
+      "loss": 0.2542,
+      "step": 1056
+    },
+    {
+      "epoch": 2.354120267260579,
+      "grad_norm": 0.19762415084293058,
+      "learning_rate": 1.2004950495049505e-05,
+      "loss": 0.2678,
+      "step": 1057
+    },
+    {
+      "epoch": 2.356347438752784,
+      "grad_norm": 0.1937697142736061,
+      "learning_rate": 1.1963696369636965e-05,
+      "loss": 0.2835,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3585746102449887,
+      "grad_norm": 0.20404730565329052,
+      "learning_rate": 1.1922442244224423e-05,
+      "loss": 0.268,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3608017817371936,
+      "grad_norm": 0.19264558829999212,
+      "learning_rate": 1.1881188118811881e-05,
+      "loss": 0.2705,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3630289532293984,
+      "grad_norm": 0.18612729824589969,
+      "learning_rate": 1.1839933993399341e-05,
+      "loss": 0.2655,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3652561247216037,
+      "grad_norm": 0.19184627981030214,
+      "learning_rate": 1.17986798679868e-05,
+      "loss": 0.2739,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3674832962138086,
+      "grad_norm": 0.20834509457158706,
+      "learning_rate": 1.1757425742574258e-05,
+      "loss": 0.2884,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3697104677060135,
+      "grad_norm": 0.20972582839416928,
+      "learning_rate": 1.1716171617161718e-05,
+      "loss": 0.2699,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3719376391982183,
+      "grad_norm": 0.1830305603494533,
+      "learning_rate": 1.1674917491749176e-05,
+      "loss": 0.2685,
+      "step": 1065
+    },
+    {
+      "epoch": 2.374164810690423,
+      "grad_norm": 0.19984211566451557,
+      "learning_rate": 1.1633663366336634e-05,
+      "loss": 0.2777,
+      "step": 1066
+    },
+    {
+      "epoch": 2.376391982182628,
+      "grad_norm": 0.19625399200236907,
+      "learning_rate": 1.1592409240924092e-05,
+      "loss": 0.2878,
+      "step": 1067
+    },
+    {
+      "epoch": 2.378619153674833,
+      "grad_norm": 0.18837463604759586,
+      "learning_rate": 1.155115511551155e-05,
+      "loss": 0.2721,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3808463251670378,
+      "grad_norm": 0.179269880400689,
+      "learning_rate": 1.150990099009901e-05,
+      "loss": 0.2597,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3830734966592426,
+      "grad_norm": 0.17905269452030195,
+      "learning_rate": 1.1468646864686469e-05,
+      "loss": 0.2669,
+      "step": 1070
+    },
+    {
+      "epoch": 2.3853006681514475,
+      "grad_norm": 0.18871772253331306,
+      "learning_rate": 1.1427392739273927e-05,
+      "loss": 0.2509,
+      "step": 1071
+    },
+    {
+      "epoch": 2.387527839643653,
+      "grad_norm": 0.18408752726862213,
+      "learning_rate": 1.1386138613861387e-05,
+      "loss": 0.2645,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3897550111358576,
+      "grad_norm": 0.1931246236770583,
+      "learning_rate": 1.1344884488448845e-05,
+      "loss": 0.27,
+      "step": 1073
+    },
+    {
+      "epoch": 2.3919821826280625,
+      "grad_norm": 0.20313530066494817,
+      "learning_rate": 1.1303630363036304e-05,
+      "loss": 0.2774,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3942093541202674,
+      "grad_norm": 0.1877854318454684,
+      "learning_rate": 1.1262376237623764e-05,
+      "loss": 0.2631,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3964365256124722,
+      "grad_norm": 0.18530096780240002,
+      "learning_rate": 1.1221122112211222e-05,
+      "loss": 0.254,
+      "step": 1076
+    },
+    {
+      "epoch": 2.398663697104677,
+      "grad_norm": 0.18435642955392806,
+      "learning_rate": 1.1179867986798682e-05,
+      "loss": 0.2558,
+      "step": 1077
+    },
+    {
+      "epoch": 2.400890868596882,
+      "grad_norm": 0.19898779382739942,
+      "learning_rate": 1.113861386138614e-05,
+      "loss": 0.288,
+      "step": 1078
+    },
+    {
+      "epoch": 2.403118040089087,
+      "grad_norm": 0.2276311483736436,
+      "learning_rate": 1.1097359735973598e-05,
+      "loss": 0.2944,
+      "step": 1079
+    },
+    {
+      "epoch": 2.4053452115812917,
+      "grad_norm": 0.18877965389508786,
+      "learning_rate": 1.1056105610561057e-05,
+      "loss": 0.2536,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4075723830734965,
+      "grad_norm": 0.21066634926735692,
+      "learning_rate": 1.1014851485148515e-05,
+      "loss": 0.2926,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4097995545657014,
+      "grad_norm": 0.18422535941666346,
+      "learning_rate": 1.0973597359735973e-05,
+      "loss": 0.2662,
+      "step": 1082
+    },
+    {
+      "epoch": 2.4120267260579062,
+      "grad_norm": 0.19053201856845206,
+      "learning_rate": 1.0932343234323433e-05,
+      "loss": 0.2668,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4142538975501115,
+      "grad_norm": 0.18892660093681088,
+      "learning_rate": 1.0891089108910891e-05,
+      "loss": 0.262,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4164810690423164,
+      "grad_norm": 0.18924935218566152,
+      "learning_rate": 1.084983498349835e-05,
+      "loss": 0.2618,
+      "step": 1085
+    },
+    {
+      "epoch": 2.4187082405345213,
+      "grad_norm": 0.18575274067232,
+      "learning_rate": 1.080858085808581e-05,
+      "loss": 0.2659,
+      "step": 1086
+    },
+    {
+      "epoch": 2.420935412026726,
+      "grad_norm": 0.19770032474640944,
+      "learning_rate": 1.0767326732673268e-05,
+      "loss": 0.2694,
+      "step": 1087
+    },
+    {
+      "epoch": 2.423162583518931,
+      "grad_norm": 0.1871465004147872,
+      "learning_rate": 1.0726072607260726e-05,
+      "loss": 0.2653,
+      "step": 1088
+    },
+    {
+      "epoch": 2.425389755011136,
+      "grad_norm": 0.19613520681940685,
+      "learning_rate": 1.0684818481848186e-05,
+      "loss": 0.2596,
+      "step": 1089
+    },
+    {
+      "epoch": 2.4276169265033407,
+      "grad_norm": 0.2013804064771092,
+      "learning_rate": 1.0643564356435644e-05,
+      "loss": 0.2744,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4298440979955456,
+      "grad_norm": 0.18448807360998867,
+      "learning_rate": 1.0602310231023104e-05,
+      "loss": 0.2739,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4320712694877504,
+      "grad_norm": 0.17436121469164215,
+      "learning_rate": 1.0561056105610562e-05,
+      "loss": 0.2556,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4342984409799553,
+      "grad_norm": 0.20230038331757244,
+      "learning_rate": 1.051980198019802e-05,
+      "loss": 0.2775,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4365256124721606,
+      "grad_norm": 0.19739011178754348,
+      "learning_rate": 1.0478547854785479e-05,
+      "loss": 0.2699,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4387527839643655,
+      "grad_norm": 0.19593366852721233,
+      "learning_rate": 1.0437293729372937e-05,
+      "loss": 0.2626,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4409799554565703,
+      "grad_norm": 0.20418637427016842,
+      "learning_rate": 1.0396039603960395e-05,
+      "loss": 0.2766,
+      "step": 1096
+    },
+    {
+      "epoch": 2.443207126948775,
+      "grad_norm": 0.1787939591686354,
+      "learning_rate": 1.0354785478547855e-05,
+      "loss": 0.2552,
+      "step": 1097
+    },
+    {
+      "epoch": 2.44543429844098,
+      "grad_norm": 0.1922023637484877,
+      "learning_rate": 1.0313531353135314e-05,
+      "loss": 0.2639,
+      "step": 1098
+    },
+    {
+      "epoch": 2.447661469933185,
+      "grad_norm": 0.18990449404250637,
+      "learning_rate": 1.0272277227722772e-05,
+      "loss": 0.2554,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4498886414253898,
+      "grad_norm": 0.19231913714571738,
+      "learning_rate": 1.0231023102310232e-05,
+      "loss": 0.2703,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4521158129175946,
+      "grad_norm": 0.18439832863625685,
+      "learning_rate": 1.018976897689769e-05,
+      "loss": 0.2708,
+      "step": 1101
+    },
+    {
+      "epoch": 2.4543429844097995,
+      "grad_norm": 0.1825082756343376,
+      "learning_rate": 1.014851485148515e-05,
+      "loss": 0.2663,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4565701559020043,
+      "grad_norm": 0.19271366095212672,
+      "learning_rate": 1.0107260726072608e-05,
+      "loss": 0.2928,
+      "step": 1103
+    },
+    {
+      "epoch": 2.458797327394209,
+      "grad_norm": 0.1929337738803978,
+      "learning_rate": 1.0066006600660067e-05,
+      "loss": 0.2701,
+      "step": 1104
+    },
+    {
+      "epoch": 2.461024498886414,
+      "grad_norm": 0.18674110929001914,
+      "learning_rate": 1.0024752475247527e-05,
+      "loss": 0.2718,
+      "step": 1105
+    },
+    {
+      "epoch": 2.463251670378619,
+      "grad_norm": 0.1742958795479238,
+      "learning_rate": 9.983498349834985e-06,
+      "loss": 0.2607,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4654788418708242,
+      "grad_norm": 0.19124442287313453,
+      "learning_rate": 9.942244224422443e-06,
+      "loss": 0.2841,
+      "step": 1107
+    },
+    {
+      "epoch": 2.467706013363029,
+      "grad_norm": 0.17466533749128837,
+      "learning_rate": 9.900990099009901e-06,
+      "loss": 0.2569,
+      "step": 1108
+    },
+    {
+      "epoch": 2.469933184855234,
+      "grad_norm": 0.18291754233017415,
+      "learning_rate": 9.85973597359736e-06,
+      "loss": 0.2658,
+      "step": 1109
+    },
+    {
+      "epoch": 2.472160356347439,
+      "grad_norm": 0.18788097066997822,
+      "learning_rate": 9.818481848184818e-06,
+      "loss": 0.271,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4743875278396437,
+      "grad_norm": 0.19428141706955826,
+      "learning_rate": 9.777227722772278e-06,
+      "loss": 0.2794,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4766146993318485,
+      "grad_norm": 0.19910230370842083,
+      "learning_rate": 9.735973597359736e-06,
+      "loss": 0.2833,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4788418708240534,
+      "grad_norm": 0.19683763848878325,
+      "learning_rate": 9.694719471947194e-06,
+      "loss": 0.2785,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4810690423162582,
+      "grad_norm": 0.19801756868357392,
+      "learning_rate": 9.653465346534654e-06,
+      "loss": 0.2719,
+      "step": 1114
+    },
+    {
+      "epoch": 2.483296213808463,
+      "grad_norm": 0.1826730168645165,
+      "learning_rate": 9.612211221122113e-06,
+      "loss": 0.2584,
+      "step": 1115
+    },
+    {
+      "epoch": 2.485523385300668,
+      "grad_norm": 0.19696852384395833,
+      "learning_rate": 9.570957095709572e-06,
+      "loss": 0.2722,
+      "step": 1116
+    },
+    {
+      "epoch": 2.4877505567928733,
+      "grad_norm": 0.18950166276846098,
+      "learning_rate": 9.52970297029703e-06,
+      "loss": 0.2706,
+      "step": 1117
+    },
+    {
+      "epoch": 2.489977728285078,
+      "grad_norm": 0.1844705006714394,
+      "learning_rate": 9.488448844884489e-06,
+      "loss": 0.2563,
+      "step": 1118
+    },
+    {
+      "epoch": 2.492204899777283,
+      "grad_norm": 0.19868072718137894,
+      "learning_rate": 9.447194719471949e-06,
+      "loss": 0.2684,
+      "step": 1119
+    },
+    {
+      "epoch": 2.494432071269488,
+      "grad_norm": 0.2066675942912551,
+      "learning_rate": 9.405940594059407e-06,
+      "loss": 0.2815,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4966592427616927,
+      "grad_norm": 0.20088816875352536,
+      "learning_rate": 9.364686468646865e-06,
+      "loss": 0.2746,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4988864142538976,
+      "grad_norm": 0.21141165012413876,
+      "learning_rate": 9.323432343234324e-06,
+      "loss": 0.2738,
+      "step": 1122
+    },
+    {
+      "epoch": 2.5011135857461024,
+      "grad_norm": 0.19699453704906772,
+      "learning_rate": 9.282178217821782e-06,
+      "loss": 0.27,
+      "step": 1123
+    },
+    {
+      "epoch": 2.5033407572383073,
+      "grad_norm": 0.184869259863741,
+      "learning_rate": 9.24092409240924e-06,
+      "loss": 0.2629,
+      "step": 1124
+    },
+    {
+      "epoch": 2.505567928730512,
+      "grad_norm": 0.20304155160431278,
+      "learning_rate": 9.1996699669967e-06,
+      "loss": 0.2603,
+      "step": 1125
+    },
+    {
+      "epoch": 2.507795100222717,
+      "grad_norm": 0.21305764778554145,
+      "learning_rate": 9.158415841584158e-06,
+      "loss": 0.2747,
+      "step": 1126
+    },
+    {
+      "epoch": 2.510022271714922,
+      "grad_norm": 0.22852912068589257,
+      "learning_rate": 9.117161716171617e-06,
+      "loss": 0.2725,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5122494432071267,
+      "grad_norm": 0.19848187457065516,
+      "learning_rate": 9.075907590759077e-06,
+      "loss": 0.2808,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5144766146993316,
+      "grad_norm": 0.1913588513644944,
+      "learning_rate": 9.034653465346535e-06,
+      "loss": 0.2631,
+      "step": 1129
+    },
+    {
+      "epoch": 2.516703786191537,
+      "grad_norm": 0.40151960277216947,
+      "learning_rate": 8.993399339933995e-06,
+      "loss": 0.2795,
+      "step": 1130
+    },
+    {
+      "epoch": 2.5189309576837418,
+      "grad_norm": 0.20250778837370625,
+      "learning_rate": 8.952145214521453e-06,
+      "loss": 0.2801,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5211581291759466,
+      "grad_norm": 0.20334890585637636,
+      "learning_rate": 8.910891089108911e-06,
+      "loss": 0.2813,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5233853006681515,
+      "grad_norm": 0.18704973977774514,
+      "learning_rate": 8.869636963696371e-06,
+      "loss": 0.2719,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5256124721603563,
+      "grad_norm": 0.19467440557527957,
+      "learning_rate": 8.82838283828383e-06,
+      "loss": 0.272,
+      "step": 1134
+    },
+    {
+      "epoch": 2.527839643652561,
+      "grad_norm": 0.2008853859089321,
+      "learning_rate": 8.787128712871288e-06,
+      "loss": 0.286,
+      "step": 1135
+    },
+    {
+      "epoch": 2.530066815144766,
+      "grad_norm": 0.19458679714398816,
+      "learning_rate": 8.745874587458746e-06,
+      "loss": 0.265,
+      "step": 1136
+    },
+    {
+      "epoch": 2.532293986636971,
+      "grad_norm": 0.20465986625053167,
+      "learning_rate": 8.704620462046204e-06,
+      "loss": 0.2813,
+      "step": 1137
+    },
+    {
+      "epoch": 2.534521158129176,
+      "grad_norm": 0.1895991283920018,
+      "learning_rate": 8.663366336633663e-06,
+      "loss": 0.2659,
+      "step": 1138
+    },
+    {
+      "epoch": 2.536748329621381,
+      "grad_norm": 0.1879019267908632,
+      "learning_rate": 8.622112211221123e-06,
+      "loss": 0.2705,
+      "step": 1139
+    },
+    {
+      "epoch": 2.538975501113586,
+      "grad_norm": 0.19743356419796973,
+      "learning_rate": 8.58085808580858e-06,
+      "loss": 0.2704,
+      "step": 1140
+    },
+    {
+      "epoch": 2.541202672605791,
+      "grad_norm": 0.1883376854672621,
+      "learning_rate": 8.53960396039604e-06,
+      "loss": 0.275,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5434298440979957,
+      "grad_norm": 0.17992466715149383,
+      "learning_rate": 8.498349834983499e-06,
+      "loss": 0.2651,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5456570155902005,
+      "grad_norm": 0.17938120420166848,
+      "learning_rate": 8.457095709570957e-06,
+      "loss": 0.2763,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5478841870824054,
+      "grad_norm": 0.1766464605464754,
+      "learning_rate": 8.415841584158417e-06,
+      "loss": 0.2814,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5501113585746102,
+      "grad_norm": 0.1733272025991543,
+      "learning_rate": 8.374587458745875e-06,
+      "loss": 0.2505,
+      "step": 1145
+    },
+    {
+      "epoch": 2.552338530066815,
+      "grad_norm": 0.19171087486125246,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2669,
+      "step": 1146
+    },
+    {
+      "epoch": 2.55456570155902,
+      "grad_norm": 0.20113910144003733,
+      "learning_rate": 8.292079207920794e-06,
+      "loss": 0.2743,
+      "step": 1147
+    },
+    {
+      "epoch": 2.556792873051225,
+      "grad_norm": 0.188959196450275,
+      "learning_rate": 8.250825082508252e-06,
+      "loss": 0.2844,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5590200445434297,
+      "grad_norm": 0.1787093466409625,
+      "learning_rate": 8.20957095709571e-06,
+      "loss": 0.2821,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5612472160356345,
+      "grad_norm": 0.1942401995026469,
+      "learning_rate": 8.168316831683168e-06,
+      "loss": 0.2829,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5634743875278394,
+      "grad_norm": 0.19634490928315082,
+      "learning_rate": 8.127062706270627e-06,
+      "loss": 0.2826,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5657015590200447,
+      "grad_norm": 0.19079365604694734,
+      "learning_rate": 8.085808580858085e-06,
+      "loss": 0.2671,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5679287305122496,
+      "grad_norm": 0.18625347494470584,
+      "learning_rate": 8.044554455445545e-06,
+      "loss": 0.2704,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5701559020044544,
+      "grad_norm": 0.19152249857310585,
+      "learning_rate": 8.003300330033003e-06,
+      "loss": 0.2607,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5723830734966593,
+      "grad_norm": 0.21034174122909713,
+      "learning_rate": 7.962046204620463e-06,
+      "loss": 0.2776,
+      "step": 1155
+    },
+    {
+      "epoch": 2.574610244988864,
+      "grad_norm": 0.19270298450827533,
+      "learning_rate": 7.920792079207921e-06,
+      "loss": 0.2633,
+      "step": 1156
+    },
+    {
+      "epoch": 2.576837416481069,
+      "grad_norm": 0.19418933010831285,
+      "learning_rate": 7.87953795379538e-06,
+      "loss": 0.2819,
+      "step": 1157
+    },
+    {
+      "epoch": 2.579064587973274,
+      "grad_norm": 0.1774967603891648,
+      "learning_rate": 7.83828382838284e-06,
+      "loss": 0.268,
+      "step": 1158
+    },
+    {
+      "epoch": 2.5812917594654787,
+      "grad_norm": 0.1952783384290122,
+      "learning_rate": 7.797029702970298e-06,
+      "loss": 0.2714,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5835189309576836,
+      "grad_norm": 0.18333390666789304,
+      "learning_rate": 7.755775577557756e-06,
+      "loss": 0.2542,
+      "step": 1160
+    },
+    {
+      "epoch": 2.585746102449889,
+      "grad_norm": 0.1915948045097857,
+      "learning_rate": 7.714521452145216e-06,
+      "loss": 0.2665,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5879732739420938,
+      "grad_norm": 0.18672447338790377,
+      "learning_rate": 7.673267326732674e-06,
+      "loss": 0.2584,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5902004454342986,
+      "grad_norm": 0.1829500342139928,
+      "learning_rate": 7.632013201320133e-06,
+      "loss": 0.2745,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5924276169265035,
+      "grad_norm": 0.18799104968969643,
+      "learning_rate": 7.590759075907592e-06,
+      "loss": 0.2627,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5946547884187083,
+      "grad_norm": 0.19458221392506975,
+      "learning_rate": 7.54950495049505e-06,
+      "loss": 0.2762,
+      "step": 1165
+    },
+    {
+      "epoch": 2.596881959910913,
+      "grad_norm": 0.191352735133007,
+      "learning_rate": 7.508250825082509e-06,
+      "loss": 0.281,
+      "step": 1166
+    },
+    {
+      "epoch": 2.599109131403118,
+      "grad_norm": 0.17408507908431892,
+      "learning_rate": 7.466996699669967e-06,
+      "loss": 0.2612,
+      "step": 1167
+    },
+    {
+      "epoch": 2.601336302895323,
+      "grad_norm": 0.186054208066729,
+      "learning_rate": 7.4257425742574256e-06,
+      "loss": 0.2736,
+      "step": 1168
+    },
+    {
+      "epoch": 2.6035634743875278,
+      "grad_norm": 0.18473369451560104,
+      "learning_rate": 7.3844884488448855e-06,
+      "loss": 0.266,
+      "step": 1169
+    },
+    {
+      "epoch": 2.6057906458797326,
+      "grad_norm": 0.1938711489246747,
+      "learning_rate": 7.343234323432344e-06,
+      "loss": 0.2637,
+      "step": 1170
+    },
+    {
+      "epoch": 2.6080178173719375,
+      "grad_norm": 0.19020930250626525,
+      "learning_rate": 7.301980198019802e-06,
+      "loss": 0.276,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6102449888641424,
+      "grad_norm": 0.1902014920990745,
+      "learning_rate": 7.260726072607261e-06,
+      "loss": 0.2931,
+      "step": 1172
+    },
+    {
+      "epoch": 2.612472160356347,
+      "grad_norm": 0.20235613156985066,
+      "learning_rate": 7.219471947194719e-06,
+      "loss": 0.2788,
+      "step": 1173
+    },
+    {
+      "epoch": 2.614699331848552,
+      "grad_norm": 0.1904905450934217,
+      "learning_rate": 7.178217821782178e-06,
+      "loss": 0.2813,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6169265033407574,
+      "grad_norm": 0.1829474505593454,
+      "learning_rate": 7.136963696369638e-06,
+      "loss": 0.2565,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6191536748329622,
+      "grad_norm": 0.18034492440862052,
+      "learning_rate": 7.095709570957096e-06,
+      "loss": 0.262,
+      "step": 1176
+    },
+    {
+      "epoch": 2.621380846325167,
+      "grad_norm": 0.18602311086032206,
+      "learning_rate": 7.054455445544554e-06,
+      "loss": 0.277,
+      "step": 1177
+    },
+    {
+      "epoch": 2.623608017817372,
+      "grad_norm": 0.1862942706304272,
+      "learning_rate": 7.013201320132014e-06,
+      "loss": 0.2583,
+      "step": 1178
+    },
+    {
+      "epoch": 2.625835189309577,
+      "grad_norm": 0.21097629481431587,
+      "learning_rate": 6.971947194719472e-06,
+      "loss": 0.274,
+      "step": 1179
+    },
+    {
+      "epoch": 2.6280623608017817,
+      "grad_norm": 0.19480593973142524,
+      "learning_rate": 6.9306930693069314e-06,
+      "loss": 0.2744,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6302895322939865,
+      "grad_norm": 0.20665384592439903,
+      "learning_rate": 6.88943894389439e-06,
+      "loss": 0.2809,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6325167037861914,
+      "grad_norm": 0.20673361836900345,
+      "learning_rate": 6.848184818481848e-06,
+      "loss": 0.2816,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6347438752783967,
+      "grad_norm": 0.19227285073839473,
+      "learning_rate": 6.806930693069308e-06,
+      "loss": 0.2628,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6369710467706016,
+      "grad_norm": 0.19492100553812766,
+      "learning_rate": 6.765676567656766e-06,
+      "loss": 0.2465,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6391982182628064,
+      "grad_norm": 0.1800432051918627,
+      "learning_rate": 6.724422442244224e-06,
+      "loss": 0.2554,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6414253897550113,
+      "grad_norm": 0.18423128116751586,
+      "learning_rate": 6.6831683168316835e-06,
+      "loss": 0.2768,
+      "step": 1186
+    },
+    {
+      "epoch": 2.643652561247216,
+      "grad_norm": 0.18976438520324423,
+      "learning_rate": 6.641914191419142e-06,
+      "loss": 0.2685,
+      "step": 1187
+    },
+    {
+      "epoch": 2.645879732739421,
+      "grad_norm": 0.1782736610121485,
+      "learning_rate": 6.6006600660066e-06,
+      "loss": 0.2491,
+      "step": 1188
+    },
+    {
+      "epoch": 2.648106904231626,
+      "grad_norm": 0.18786731467851306,
+      "learning_rate": 6.55940594059406e-06,
+      "loss": 0.2634,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6503340757238307,
+      "grad_norm": 0.17288807093480227,
+      "learning_rate": 6.518151815181518e-06,
+      "loss": 0.2666,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6525612472160356,
+      "grad_norm": 0.18393738485159578,
+      "learning_rate": 6.476897689768978e-06,
+      "loss": 0.2551,
+      "step": 1191
+    },
+    {
+      "epoch": 2.6547884187082404,
+      "grad_norm": 0.18030281904212597,
+      "learning_rate": 6.4356435643564364e-06,
+      "loss": 0.2658,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6570155902004453,
+      "grad_norm": 0.1827429575224797,
+      "learning_rate": 6.394389438943895e-06,
+      "loss": 0.2722,
+      "step": 1193
+    },
+    {
+      "epoch": 2.65924276169265,
+      "grad_norm": 0.17988767728703248,
+      "learning_rate": 6.353135313531354e-06,
+      "loss": 0.2649,
+      "step": 1194
+    },
+    {
+      "epoch": 2.661469933184855,
+      "grad_norm": 0.1747455257273141,
+      "learning_rate": 6.311881188118812e-06,
+      "loss": 0.2609,
+      "step": 1195
+    },
+    {
+      "epoch": 2.66369710467706,
+      "grad_norm": 0.1890842571379513,
+      "learning_rate": 6.27062706270627e-06,
+      "loss": 0.2836,
+      "step": 1196
+    },
+    {
+      "epoch": 2.665924276169265,
+      "grad_norm": 0.21595638844541948,
+      "learning_rate": 6.2293729372937294e-06,
+      "loss": 0.2995,
+      "step": 1197
+    },
+    {
+      "epoch": 2.66815144766147,
+      "grad_norm": 0.17746447661036516,
+      "learning_rate": 6.1881188118811885e-06,
+      "loss": 0.2693,
+      "step": 1198
+    },
+    {
+      "epoch": 2.670378619153675,
+      "grad_norm": 0.17735444851193852,
+      "learning_rate": 6.146864686468648e-06,
+      "loss": 0.2723,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6726057906458798,
+      "grad_norm": 0.18981324382870987,
+      "learning_rate": 6.105610561056106e-06,
+      "loss": 0.2752,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6748329621380846,
+      "grad_norm": 0.18729354512349214,
+      "learning_rate": 6.064356435643564e-06,
+      "loss": 0.2645,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6770601336302895,
+      "grad_norm": 0.18309412763714775,
+      "learning_rate": 6.023102310231023e-06,
+      "loss": 0.2702,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6792873051224944,
+      "grad_norm": 0.18164682701271706,
+      "learning_rate": 5.981848184818482e-06,
+      "loss": 0.2635,
+      "step": 1203
+    },
+    {
+      "epoch": 2.681514476614699,
+      "grad_norm": 0.1899876797983057,
+      "learning_rate": 5.940594059405941e-06,
+      "loss": 0.2766,
+      "step": 1204
+    },
+    {
+      "epoch": 2.683741648106904,
+      "grad_norm": 0.1827526700599873,
+      "learning_rate": 5.8993399339934e-06,
+      "loss": 0.2671,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6859688195991094,
+      "grad_norm": 0.19843184360259508,
+      "learning_rate": 5.858085808580859e-06,
+      "loss": 0.2805,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6881959910913142,
+      "grad_norm": 0.17624294896059342,
+      "learning_rate": 5.816831683168317e-06,
+      "loss": 0.2674,
+      "step": 1207
+    },
+    {
+      "epoch": 2.690423162583519,
+      "grad_norm": 0.19016213300099563,
+      "learning_rate": 5.775577557755775e-06,
+      "loss": 0.2963,
+      "step": 1208
+    },
+    {
+      "epoch": 2.692650334075724,
+      "grad_norm": 0.1917312001534066,
+      "learning_rate": 5.7343234323432344e-06,
+      "loss": 0.2818,
+      "step": 1209
+    },
+    {
+      "epoch": 2.694877505567929,
+      "grad_norm": 0.1798024925266007,
+      "learning_rate": 5.6930693069306936e-06,
+      "loss": 0.2589,
+      "step": 1210
+    },
+    {
+      "epoch": 2.6971046770601337,
+      "grad_norm": 0.20136922752025474,
+      "learning_rate": 5.651815181518152e-06,
+      "loss": 0.2811,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6993318485523385,
+      "grad_norm": 0.18163309652399173,
+      "learning_rate": 5.610561056105611e-06,
+      "loss": 0.2856,
+      "step": 1212
+    },
+    {
+      "epoch": 2.7015590200445434,
+      "grad_norm": 0.18234375247944756,
+      "learning_rate": 5.56930693069307e-06,
+      "loss": 0.2602,
+      "step": 1213
+    },
+    {
+      "epoch": 2.7037861915367483,
+      "grad_norm": 0.17889900699377081,
+      "learning_rate": 5.528052805280528e-06,
+      "loss": 0.2639,
+      "step": 1214
+    },
+    {
+      "epoch": 2.706013363028953,
+      "grad_norm": 0.18490262644707528,
+      "learning_rate": 5.4867986798679865e-06,
+      "loss": 0.2718,
+      "step": 1215
+    },
+    {
+      "epoch": 2.708240534521158,
+      "grad_norm": 0.1954391315622027,
+      "learning_rate": 5.445544554455446e-06,
+      "loss": 0.2854,
+      "step": 1216
+    },
+    {
+      "epoch": 2.710467706013363,
+      "grad_norm": 0.19260250643946217,
+      "learning_rate": 5.404290429042905e-06,
+      "loss": 0.2716,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7126948775055677,
+      "grad_norm": 0.18831692036934417,
+      "learning_rate": 5.363036303630363e-06,
+      "loss": 0.2552,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7149220489977726,
+      "grad_norm": 0.17430274431770415,
+      "learning_rate": 5.321782178217822e-06,
+      "loss": 0.2561,
+      "step": 1219
+    },
+    {
+      "epoch": 2.717149220489978,
+      "grad_norm": 0.1742355682043371,
+      "learning_rate": 5.280528052805281e-06,
+      "loss": 0.2587,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7193763919821827,
+      "grad_norm": 0.17609385974503883,
+      "learning_rate": 5.2392739273927395e-06,
+      "loss": 0.2632,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7216035634743876,
+      "grad_norm": 0.18316692056914868,
+      "learning_rate": 5.198019801980198e-06,
+      "loss": 0.2689,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7238307349665924,
+      "grad_norm": 0.18141791204316718,
+      "learning_rate": 5.156765676567657e-06,
+      "loss": 0.2699,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7260579064587973,
+      "grad_norm": 0.18466460423749684,
+      "learning_rate": 5.115511551155116e-06,
+      "loss": 0.271,
+      "step": 1224
+    },
+    {
+      "epoch": 2.728285077951002,
+      "grad_norm": 0.17871322125662872,
+      "learning_rate": 5.074257425742575e-06,
+      "loss": 0.2707,
+      "step": 1225
+    },
+    {
+      "epoch": 2.730512249443207,
+      "grad_norm": 0.18054918919202892,
+      "learning_rate": 5.033003300330033e-06,
+      "loss": 0.2742,
+      "step": 1226
+    },
+    {
+      "epoch": 2.732739420935412,
+      "grad_norm": 0.1909635970563273,
+      "learning_rate": 4.991749174917492e-06,
+      "loss": 0.2834,
+      "step": 1227
+    },
+    {
+      "epoch": 2.734966592427617,
+      "grad_norm": 0.1865446441034419,
+      "learning_rate": 4.950495049504951e-06,
+      "loss": 0.2543,
+      "step": 1228
+    },
+    {
+      "epoch": 2.737193763919822,
+      "grad_norm": 0.18186036719557713,
+      "learning_rate": 4.909240924092409e-06,
+      "loss": 0.2467,
+      "step": 1229
+    },
+    {
+      "epoch": 2.739420935412027,
+      "grad_norm": 0.17895945173012268,
+      "learning_rate": 4.867986798679868e-06,
+      "loss": 0.258,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7416481069042318,
+      "grad_norm": 0.21399121434691468,
+      "learning_rate": 4.826732673267327e-06,
+      "loss": 0.2685,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7438752783964366,
+      "grad_norm": 0.18181551081292457,
+      "learning_rate": 4.785478547854786e-06,
+      "loss": 0.2677,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7461024498886415,
+      "grad_norm": 0.18119449697633516,
+      "learning_rate": 4.7442244224422445e-06,
+      "loss": 0.2631,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7483296213808464,
+      "grad_norm": 0.18462309715029207,
+      "learning_rate": 4.702970297029704e-06,
+      "loss": 0.2687,
+      "step": 1234
+    },
+    {
+      "epoch": 2.750556792873051,
+      "grad_norm": 0.18693994168103942,
+      "learning_rate": 4.661716171617162e-06,
+      "loss": 0.2738,
+      "step": 1235
+    },
+    {
+      "epoch": 2.752783964365256,
+      "grad_norm": 0.1811090852572252,
+      "learning_rate": 4.62046204620462e-06,
+      "loss": 0.266,
+      "step": 1236
+    },
+    {
+      "epoch": 2.755011135857461,
+      "grad_norm": 0.1863451840223255,
+      "learning_rate": 4.579207920792079e-06,
+      "loss": 0.2693,
+      "step": 1237
+    },
+    {
+      "epoch": 2.757238307349666,
+      "grad_norm": 0.18547488130828133,
+      "learning_rate": 4.537953795379538e-06,
+      "loss": 0.286,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7594654788418707,
+      "grad_norm": 0.20698222310531328,
+      "learning_rate": 4.496699669966997e-06,
+      "loss": 0.2702,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7616926503340755,
+      "grad_norm": 0.1758322265902654,
+      "learning_rate": 4.455445544554456e-06,
+      "loss": 0.2541,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7639198218262804,
+      "grad_norm": 0.1844287145188825,
+      "learning_rate": 4.414191419141915e-06,
+      "loss": 0.2718,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7661469933184857,
+      "grad_norm": 0.18272135591893765,
+      "learning_rate": 4.372937293729373e-06,
+      "loss": 0.2703,
+      "step": 1242
+    },
+    {
+      "epoch": 2.7683741648106905,
+      "grad_norm": 0.19633638949830934,
+      "learning_rate": 4.331683168316831e-06,
+      "loss": 0.2827,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7706013363028954,
+      "grad_norm": 0.1972991253891059,
+      "learning_rate": 4.29042904290429e-06,
+      "loss": 0.2833,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7728285077951003,
+      "grad_norm": 0.1848672068698584,
+      "learning_rate": 4.2491749174917495e-06,
+      "loss": 0.2611,
+      "step": 1245
+    },
+    {
+      "epoch": 2.775055679287305,
+      "grad_norm": 0.17608402133914777,
+      "learning_rate": 4.207920792079209e-06,
+      "loss": 0.2516,
+      "step": 1246
+    },
+    {
+      "epoch": 2.77728285077951,
+      "grad_norm": 0.18690090446720733,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.263,
+      "step": 1247
+    },
+    {
+      "epoch": 2.779510022271715,
+      "grad_norm": 0.1976897326681855,
+      "learning_rate": 4.125412541254126e-06,
+      "loss": 0.2741,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7817371937639197,
+      "grad_norm": 0.18113007486279845,
+      "learning_rate": 4.084158415841584e-06,
+      "loss": 0.2709,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7839643652561246,
+      "grad_norm": 0.1738680334739268,
+      "learning_rate": 4.0429042904290425e-06,
+      "loss": 0.2609,
+      "step": 1250
+    },
+    {
+      "epoch": 2.78619153674833,
+      "grad_norm": 0.1895039359894699,
+      "learning_rate": 4.001650165016502e-06,
+      "loss": 0.272,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7884187082405347,
+      "grad_norm": 0.18086962959438846,
+      "learning_rate": 3.960396039603961e-06,
+      "loss": 0.2563,
+      "step": 1252
+    },
+    {
+      "epoch": 2.7906458797327396,
+      "grad_norm": 0.18645414962303633,
+      "learning_rate": 3.91914191419142e-06,
+      "loss": 0.2929,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7928730512249444,
+      "grad_norm": 0.17538676931393385,
+      "learning_rate": 3.877887788778878e-06,
+      "loss": 0.261,
+      "step": 1254
+    },
+    {
+      "epoch": 2.7951002227171493,
+      "grad_norm": 0.17785412889632757,
+      "learning_rate": 3.836633663366337e-06,
+      "loss": 0.2689,
+      "step": 1255
+    },
+    {
+      "epoch": 2.797327394209354,
+      "grad_norm": 0.18324305459455334,
+      "learning_rate": 3.795379537953796e-06,
+      "loss": 0.2697,
+      "step": 1256
+    },
+    {
+      "epoch": 2.799554565701559,
+      "grad_norm": 0.18648774005120053,
+      "learning_rate": 3.7541254125412545e-06,
+      "loss": 0.2715,
+      "step": 1257
+    },
+    {
+      "epoch": 2.801781737193764,
+      "grad_norm": 0.18199394587678758,
+      "learning_rate": 3.7128712871287128e-06,
+      "loss": 0.2644,
+      "step": 1258
+    },
+    {
+      "epoch": 2.8040089086859687,
+      "grad_norm": 0.1778832472052361,
+      "learning_rate": 3.671617161716172e-06,
+      "loss": 0.2775,
+      "step": 1259
+    },
+    {
+      "epoch": 2.8062360801781736,
+      "grad_norm": 0.18580722059990096,
+      "learning_rate": 3.6303630363036306e-06,
+      "loss": 0.2862,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8084632516703785,
+      "grad_norm": 0.18284126010620624,
+      "learning_rate": 3.589108910891089e-06,
+      "loss": 0.2777,
+      "step": 1261
+    },
+    {
+      "epoch": 2.8106904231625833,
+      "grad_norm": 0.17793340543250122,
+      "learning_rate": 3.547854785478548e-06,
+      "loss": 0.2695,
+      "step": 1262
+    },
+    {
+      "epoch": 2.812917594654788,
+      "grad_norm": 0.17026032822717652,
+      "learning_rate": 3.506600660066007e-06,
+      "loss": 0.2659,
+      "step": 1263
+    },
+    {
+      "epoch": 2.815144766146993,
+      "grad_norm": 0.2009194626526499,
+      "learning_rate": 3.4653465346534657e-06,
+      "loss": 0.2656,
+      "step": 1264
+    },
+    {
+      "epoch": 2.8173719376391984,
+      "grad_norm": 0.18505755658284295,
+      "learning_rate": 3.424092409240924e-06,
+      "loss": 0.2689,
+      "step": 1265
+    },
+    {
+      "epoch": 2.819599109131403,
+      "grad_norm": 0.19346965340451247,
+      "learning_rate": 3.382838283828383e-06,
+      "loss": 0.2709,
+      "step": 1266
+    },
+    {
+      "epoch": 2.821826280623608,
+      "grad_norm": 0.18373096351793147,
+      "learning_rate": 3.3415841584158418e-06,
+      "loss": 0.2724,
+      "step": 1267
+    },
+    {
+      "epoch": 2.824053452115813,
+      "grad_norm": 0.17744038770228643,
+      "learning_rate": 3.3003300330033e-06,
+      "loss": 0.26,
+      "step": 1268
+    },
+    {
+      "epoch": 2.826280623608018,
+      "grad_norm": 0.17737412754439033,
+      "learning_rate": 3.259075907590759e-06,
+      "loss": 0.2668,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8285077951002227,
+      "grad_norm": 0.17079177918726388,
+      "learning_rate": 3.2178217821782182e-06,
+      "loss": 0.2598,
+      "step": 1270
+    },
+    {
+      "epoch": 2.8307349665924275,
+      "grad_norm": 0.1797365633284667,
+      "learning_rate": 3.176567656765677e-06,
+      "loss": 0.2507,
+      "step": 1271
+    },
+    {
+      "epoch": 2.8329621380846324,
+      "grad_norm": 0.183335091712768,
+      "learning_rate": 3.135313531353135e-06,
+      "loss": 0.2788,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8351893095768377,
+      "grad_norm": 0.17765873948179847,
+      "learning_rate": 3.0940594059405943e-06,
+      "loss": 0.251,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8374164810690425,
+      "grad_norm": 0.17938524809533818,
+      "learning_rate": 3.052805280528053e-06,
+      "loss": 0.2744,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8396436525612474,
+      "grad_norm": 0.17995840306068717,
+      "learning_rate": 3.0115511551155116e-06,
+      "loss": 0.2625,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8418708240534523,
+      "grad_norm": 0.18844273854538926,
+      "learning_rate": 2.9702970297029703e-06,
+      "loss": 0.2814,
+      "step": 1276
+    },
+    {
+      "epoch": 2.844097995545657,
+      "grad_norm": 0.17983513000984847,
+      "learning_rate": 2.9290429042904294e-06,
+      "loss": 0.2647,
+      "step": 1277
+    },
+    {
+      "epoch": 2.846325167037862,
+      "grad_norm": 0.18854659104915525,
+      "learning_rate": 2.8877887788778877e-06,
+      "loss": 0.2786,
+      "step": 1278
+    },
+    {
+      "epoch": 2.848552338530067,
+      "grad_norm": 0.19009114215108075,
+      "learning_rate": 2.8465346534653468e-06,
+      "loss": 0.2833,
+      "step": 1279
+    },
+    {
+      "epoch": 2.8507795100222717,
+      "grad_norm": 0.17542012697195186,
+      "learning_rate": 2.8052805280528055e-06,
+      "loss": 0.27,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8530066815144766,
+      "grad_norm": 0.1735572771297532,
+      "learning_rate": 2.764026402640264e-06,
+      "loss": 0.2591,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8552338530066814,
+      "grad_norm": 0.17811981123044532,
+      "learning_rate": 2.722772277227723e-06,
+      "loss": 0.2686,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8574610244988863,
+      "grad_norm": 0.1879360008971378,
+      "learning_rate": 2.6815181518151815e-06,
+      "loss": 0.2664,
+      "step": 1283
+    },
+    {
+      "epoch": 2.859688195991091,
+      "grad_norm": 0.17888528677298804,
+      "learning_rate": 2.6402640264026406e-06,
+      "loss": 0.2701,
+      "step": 1284
+    },
+    {
+      "epoch": 2.861915367483296,
+      "grad_norm": 0.17662474501142036,
+      "learning_rate": 2.599009900990099e-06,
+      "loss": 0.2826,
+      "step": 1285
+    },
+    {
+      "epoch": 2.864142538975501,
+      "grad_norm": 0.18906031553549424,
+      "learning_rate": 2.557755775577558e-06,
+      "loss": 0.2866,
+      "step": 1286
+    },
+    {
+      "epoch": 2.866369710467706,
+      "grad_norm": 0.17354639342193298,
+      "learning_rate": 2.5165016501650166e-06,
+      "loss": 0.2502,
+      "step": 1287
+    },
+    {
+      "epoch": 2.868596881959911,
+      "grad_norm": 0.17707561914042097,
+      "learning_rate": 2.4752475247524753e-06,
+      "loss": 0.2705,
+      "step": 1288
+    },
+    {
+      "epoch": 2.870824053452116,
+      "grad_norm": 0.17357571373149694,
+      "learning_rate": 2.433993399339934e-06,
+      "loss": 0.2598,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8730512249443207,
+      "grad_norm": 0.18095560062308227,
+      "learning_rate": 2.392739273927393e-06,
+      "loss": 0.2635,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8752783964365256,
+      "grad_norm": 0.17612400842571985,
+      "learning_rate": 2.351485148514852e-06,
+      "loss": 0.2554,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8775055679287305,
+      "grad_norm": 0.18632630258694613,
+      "learning_rate": 2.31023102310231e-06,
+      "loss": 0.2754,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8797327394209353,
+      "grad_norm": 0.176004846745384,
+      "learning_rate": 2.268976897689769e-06,
+      "loss": 0.2661,
+      "step": 1293
+    },
+    {
+      "epoch": 2.88195991091314,
+      "grad_norm": 0.1844978464314028,
+      "learning_rate": 2.227722772277228e-06,
+      "loss": 0.2703,
+      "step": 1294
+    },
+    {
+      "epoch": 2.884187082405345,
+      "grad_norm": 0.17388026417343408,
+      "learning_rate": 2.1864686468646865e-06,
+      "loss": 0.2594,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8864142538975504,
+      "grad_norm": 0.1746902632671786,
+      "learning_rate": 2.145214521452145e-06,
+      "loss": 0.2604,
+      "step": 1296
+    },
+    {
+      "epoch": 2.888641425389755,
+      "grad_norm": 0.17388558045342958,
+      "learning_rate": 2.1039603960396043e-06,
+      "loss": 0.2735,
+      "step": 1297
+    },
+    {
+      "epoch": 2.89086859688196,
+      "grad_norm": 0.1805564659412802,
+      "learning_rate": 2.062706270627063e-06,
+      "loss": 0.2795,
+      "step": 1298
+    },
+    {
+      "epoch": 2.893095768374165,
+      "grad_norm": 0.18299640212213222,
+      "learning_rate": 2.0214521452145212e-06,
+      "loss": 0.2788,
+      "step": 1299
+    },
+    {
+      "epoch": 2.89532293986637,
+      "grad_norm": 0.1726098063623092,
+      "learning_rate": 1.9801980198019803e-06,
+      "loss": 0.2695,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8975501113585747,
+      "grad_norm": 0.7589122167800628,
+      "learning_rate": 1.938943894389439e-06,
+      "loss": 0.2756,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8997772828507795,
+      "grad_norm": 0.17268798276447622,
+      "learning_rate": 1.897689768976898e-06,
+      "loss": 0.2676,
+      "step": 1302
+    },
+    {
+      "epoch": 2.9020044543429844,
+      "grad_norm": 0.1761435400733331,
+      "learning_rate": 1.8564356435643564e-06,
+      "loss": 0.2631,
+      "step": 1303
+    },
+    {
+      "epoch": 2.9042316258351892,
+      "grad_norm": 0.19236534492394342,
+      "learning_rate": 1.8151815181518153e-06,
+      "loss": 0.2894,
+      "step": 1304
+    },
+    {
+      "epoch": 2.906458797327394,
+      "grad_norm": 0.1837113832497075,
+      "learning_rate": 1.773927392739274e-06,
+      "loss": 0.2577,
+      "step": 1305
+    },
+    {
+      "epoch": 2.908685968819599,
+      "grad_norm": 0.18799332234673125,
+      "learning_rate": 1.7326732673267329e-06,
+      "loss": 0.2895,
+      "step": 1306
+    },
+    {
+      "epoch": 2.910913140311804,
+      "grad_norm": 0.16575965823267147,
+      "learning_rate": 1.6914191419141915e-06,
+      "loss": 0.2586,
+      "step": 1307
+    },
+    {
+      "epoch": 2.9131403118040087,
+      "grad_norm": 0.1821331264805557,
+      "learning_rate": 1.65016501650165e-06,
+      "loss": 0.2901,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9153674832962135,
+      "grad_norm": 0.20278623871064896,
+      "learning_rate": 1.6089108910891091e-06,
+      "loss": 0.2547,
+      "step": 1309
+    },
+    {
+      "epoch": 2.917594654788419,
+      "grad_norm": 0.18693643097264878,
+      "learning_rate": 1.5676567656765676e-06,
+      "loss": 0.2665,
+      "step": 1310
+    },
+    {
+      "epoch": 2.9198218262806237,
+      "grad_norm": 0.18251206885926352,
+      "learning_rate": 1.5264026402640265e-06,
+      "loss": 0.305,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9220489977728286,
+      "grad_norm": 0.1795333615670338,
+      "learning_rate": 1.4851485148514852e-06,
+      "loss": 0.2746,
+      "step": 1312
+    },
+    {
+      "epoch": 2.9242761692650334,
+      "grad_norm": 0.17714431772046693,
+      "learning_rate": 1.4438943894389438e-06,
+      "loss": 0.2627,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9265033407572383,
+      "grad_norm": 0.17660747696135468,
+      "learning_rate": 1.4026402640264027e-06,
+      "loss": 0.272,
+      "step": 1314
+    },
+    {
+      "epoch": 2.928730512249443,
+      "grad_norm": 0.18249194400005841,
+      "learning_rate": 1.3613861386138614e-06,
+      "loss": 0.2667,
+      "step": 1315
+    },
+    {
+      "epoch": 2.930957683741648,
+      "grad_norm": 0.18093238717836005,
+      "learning_rate": 1.3201320132013203e-06,
+      "loss": 0.2756,
+      "step": 1316
+    },
+    {
+      "epoch": 2.933184855233853,
+      "grad_norm": 0.1857417372113312,
+      "learning_rate": 1.278877887788779e-06,
+      "loss": 0.2765,
+      "step": 1317
+    },
+    {
+      "epoch": 2.935412026726058,
+      "grad_norm": 0.18733495569539454,
+      "learning_rate": 1.2376237623762377e-06,
+      "loss": 0.2798,
+      "step": 1318
+    },
+    {
+      "epoch": 2.937639198218263,
+      "grad_norm": 0.17193783973837945,
+      "learning_rate": 1.1963696369636966e-06,
+      "loss": 0.2521,
+      "step": 1319
+    },
+    {
+      "epoch": 2.939866369710468,
+      "grad_norm": 0.16811094774035254,
+      "learning_rate": 1.155115511551155e-06,
+      "loss": 0.2579,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9420935412026727,
+      "grad_norm": 0.1793421904458536,
+      "learning_rate": 1.113861386138614e-06,
+      "loss": 0.2698,
+      "step": 1321
+    },
+    {
+      "epoch": 2.9443207126948776,
+      "grad_norm": 0.17505515975567199,
+      "learning_rate": 1.0726072607260726e-06,
+      "loss": 0.2685,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9465478841870825,
+      "grad_norm": 0.18282625596514968,
+      "learning_rate": 1.0313531353135315e-06,
+      "loss": 0.2736,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9487750556792873,
+      "grad_norm": 0.16894345591206814,
+      "learning_rate": 9.900990099009902e-07,
+      "loss": 0.2564,
+      "step": 1324
+    },
+    {
+      "epoch": 2.951002227171492,
+      "grad_norm": 0.17025646518611448,
+      "learning_rate": 9.48844884488449e-07,
+      "loss": 0.2599,
+      "step": 1325
+    },
+    {
+      "epoch": 2.953229398663697,
+      "grad_norm": 0.17513515627006346,
+      "learning_rate": 9.075907590759076e-07,
+      "loss": 0.2791,
+      "step": 1326
+    },
+    {
+      "epoch": 2.955456570155902,
+      "grad_norm": 0.1828597861925402,
+      "learning_rate": 8.663366336633664e-07,
+      "loss": 0.2889,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9576837416481068,
+      "grad_norm": 0.17806935040212402,
+      "learning_rate": 8.25082508250825e-07,
+      "loss": 0.2766,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9599109131403116,
+      "grad_norm": 0.17322189701290536,
+      "learning_rate": 7.838283828382838e-07,
+      "loss": 0.2774,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9621380846325165,
+      "grad_norm": 0.17996043400629883,
+      "learning_rate": 7.425742574257426e-07,
+      "loss": 0.2734,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9643652561247213,
+      "grad_norm": 0.17947187484562413,
+      "learning_rate": 7.013201320132014e-07,
+      "loss": 0.2713,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9665924276169267,
+      "grad_norm": 0.19967183239652675,
+      "learning_rate": 6.600660066006602e-07,
+      "loss": 0.2645,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9688195991091315,
+      "grad_norm": 0.16870682519840827,
+      "learning_rate": 6.188118811881188e-07,
+      "loss": 0.2653,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9710467706013364,
+      "grad_norm": 0.17344643257978654,
+      "learning_rate": 5.775577557755775e-07,
+      "loss": 0.2486,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9732739420935412,
+      "grad_norm": 0.1818486260078175,
+      "learning_rate": 5.363036303630363e-07,
+      "loss": 0.2704,
+      "step": 1335
+    },
+    {
+      "epoch": 2.975501113585746,
+      "grad_norm": 0.18264295704074213,
+      "learning_rate": 4.950495049504951e-07,
+      "loss": 0.2847,
+      "step": 1336
+    },
+    {
+      "epoch": 2.977728285077951,
+      "grad_norm": 0.18007690267794807,
+      "learning_rate": 4.537953795379538e-07,
+      "loss": 0.2623,
+      "step": 1337
+    },
+    {
+      "epoch": 2.979955456570156,
+      "grad_norm": 0.1732717832641267,
+      "learning_rate": 4.125412541254125e-07,
+      "loss": 0.2572,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9821826280623607,
+      "grad_norm": 0.16832712239665867,
+      "learning_rate": 3.712871287128713e-07,
+      "loss": 0.2553,
+      "step": 1339
+    },
+    {
+      "epoch": 2.984409799554566,
+      "grad_norm": 0.1944252415604603,
+      "learning_rate": 3.300330033003301e-07,
+      "loss": 0.2639,
+      "step": 1340
+    },
+    {
+      "epoch": 2.986636971046771,
+      "grad_norm": 0.20639972423223196,
+      "learning_rate": 2.8877887788778876e-07,
+      "loss": 0.2806,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9888641425389757,
+      "grad_norm": 0.1910915850331487,
+      "learning_rate": 2.4752475247524754e-07,
+      "loss": 0.2728,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9910913140311806,
+      "grad_norm": 0.1756097477275674,
+      "learning_rate": 2.0627062706270625e-07,
+      "loss": 0.272,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9933184855233854,
+      "grad_norm": 0.17215083407744072,
+      "learning_rate": 1.6501650165016504e-07,
+      "loss": 0.2705,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9955456570155903,
+      "grad_norm": 0.18331819283347497,
+      "learning_rate": 1.2376237623762377e-07,
+      "loss": 0.2926,
+      "step": 1345
+    },
+    {
+      "epoch": 2.997772828507795,
+      "grad_norm": 0.17390758148350594,
+      "learning_rate": 8.250825082508252e-08,
+      "loss": 0.2698,
+      "step": 1346
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.17310367320777945,
+      "learning_rate": 4.125412541254126e-08,
+      "loss": 0.2685,
+      "step": 1347
+    },
+    {
+      "epoch": 3.0,
+      "step": 1347,
+      "total_flos": 1.4980518380283363e+19,
+      "train_loss": 0.4525371772342907,
+      "train_runtime": 41466.0073,
+      "train_samples_per_second": 0.52,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1347,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4980518380283363e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}