{
  "best_global_step": 12395,
  "best_metric": 0.9532667876588021,
  "best_model_checkpoint": "./resnet50-fashion/checkpoint-12395",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 12395,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02016942315449778,
      "grad_norm": 10.520899772644043,
      "learning_rate": 4.9802339653085925e-05,
      "loss": 3.5885751342773435,
      "step": 50
    },
    {
      "epoch": 0.04033884630899556,
      "grad_norm": 11.112278938293457,
      "learning_rate": 4.9600645421540946e-05,
      "loss": 3.066470947265625,
      "step": 100
    },
    {
      "epoch": 0.060508269463493344,
      "grad_norm": 6.844457149505615,
      "learning_rate": 4.939895118999597e-05,
      "loss": 2.5258544921875,
      "step": 150
    },
    {
      "epoch": 0.08067769261799113,
      "grad_norm": 13.254263877868652,
      "learning_rate": 4.919725695845099e-05,
      "loss": 2.1543016052246093,
      "step": 200
    },
    {
      "epoch": 0.10084711577248891,
      "grad_norm": 6.580667972564697,
      "learning_rate": 4.8995562726906015e-05,
      "loss": 1.9740011596679687,
      "step": 250
    },
    {
      "epoch": 0.12101653892698669,
      "grad_norm": 7.654535293579102,
      "learning_rate": 4.8793868495361035e-05,
      "loss": 1.7021075439453126,
      "step": 300
    },
    {
      "epoch": 0.14118596208148446,
      "grad_norm": 6.575769424438477,
      "learning_rate": 4.8592174263816056e-05,
      "loss": 1.6266166687011718,
      "step": 350
    },
    {
      "epoch": 0.16135538523598225,
      "grad_norm": 4.312820911407471,
      "learning_rate": 4.839048003227108e-05,
      "loss": 1.4934498596191406,
      "step": 400
    },
    {
      "epoch": 0.18152480839048002,
      "grad_norm": 4.195486068725586,
      "learning_rate": 4.81887858007261e-05,
      "loss": 1.33470703125,
      "step": 450
    },
    {
      "epoch": 0.20169423154497781,
      "grad_norm": 7.272072792053223,
      "learning_rate": 4.7987091569181125e-05,
      "loss": 1.2691184234619142,
      "step": 500
    },
    {
      "epoch": 0.22186365469947558,
      "grad_norm": 5.7398200035095215,
      "learning_rate": 4.7785397337636145e-05,
      "loss": 1.136511993408203,
      "step": 550
    },
    {
      "epoch": 0.24203307785397338,
      "grad_norm": 6.210187911987305,
      "learning_rate": 4.7583703106091166e-05,
      "loss": 1.0565355682373048,
      "step": 600
    },
    {
      "epoch": 0.26220250100847114,
      "grad_norm": 7.919003963470459,
      "learning_rate": 4.7382008874546193e-05,
      "loss": 0.9344426727294922,
      "step": 650
    },
    {
      "epoch": 0.2823719241629689,
      "grad_norm": 3.5390396118164062,
      "learning_rate": 4.7180314643001214e-05,
      "loss": 0.8554486846923828,
      "step": 700
    },
    {
      "epoch": 0.30254134731746674,
      "grad_norm": 7.0212273597717285,
      "learning_rate": 4.6978620411456235e-05,
      "loss": 0.7601541137695312,
      "step": 750
    },
    {
      "epoch": 0.3227107704719645,
      "grad_norm": 4.230292320251465,
      "learning_rate": 4.6776926179911255e-05,
      "loss": 0.6942082977294922,
      "step": 800
    },
    {
      "epoch": 0.34288019362646227,
      "grad_norm": 7.144539833068848,
      "learning_rate": 4.6575231948366276e-05,
      "loss": 0.6801434326171875,
      "step": 850
    },
    {
      "epoch": 0.36304961678096004,
      "grad_norm": 2.7382545471191406,
      "learning_rate": 4.6373537716821303e-05,
      "loss": 0.7084119415283203,
      "step": 900
    },
    {
      "epoch": 0.38321903993545786,
      "grad_norm": 6.787492752075195,
      "learning_rate": 4.6171843485276324e-05,
      "loss": 0.6313615417480469,
      "step": 950
    },
    {
      "epoch": 0.40338846308995563,
      "grad_norm": 6.617012023925781,
      "learning_rate": 4.5970149253731345e-05,
      "loss": 0.6253053283691407,
      "step": 1000
    },
    {
      "epoch": 0.4235578862444534,
      "grad_norm": 2.82509708404541,
      "learning_rate": 4.5768455022186365e-05,
      "loss": 0.5697365188598633,
      "step": 1050
    },
    {
      "epoch": 0.44372730939895116,
      "grad_norm": 6.462093353271484,
      "learning_rate": 4.5566760790641386e-05,
      "loss": 0.5618856811523437,
      "step": 1100
    },
    {
      "epoch": 0.463896732553449,
      "grad_norm": 8.217955589294434,
      "learning_rate": 4.5365066559096413e-05,
      "loss": 0.5241514587402344,
      "step": 1150
    },
    {
      "epoch": 0.48406615570794675,
      "grad_norm": 5.901000499725342,
      "learning_rate": 4.5163372327551434e-05,
      "loss": 0.5597282028198243,
      "step": 1200
    },
    {
      "epoch": 0.5042355788624445,
      "grad_norm": 5.706325054168701,
      "learning_rate": 4.4961678096006455e-05,
      "loss": 0.500801773071289,
      "step": 1250
    },
    {
      "epoch": 0.5244050020169423,
      "grad_norm": 3.6276135444641113,
      "learning_rate": 4.475998386446148e-05,
      "loss": 0.4998150253295898,
      "step": 1300
    },
    {
      "epoch": 0.5445744251714401,
      "grad_norm": 1.5081746578216553,
      "learning_rate": 4.45582896329165e-05,
      "loss": 0.4404385757446289,
      "step": 1350
    },
    {
      "epoch": 0.5647438483259378,
      "grad_norm": 2.9040212631225586,
      "learning_rate": 4.4356595401371523e-05,
      "loss": 0.42787273406982423,
      "step": 1400
    },
    {
      "epoch": 0.5849132714804357,
      "grad_norm": 6.973745822906494,
      "learning_rate": 4.4154901169826544e-05,
      "loss": 0.4308282470703125,
      "step": 1450
    },
    {
      "epoch": 0.6050826946349335,
      "grad_norm": 5.634387969970703,
      "learning_rate": 4.3953206938281565e-05,
      "loss": 0.3713912582397461,
      "step": 1500
    },
    {
      "epoch": 0.6252521177894312,
      "grad_norm": 2.7692153453826904,
      "learning_rate": 4.375151270673659e-05,
      "loss": 0.38077377319335937,
      "step": 1550
    },
    {
      "epoch": 0.645421540943929,
      "grad_norm": 3.9123497009277344,
      "learning_rate": 4.354981847519161e-05,
      "loss": 0.35636608123779295,
      "step": 1600
    },
    {
      "epoch": 0.6655909640984268,
      "grad_norm": 1.6538969278335571,
      "learning_rate": 4.3348124243646633e-05,
      "loss": 0.3324755096435547,
      "step": 1650
    },
    {
      "epoch": 0.6857603872529245,
      "grad_norm": 2.7436976432800293,
      "learning_rate": 4.3146430012101654e-05,
      "loss": 0.3718099594116211,
      "step": 1700
    },
    {
      "epoch": 0.7059298104074223,
      "grad_norm": 2.514030933380127,
      "learning_rate": 4.2944735780556675e-05,
      "loss": 0.365201416015625,
      "step": 1750
    },
    {
      "epoch": 0.7260992335619201,
      "grad_norm": 3.3714559078216553,
      "learning_rate": 4.2743041549011695e-05,
      "loss": 0.3048884391784668,
      "step": 1800
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 3.61609148979187,
      "learning_rate": 4.254134731746672e-05,
      "loss": 0.4202354049682617,
      "step": 1850
    },
    {
      "epoch": 0.7664380798709157,
      "grad_norm": 5.999142169952393,
      "learning_rate": 4.2339653085921743e-05,
      "loss": 0.4184662628173828,
      "step": 1900
    },
    {
      "epoch": 0.7866075030254135,
      "grad_norm": 7.747167587280273,
      "learning_rate": 4.213795885437677e-05,
      "loss": 0.2971736145019531,
      "step": 1950
    },
    {
      "epoch": 0.8067769261799113,
      "grad_norm": 3.178110122680664,
      "learning_rate": 4.193626462283179e-05,
      "loss": 0.3129091262817383,
      "step": 2000
    },
    {
      "epoch": 0.826946349334409,
      "grad_norm": 7.176916599273682,
      "learning_rate": 4.173457039128681e-05,
      "loss": 0.3405015182495117,
      "step": 2050
    },
    {
      "epoch": 0.8471157724889068,
      "grad_norm": 2.1656532287597656,
      "learning_rate": 4.153287615974183e-05,
      "loss": 0.32712291717529296,
      "step": 2100
    },
    {
      "epoch": 0.8672851956434046,
      "grad_norm": 7.728622913360596,
      "learning_rate": 4.1331181928196853e-05,
      "loss": 0.3142762756347656,
      "step": 2150
    },
    {
      "epoch": 0.8874546187979023,
      "grad_norm": 3.9452037811279297,
      "learning_rate": 4.1129487696651874e-05,
      "loss": 0.3304166793823242,
      "step": 2200
    },
    {
      "epoch": 0.9076240419524002,
      "grad_norm": 4.345726013183594,
      "learning_rate": 4.09277934651069e-05,
      "loss": 0.29476640701293944,
      "step": 2250
    },
    {
      "epoch": 0.927793465106898,
      "grad_norm": 9.373312950134277,
      "learning_rate": 4.072609923356192e-05,
      "loss": 0.29666582107543943,
      "step": 2300
    },
    {
      "epoch": 0.9479628882613957,
      "grad_norm": 15.719057083129883,
      "learning_rate": 4.052440500201695e-05,
      "loss": 0.31481658935546875,
      "step": 2350
    },
    {
      "epoch": 0.9681323114158935,
      "grad_norm": 11.424883842468262,
      "learning_rate": 4.0322710770471963e-05,
      "loss": 0.40077999114990237,
      "step": 2400
    },
    {
      "epoch": 0.9883017345703913,
      "grad_norm": 3.748368740081787,
      "learning_rate": 4.0121016538926984e-05,
      "loss": 0.3299874496459961,
      "step": 2450
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9299001814882033,
      "eval_loss": 0.27112650871276855,
      "eval_runtime": 20.3548,
      "eval_samples_per_second": 216.558,
      "eval_steps_per_second": 13.559,
      "step": 2479
    },
    {
      "epoch": 1.008471157724889,
      "grad_norm": 6.341300964355469,
      "learning_rate": 3.991932230738201e-05,
      "loss": 0.3089567756652832,
      "step": 2500
    },
    {
      "epoch": 1.0286405808793868,
      "grad_norm": 2.0980629920959473,
      "learning_rate": 3.971762807583703e-05,
      "loss": 0.3381888961791992,
      "step": 2550
    },
    {
      "epoch": 1.0488100040338846,
      "grad_norm": 0.7737718820571899,
      "learning_rate": 3.951593384429206e-05,
      "loss": 0.2835183525085449,
      "step": 2600
    },
    {
      "epoch": 1.0689794271883823,
      "grad_norm": 6.585007190704346,
      "learning_rate": 3.931423961274708e-05,
      "loss": 0.2929516220092773,
      "step": 2650
    },
    {
      "epoch": 1.0891488503428801,
      "grad_norm": 2.3810129165649414,
      "learning_rate": 3.91125453812021e-05,
      "loss": 0.2525530433654785,
      "step": 2700
    },
    {
      "epoch": 1.1093182734973779,
      "grad_norm": 3.5129992961883545,
      "learning_rate": 3.891085114965712e-05,
      "loss": 0.27019474029541013,
      "step": 2750
    },
    {
      "epoch": 1.1294876966518759,
      "grad_norm": 5.742709159851074,
      "learning_rate": 3.870915691811214e-05,
      "loss": 0.2933888626098633,
      "step": 2800
    },
    {
      "epoch": 1.1496571198063736,
      "grad_norm": 3.1736836433410645,
      "learning_rate": 3.850746268656716e-05,
      "loss": 0.2935880470275879,
      "step": 2850
    },
    {
      "epoch": 1.1698265429608714,
      "grad_norm": 6.119251251220703,
      "learning_rate": 3.830576845502219e-05,
      "loss": 0.3064602470397949,
      "step": 2900
    },
    {
      "epoch": 1.1899959661153692,
      "grad_norm": 3.5085670948028564,
      "learning_rate": 3.810407422347721e-05,
      "loss": 0.25757852554321287,
      "step": 2950
    },
    {
      "epoch": 1.210165389269867,
      "grad_norm": 3.2313594818115234,
      "learning_rate": 3.790237999193224e-05,
      "loss": 0.2911284065246582,
      "step": 3000
    },
    {
      "epoch": 1.2303348124243647,
      "grad_norm": 4.116666316986084,
      "learning_rate": 3.770068576038725e-05,
      "loss": 0.3029876136779785,
      "step": 3050
    },
    {
      "epoch": 1.2505042355788625,
      "grad_norm": 0.5835158824920654,
      "learning_rate": 3.749899152884227e-05,
      "loss": 0.22566375732421876,
      "step": 3100
    },
    {
      "epoch": 1.2706736587333602,
      "grad_norm": 4.364255428314209,
      "learning_rate": 3.72972972972973e-05,
      "loss": 0.20720888137817384,
      "step": 3150
    },
    {
      "epoch": 1.290843081887858,
      "grad_norm": 3.6221392154693604,
      "learning_rate": 3.709560306575232e-05,
      "loss": 0.22576507568359375,
      "step": 3200
    },
    {
      "epoch": 1.3110125050423558,
      "grad_norm": 10.07580852508545,
      "learning_rate": 3.689390883420734e-05,
      "loss": 0.2189176368713379,
      "step": 3250
    },
    {
      "epoch": 1.3311819281968535,
      "grad_norm": 6.97329568862915,
      "learning_rate": 3.669221460266237e-05,
      "loss": 0.338087158203125,
      "step": 3300
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 1.1882565021514893,
      "learning_rate": 3.649052037111739e-05,
      "loss": 0.21201395034790038,
      "step": 3350
    },
    {
      "epoch": 1.371520774505849,
      "grad_norm": 2.1187331676483154,
      "learning_rate": 3.628882613957241e-05,
      "loss": 0.22095136642456054,
      "step": 3400
    },
    {
      "epoch": 1.3916901976603469,
      "grad_norm": 4.908150672912598,
      "learning_rate": 3.608713190802743e-05,
      "loss": 0.24210647583007813,
      "step": 3450
    },
    {
      "epoch": 1.4118596208148446,
      "grad_norm": 2.4119579792022705,
      "learning_rate": 3.588543767648245e-05,
      "loss": 0.2516419792175293,
      "step": 3500
    },
    {
      "epoch": 1.4320290439693424,
      "grad_norm": 5.924858570098877,
      "learning_rate": 3.568374344493748e-05,
      "loss": 0.2286382484436035,
      "step": 3550
    },
    {
      "epoch": 1.4521984671238402,
      "grad_norm": 6.0561041831970215,
      "learning_rate": 3.54820492133925e-05,
      "loss": 0.272211971282959,
      "step": 3600
    },
    {
      "epoch": 1.472367890278338,
      "grad_norm": 11.075820922851562,
      "learning_rate": 3.528035498184752e-05,
      "loss": 0.18730430603027343,
      "step": 3650
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 6.133443355560303,
      "learning_rate": 3.507866075030255e-05,
      "loss": 0.2744808006286621,
      "step": 3700
    },
    {
      "epoch": 1.5127067365873335,
      "grad_norm": 2.7195820808410645,
      "learning_rate": 3.487696651875756e-05,
      "loss": 0.2309401512145996,
      "step": 3750
    },
    {
      "epoch": 1.5328761597418312,
      "grad_norm": 5.68732213973999,
      "learning_rate": 3.467527228721259e-05,
      "loss": 0.22877124786376954,
      "step": 3800
    },
    {
      "epoch": 1.5530455828963292,
      "grad_norm": 5.202581882476807,
      "learning_rate": 3.447357805566761e-05,
      "loss": 0.27316980361938475,
      "step": 3850
    },
    {
      "epoch": 1.573215006050827,
      "grad_norm": 0.48496291041374207,
      "learning_rate": 3.427188382412263e-05,
      "loss": 0.20376949310302733,
      "step": 3900
    },
    {
      "epoch": 1.5933844292053247,
      "grad_norm": 3.1519060134887695,
      "learning_rate": 3.407018959257766e-05,
      "loss": 0.20258005142211913,
      "step": 3950
    },
    {
      "epoch": 1.6135538523598225,
      "grad_norm": 0.649708092212677,
      "learning_rate": 3.386849536103268e-05,
      "loss": 0.2322039031982422,
      "step": 4000
    },
    {
      "epoch": 1.6337232755143203,
      "grad_norm": 2.9072558879852295,
      "learning_rate": 3.36668011294877e-05,
      "loss": 0.20238967895507812,
      "step": 4050
    },
    {
      "epoch": 1.653892698668818,
      "grad_norm": 1.227184772491455,
      "learning_rate": 3.346510689794272e-05,
      "loss": 0.19608688354492188,
      "step": 4100
    },
    {
      "epoch": 1.6740621218233158,
      "grad_norm": 5.7164483070373535,
      "learning_rate": 3.326341266639774e-05,
      "loss": 0.2062531852722168,
      "step": 4150
    },
    {
      "epoch": 1.6942315449778136,
      "grad_norm": 4.255147457122803,
      "learning_rate": 3.306171843485277e-05,
      "loss": 0.24369115829467775,
      "step": 4200
    },
    {
      "epoch": 1.7144009681323116,
      "grad_norm": 3.4695684909820557,
      "learning_rate": 3.286002420330779e-05,
      "loss": 0.19526601791381837,
      "step": 4250
    },
    {
      "epoch": 1.7345703912868093,
      "grad_norm": 3.0551869869232178,
      "learning_rate": 3.265832997176281e-05,
      "loss": 0.2388477325439453,
      "step": 4300
    },
    {
      "epoch": 1.754739814441307,
      "grad_norm": 5.207853317260742,
      "learning_rate": 3.2456635740217836e-05,
      "loss": 0.22748014450073242,
      "step": 4350
    },
    {
      "epoch": 1.7749092375958049,
      "grad_norm": 6.68976354598999,
      "learning_rate": 3.225494150867285e-05,
      "loss": 0.2202134132385254,
      "step": 4400
    },
    {
      "epoch": 1.7950786607503026,
      "grad_norm": 1.4704676866531372,
      "learning_rate": 3.205324727712787e-05,
      "loss": 0.19456554412841798,
      "step": 4450
    },
    {
      "epoch": 1.8152480839048004,
      "grad_norm": 7.315227508544922,
      "learning_rate": 3.18515530455829e-05,
      "loss": 0.3203375244140625,
      "step": 4500
    },
    {
      "epoch": 1.8354175070592982,
      "grad_norm": 1.7864209413528442,
      "learning_rate": 3.164985881403792e-05,
      "loss": 0.18399269104003907,
      "step": 4550
    },
    {
      "epoch": 1.855586930213796,
      "grad_norm": 5.953585624694824,
      "learning_rate": 3.1448164582492946e-05,
      "loss": 0.25314868927001954,
      "step": 4600
    },
    {
      "epoch": 1.8757563533682937,
      "grad_norm": 4.386700630187988,
      "learning_rate": 3.124647035094797e-05,
      "loss": 0.19606996536254884,
      "step": 4650
    },
    {
      "epoch": 1.8959257765227915,
      "grad_norm": 6.6788330078125,
      "learning_rate": 3.104477611940299e-05,
      "loss": 0.19745792388916017,
      "step": 4700
    },
    {
      "epoch": 1.9160951996772893,
      "grad_norm": 5.657231330871582,
      "learning_rate": 3.084308188785801e-05,
      "loss": 0.17997068405151367,
      "step": 4750
    },
    {
      "epoch": 1.936264622831787,
      "grad_norm": 1.3474644422531128,
      "learning_rate": 3.064138765631303e-05,
      "loss": 0.19095916748046876,
      "step": 4800
    },
    {
      "epoch": 1.9564340459862848,
      "grad_norm": 4.156283378601074,
      "learning_rate": 3.043969342476805e-05,
      "loss": 0.21400732040405274,
      "step": 4850
    },
    {
      "epoch": 1.9766034691407826,
      "grad_norm": 0.9628849029541016,
      "learning_rate": 3.0237999193223077e-05,
      "loss": 0.23152576446533202,
      "step": 4900
    },
    {
      "epoch": 1.9967728922952803,
      "grad_norm": 8.431641578674316,
      "learning_rate": 3.0036304961678098e-05,
      "loss": 0.19971717834472658,
      "step": 4950
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9441923774954628,
      "eval_loss": 0.2141035795211792,
      "eval_runtime": 20.5277,
      "eval_samples_per_second": 214.734,
      "eval_steps_per_second": 13.445,
      "step": 4958
    },
    {
      "epoch": 2.016942315449778,
      "grad_norm": 7.967335224151611,
      "learning_rate": 2.9834610730133122e-05,
      "loss": 0.20258569717407227,
      "step": 5000
    },
    {
      "epoch": 2.037111738604276,
      "grad_norm": 1.8908740282058716,
      "learning_rate": 2.9632916498588142e-05,
      "loss": 0.2126533317565918,
      "step": 5050
    },
    {
      "epoch": 2.0572811617587736,
      "grad_norm": 0.6554378867149353,
      "learning_rate": 2.9431222267043163e-05,
      "loss": 0.21582073211669922,
      "step": 5100
    },
    {
      "epoch": 2.0774505849132714,
      "grad_norm": 2.4483304023742676,
      "learning_rate": 2.9229528035498187e-05,
      "loss": 0.21905384063720704,
      "step": 5150
    },
    {
      "epoch": 2.097620008067769,
      "grad_norm": 3.9244415760040283,
      "learning_rate": 2.9027833803953208e-05,
      "loss": 0.16603147506713867,
      "step": 5200
    },
    {
      "epoch": 2.117789431222267,
      "grad_norm": 0.17323902249336243,
      "learning_rate": 2.882613957240823e-05,
      "loss": 0.1652754020690918,
      "step": 5250
    },
    {
      "epoch": 2.1379588543767647,
      "grad_norm": 4.670708179473877,
      "learning_rate": 2.8624445340863252e-05,
      "loss": 0.17886968612670898,
      "step": 5300
    },
    {
      "epoch": 2.1581282775312625,
      "grad_norm": 1.9104596376419067,
      "learning_rate": 2.8422751109318273e-05,
      "loss": 0.17843175888061524,
      "step": 5350
    },
    {
      "epoch": 2.1782977006857602,
      "grad_norm": 3.732868194580078,
      "learning_rate": 2.82210568777733e-05,
      "loss": 0.163537654876709,
      "step": 5400
    },
    {
      "epoch": 2.198467123840258,
      "grad_norm": 3.159740924835205,
      "learning_rate": 2.801936264622832e-05,
      "loss": 0.18888442993164062,
      "step": 5450
    },
    {
      "epoch": 2.2186365469947558,
      "grad_norm": 3.3042259216308594,
      "learning_rate": 2.781766841468334e-05,
      "loss": 0.1473686122894287,
      "step": 5500
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 4.918742656707764,
      "learning_rate": 2.7615974183138366e-05,
      "loss": 0.26290460586547854,
      "step": 5550
    },
    {
      "epoch": 2.2589753933037517,
      "grad_norm": 0.3327418267726898,
      "learning_rate": 2.7414279951593386e-05,
      "loss": 0.17103816986083983,
      "step": 5600
    },
    {
      "epoch": 2.279144816458249,
      "grad_norm": 0.12691690027713776,
      "learning_rate": 2.7212585720048407e-05,
      "loss": 0.15552584648132325,
      "step": 5650
    },
    {
      "epoch": 2.2993142396127473,
      "grad_norm": 2.5150461196899414,
      "learning_rate": 2.701089148850343e-05,
      "loss": 0.1537048625946045,
      "step": 5700
    },
    {
      "epoch": 2.319483662767245,
      "grad_norm": 0.4236988127231598,
      "learning_rate": 2.6809197256958452e-05,
      "loss": 0.15081631660461425,
      "step": 5750
    },
    {
      "epoch": 2.339653085921743,
      "grad_norm": 4.3972907066345215,
      "learning_rate": 2.6607503025413476e-05,
      "loss": 0.15966953277587892,
      "step": 5800
    },
    {
      "epoch": 2.3598225090762406,
      "grad_norm": 0.20289625227451324,
      "learning_rate": 2.6405808793868496e-05,
      "loss": 0.20613027572631837,
      "step": 5850
    },
    {
      "epoch": 2.3799919322307383,
      "grad_norm": 13.914892196655273,
      "learning_rate": 2.6204114562323517e-05,
      "loss": 0.1521580123901367,
      "step": 5900
    },
    {
      "epoch": 2.400161355385236,
      "grad_norm": 1.8845982551574707,
      "learning_rate": 2.6002420330778545e-05,
      "loss": 0.18961977005004882,
      "step": 5950
    },
    {
      "epoch": 2.420330778539734,
      "grad_norm": 6.2201433181762695,
      "learning_rate": 2.5800726099233562e-05,
      "loss": 0.17040407180786132,
      "step": 6000
    },
    {
      "epoch": 2.4405002016942317,
      "grad_norm": 1.7300294637680054,
      "learning_rate": 2.5599031867688582e-05,
      "loss": 0.15661882400512694,
      "step": 6050
    },
    {
      "epoch": 2.4606696248487294,
      "grad_norm": 0.38582950830459595,
      "learning_rate": 2.539733763614361e-05,
      "loss": 0.18057403564453126,
      "step": 6100
    },
    {
      "epoch": 2.480839048003227,
      "grad_norm": 3.5432305335998535,
      "learning_rate": 2.519564340459863e-05,
      "loss": 0.21012399673461915,
      "step": 6150
    },
    {
      "epoch": 2.501008471157725,
      "grad_norm": 3.0747153759002686,
      "learning_rate": 2.499394917305365e-05,
      "loss": 0.1722999954223633,
      "step": 6200
    },
    {
      "epoch": 2.5211778943122227,
      "grad_norm": 1.0176174640655518,
      "learning_rate": 2.4792254941508675e-05,
      "loss": 0.19214908599853517,
      "step": 6250
    },
    {
      "epoch": 2.5413473174667205,
      "grad_norm": 1.0402295589447021,
      "learning_rate": 2.45905607099637e-05,
      "loss": 0.14209456443786622,
      "step": 6300
    },
    {
      "epoch": 2.5615167406212183,
      "grad_norm": 0.44618988037109375,
      "learning_rate": 2.4388866478418716e-05,
      "loss": 0.14631601333618163,
      "step": 6350
    },
    {
      "epoch": 2.581686163775716,
      "grad_norm": 0.32430288195610046,
      "learning_rate": 2.418717224687374e-05,
      "loss": 0.15652142524719237,
      "step": 6400
    },
    {
      "epoch": 2.601855586930214,
      "grad_norm": 4.26512336730957,
      "learning_rate": 2.3985478015328765e-05,
      "loss": 0.17958589553833007,
      "step": 6450
    },
    {
      "epoch": 2.6220250100847116,
      "grad_norm": 5.863807678222656,
      "learning_rate": 2.3783783783783785e-05,
      "loss": 0.12491923332214355,
      "step": 6500
    },
    {
      "epoch": 2.6421944332392093,
      "grad_norm": 1.6168506145477295,
      "learning_rate": 2.3582089552238806e-05,
      "loss": 0.16195316314697267,
      "step": 6550
    },
    {
      "epoch": 2.662363856393707,
      "grad_norm": 0.24692083895206451,
      "learning_rate": 2.338039532069383e-05,
      "loss": 0.21834095001220702,
      "step": 6600
    },
    {
      "epoch": 2.682533279548205,
      "grad_norm": 2.46010684967041,
      "learning_rate": 2.317870108914885e-05,
      "loss": 0.14829426765441894,
      "step": 6650
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 2.4015772342681885,
      "learning_rate": 2.2977006857603875e-05,
      "loss": 0.19773597717285157,
      "step": 6700
    },
    {
      "epoch": 2.7228721258572004,
      "grad_norm": 0.30142679810523987,
      "learning_rate": 2.2775312626058895e-05,
      "loss": 0.19015130996704102,
      "step": 6750
    },
    {
      "epoch": 2.743041549011698,
      "grad_norm": 3.2308762073516846,
      "learning_rate": 2.257361839451392e-05,
      "loss": 0.15893199920654297,
      "step": 6800
    },
    {
      "epoch": 2.763210972166196,
      "grad_norm": 4.456058979034424,
      "learning_rate": 2.237192416296894e-05,
      "loss": 0.21003019332885742,
      "step": 6850
    },
    {
      "epoch": 2.7833803953206937,
      "grad_norm": 0.072988361120224,
      "learning_rate": 2.2170229931423964e-05,
      "loss": 0.17291305541992188,
      "step": 6900
    },
    {
      "epoch": 2.8035498184751915,
      "grad_norm": 1.3505617380142212,
      "learning_rate": 2.1968535699878985e-05,
      "loss": 0.16227161407470703,
      "step": 6950
    },
    {
      "epoch": 2.8237192416296892,
      "grad_norm": 6.348002910614014,
      "learning_rate": 2.1766841468334005e-05,
      "loss": 0.18089292526245118,
      "step": 7000
    },
    {
      "epoch": 2.8438886647841874,
      "grad_norm": 4.224061965942383,
      "learning_rate": 2.156514723678903e-05,
      "loss": 0.21152067184448242,
      "step": 7050
    },
    {
      "epoch": 2.8640580879386848,
      "grad_norm": 6.137605667114258,
      "learning_rate": 2.1363453005244053e-05,
      "loss": 0.1759657669067383,
      "step": 7100
    },
    {
      "epoch": 2.884227511093183,
      "grad_norm": 0.8793334364891052,
      "learning_rate": 2.1161758773699074e-05,
      "loss": 0.1867251205444336,
      "step": 7150
    },
    {
      "epoch": 2.9043969342476803,
      "grad_norm": 6.664191246032715,
      "learning_rate": 2.0960064542154095e-05,
      "loss": 0.16023876190185546,
      "step": 7200
    },
    {
      "epoch": 2.9245663574021785,
      "grad_norm": 7.107497692108154,
      "learning_rate": 2.075837031060912e-05,
      "loss": 0.16876220703125,
      "step": 7250
    },
    {
      "epoch": 2.944735780556676,
      "grad_norm": 6.290689468383789,
      "learning_rate": 2.0556676079064143e-05,
      "loss": 0.14115575790405274,
      "step": 7300
    },
    {
      "epoch": 2.964905203711174,
      "grad_norm": 2.726917266845703,
      "learning_rate": 2.035498184751916e-05,
      "loss": 0.15181530952453615,
      "step": 7350
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.1575377732515335,
      "learning_rate": 2.0153287615974184e-05,
      "loss": 0.204818058013916,
      "step": 7400
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9469147005444646,
      "eval_loss": 0.20167165994644165,
      "eval_runtime": 20.7163,
      "eval_samples_per_second": 212.78,
      "eval_steps_per_second": 13.323,
      "step": 7437
    },
    {
      "epoch": 3.0052440500201696,
      "grad_norm": 4.361955165863037,
      "learning_rate": 1.9951593384429208e-05,
      "loss": 0.1282938003540039,
      "step": 7450
    },
    {
      "epoch": 3.0254134731746674,
      "grad_norm": 2.4413537979125977,
      "learning_rate": 1.974989915288423e-05,
      "loss": 0.14994653701782226,
      "step": 7500
    },
    {
      "epoch": 3.045582896329165,
      "grad_norm": 0.9888622760772705,
      "learning_rate": 1.954820492133925e-05,
      "loss": 0.18875495910644532,
      "step": 7550
    },
    {
      "epoch": 3.065752319483663,
      "grad_norm": 1.7957770824432373,
      "learning_rate": 1.9346510689794273e-05,
      "loss": 0.16345306396484374,
      "step": 7600
    },
    {
      "epoch": 3.0859217426381607,
      "grad_norm": 10.348389625549316,
      "learning_rate": 1.9144816458249294e-05,
      "loss": 0.16646900177001953,
      "step": 7650
    },
    {
      "epoch": 3.1060911657926584,
      "grad_norm": 1.3166776895523071,
      "learning_rate": 1.8943122226704318e-05,
      "loss": 0.11046904563903809,
      "step": 7700
    },
    {
      "epoch": 3.126260588947156,
      "grad_norm": 4.864676475524902,
      "learning_rate": 1.874142799515934e-05,
      "loss": 0.14641575813293456,
      "step": 7750
    },
    {
      "epoch": 3.146430012101654,
      "grad_norm": 0.590563178062439,
      "learning_rate": 1.8539733763614363e-05,
      "loss": 0.1471682071685791,
      "step": 7800
    },
    {
      "epoch": 3.1665994352561517,
      "grad_norm": 2.9072699546813965,
      "learning_rate": 1.8338039532069383e-05,
      "loss": 0.14556562423706054,
      "step": 7850
    },
    {
      "epoch": 3.1867688584106495,
      "grad_norm": 1.3315081596374512,
      "learning_rate": 1.8136345300524407e-05,
      "loss": 0.09854414939880371,
      "step": 7900
    },
    {
      "epoch": 3.2069382815651473,
      "grad_norm": 5.771864414215088,
      "learning_rate": 1.7934651068979428e-05,
      "loss": 0.18490333557128907,
      "step": 7950
    },
    {
      "epoch": 3.227107704719645,
      "grad_norm": 1.54212486743927,
      "learning_rate": 1.773295683743445e-05,
      "loss": 0.11037788391113282,
      "step": 8000
    },
    {
      "epoch": 3.247277127874143,
      "grad_norm": 0.1329752802848816,
      "learning_rate": 1.7531262605889473e-05,
      "loss": 0.150218448638916,
      "step": 8050
    },
    {
      "epoch": 3.2674465510286406,
      "grad_norm": 4.867387771606445,
      "learning_rate": 1.7329568374344497e-05,
      "loss": 0.17623374938964845,
      "step": 8100
    },
    {
      "epoch": 3.2876159741831383,
      "grad_norm": 4.709106922149658,
      "learning_rate": 1.7127874142799517e-05,
      "loss": 0.1503092098236084,
      "step": 8150
    },
    {
      "epoch": 3.307785397337636,
      "grad_norm": 2.4587321281433105,
      "learning_rate": 1.6926179911254538e-05,
      "loss": 0.17221961975097655,
      "step": 8200
    },
    {
      "epoch": 3.327954820492134,
      "grad_norm": 0.18433253467082977,
      "learning_rate": 1.6724485679709562e-05,
      "loss": 0.15626630783081055,
      "step": 8250
    },
    {
      "epoch": 3.3481242436466316,
      "grad_norm": 5.018091678619385,
      "learning_rate": 1.6522791448164586e-05,
      "loss": 0.11743727684020996,
      "step": 8300
    },
    {
      "epoch": 3.3682936668011294,
      "grad_norm": 5.374708652496338,
      "learning_rate": 1.6321097216619603e-05,
      "loss": 0.1508438777923584,
      "step": 8350
    },
    {
      "epoch": 3.388463089955627,
      "grad_norm": 2.271517753601074,
      "learning_rate": 1.6119402985074627e-05,
      "loss": 0.13781851768493653,
      "step": 8400
    },
    {
      "epoch": 3.408632513110125,
      "grad_norm": 5.559137344360352,
      "learning_rate": 1.591770875352965e-05,
      "loss": 0.11309442520141602,
      "step": 8450
    },
    {
      "epoch": 3.4288019362646227,
      "grad_norm": 0.41697025299072266,
      "learning_rate": 1.5716014521984672e-05,
      "loss": 0.1275617790222168,
      "step": 8500
    },
    {
      "epoch": 3.4489713594191205,
      "grad_norm": 2.4398200511932373,
      "learning_rate": 1.5514320290439693e-05,
      "loss": 0.11772675514221191,
      "step": 8550
    },
    {
      "epoch": 3.4691407825736182,
      "grad_norm": 5.258758544921875,
      "learning_rate": 1.5312626058894717e-05,
      "loss": 0.1622784423828125,
      "step": 8600
    },
    {
      "epoch": 3.489310205728116,
      "grad_norm": 3.7077977657318115,
      "learning_rate": 1.5110931827349739e-05,
      "loss": 0.17116220474243163,
      "step": 8650
    },
    {
      "epoch": 3.509479628882614,
      "grad_norm": 0.45965898036956787,
      "learning_rate": 1.4909237595804761e-05,
      "loss": 0.16231245040893555,
      "step": 8700
    },
    {
      "epoch": 3.5296490520371115,
      "grad_norm": 1.7838913202285767,
      "learning_rate": 1.4707543364259782e-05,
      "loss": 0.1521257972717285,
      "step": 8750
    },
    {
      "epoch": 3.5498184751916098,
      "grad_norm": 0.8296908736228943,
      "learning_rate": 1.4505849132714804e-05,
      "loss": 0.1450238609313965,
      "step": 8800
    },
    {
      "epoch": 3.569987898346107,
      "grad_norm": 2.603304862976074,
      "learning_rate": 1.4304154901169828e-05,
      "loss": 0.18242509841918944,
      "step": 8850
    },
    {
      "epoch": 3.5901573215006053,
      "grad_norm": 0.5085131525993347,
      "learning_rate": 1.410246066962485e-05,
      "loss": 0.1511705684661865,
      "step": 8900
    },
    {
      "epoch": 3.6103267446551026,
      "grad_norm": 0.39615824818611145,
      "learning_rate": 1.3900766438079871e-05,
      "loss": 0.18586685180664062,
      "step": 8950
    },
    {
      "epoch": 3.630496167809601,
      "grad_norm": 3.9840400218963623,
      "learning_rate": 1.3699072206534894e-05,
      "loss": 0.12510306358337403,
      "step": 9000
    },
    {
      "epoch": 3.650665590964098,
      "grad_norm": 1.856611728668213,
      "learning_rate": 1.3497377974989916e-05,
      "loss": 0.13308097839355468,
      "step": 9050
    },
    {
      "epoch": 3.6708350141185964,
      "grad_norm": 3.8172595500946045,
      "learning_rate": 1.329568374344494e-05,
      "loss": 0.15418212890625,
      "step": 9100
    },
    {
      "epoch": 3.691004437273094,
      "grad_norm": 1.3330752849578857,
      "learning_rate": 1.3093989511899959e-05,
      "loss": 0.11185816764831542,
      "step": 9150
    },
    {
      "epoch": 3.711173860427592,
      "grad_norm": 5.0063652992248535,
      "learning_rate": 1.2892295280354983e-05,
      "loss": 0.14064055442810058,
      "step": 9200
    },
    {
      "epoch": 3.7313432835820897,
      "grad_norm": 5.983749866485596,
      "learning_rate": 1.2690601048810005e-05,
      "loss": 0.1424097728729248,
      "step": 9250
    },
    {
      "epoch": 3.7515127067365874,
      "grad_norm": 0.6912927627563477,
      "learning_rate": 1.2488906817265026e-05,
      "loss": 0.1355896282196045,
      "step": 9300
    },
    {
      "epoch": 3.771682129891085,
      "grad_norm": 0.789875864982605,
      "learning_rate": 1.228721258572005e-05,
      "loss": 0.1569892692565918,
      "step": 9350
    },
    {
      "epoch": 3.791851553045583,
      "grad_norm": 25.235803604125977,
      "learning_rate": 1.208551835417507e-05,
      "loss": 0.16975107192993164,
      "step": 9400
    },
    {
      "epoch": 3.8120209762000807,
      "grad_norm": 10.807896614074707,
      "learning_rate": 1.1883824122630095e-05,
      "loss": 0.12455953598022461,
      "step": 9450
    },
    {
      "epoch": 3.8321903993545785,
      "grad_norm": 1.9222596883773804,
      "learning_rate": 1.1682129891085115e-05,
      "loss": 0.13488195419311524,
      "step": 9500
    },
    {
      "epoch": 3.8523598225090763,
      "grad_norm": 2.0477261543273926,
      "learning_rate": 1.1480435659540138e-05,
      "loss": 0.1264381980895996,
      "step": 9550
    },
    {
      "epoch": 3.872529245663574,
      "grad_norm": 0.35994377732276917,
      "learning_rate": 1.127874142799516e-05,
      "loss": 0.1540088176727295,
      "step": 9600
    },
    {
      "epoch": 3.892698668818072,
      "grad_norm": 4.0362019538879395,
      "learning_rate": 1.1077047196450182e-05,
      "loss": 0.15407589912414552,
      "step": 9650
    },
    {
      "epoch": 3.9128680919725696,
      "grad_norm": 0.061721403151750565,
      "learning_rate": 1.0875352964905205e-05,
      "loss": 0.11756512641906738,
      "step": 9700
    },
    {
      "epoch": 3.9330375151270673,
      "grad_norm": 9.173218727111816,
      "learning_rate": 1.0673658733360227e-05,
      "loss": 0.12152119636535645,
      "step": 9750
    },
    {
      "epoch": 3.953206938281565,
      "grad_norm": 1.0651715993881226,
      "learning_rate": 1.0471964501815248e-05,
      "loss": 0.13009054183959962,
      "step": 9800
    },
    {
      "epoch": 3.973376361436063,
      "grad_norm": 8.925455093383789,
      "learning_rate": 1.0270270270270272e-05,
      "loss": 0.13634368896484375,
      "step": 9850
    },
    {
      "epoch": 3.9935457845905606,
      "grad_norm": 7.055464267730713,
      "learning_rate": 1.0068576038725292e-05,
      "loss": 0.12510560989379882,
      "step": 9900
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9498638838475499,
      "eval_loss": 0.19086875021457672,
      "eval_runtime": 20.6378,
      "eval_samples_per_second": 213.589,
      "eval_steps_per_second": 13.374,
      "step": 9916
    },
    {
      "epoch": 4.013715207745059,
      "grad_norm": 1.8108693361282349,
      "learning_rate": 9.866881807180316e-06,
      "loss": 0.13720301628112794,
      "step": 9950
    },
    {
      "epoch": 4.033884630899556,
      "grad_norm": 1.0627018213272095,
      "learning_rate": 9.665187575635337e-06,
      "loss": 0.13952345848083497,
      "step": 10000
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 0.9664111733436584,
      "learning_rate": 9.46349334409036e-06,
      "loss": 0.15156702995300292,
      "step": 10050
    },
    {
      "epoch": 4.074223477208552,
      "grad_norm": 43.628841400146484,
      "learning_rate": 9.261799112545382e-06,
      "loss": 0.12908251762390135,
      "step": 10100
    },
    {
      "epoch": 4.09439290036305,
      "grad_norm": 1.320526123046875,
      "learning_rate": 9.060104881000404e-06,
      "loss": 0.13279842376708983,
      "step": 10150
    },
    {
      "epoch": 4.114562323517547,
      "grad_norm": 2.7222747802734375,
      "learning_rate": 8.858410649455426e-06,
      "loss": 0.0985030746459961,
      "step": 10200
    },
    {
      "epoch": 4.1347317466720455,
      "grad_norm": 14.308988571166992,
      "learning_rate": 8.656716417910449e-06,
      "loss": 0.14445188522338867,
      "step": 10250
    },
    {
      "epoch": 4.154901169826543,
      "grad_norm": 11.831380844116211,
      "learning_rate": 8.45502218636547e-06,
      "loss": 0.1422930145263672,
      "step": 10300
    },
    {
      "epoch": 4.175070592981041,
      "grad_norm": 0.28589361906051636,
      "learning_rate": 8.253327954820494e-06,
      "loss": 0.12077164649963379,
      "step": 10350
    },
    {
      "epoch": 4.195240016135538,
      "grad_norm": 1.5173996686935425,
      "learning_rate": 8.051633723275514e-06,
      "loss": 0.09480426788330078,
      "step": 10400
    },
    {
      "epoch": 4.2154094392900365,
      "grad_norm": 3.5482044219970703,
      "learning_rate": 7.849939491730538e-06,
      "loss": 0.13487308502197265,
      "step": 10450
    },
    {
      "epoch": 4.235578862444534,
      "grad_norm": 4.181003093719482,
      "learning_rate": 7.648245260185559e-06,
      "loss": 0.10856437683105469,
      "step": 10500
    },
    {
      "epoch": 4.255748285599032,
      "grad_norm": 3.127241849899292,
      "learning_rate": 7.446551028640582e-06,
      "loss": 0.11362249374389649,
      "step": 10550
    },
    {
      "epoch": 4.275917708753529,
      "grad_norm": 16.710121154785156,
      "learning_rate": 7.2448567970956035e-06,
      "loss": 0.16090269088745118,
      "step": 10600
    },
    {
      "epoch": 4.296087131908028,
      "grad_norm": 7.735565662384033,
      "learning_rate": 7.043162565550626e-06,
      "loss": 0.1192931079864502,
      "step": 10650
    },
    {
      "epoch": 4.316256555062525,
      "grad_norm": 1.4519258737564087,
      "learning_rate": 6.841468334005647e-06,
      "loss": 0.1244455623626709,
      "step": 10700
    },
    {
      "epoch": 4.336425978217023,
      "grad_norm": 0.2084610015153885,
      "learning_rate": 6.6397741024606705e-06,
      "loss": 0.12691824913024902,
      "step": 10750
    },
    {
      "epoch": 4.3565954013715205,
      "grad_norm": 5.141778469085693,
      "learning_rate": 6.438079870915692e-06,
      "loss": 0.15231054306030273,
      "step": 10800
    },
    {
      "epoch": 4.376764824526019,
      "grad_norm": 4.018757343292236,
      "learning_rate": 6.236385639370714e-06,
      "loss": 0.13898849487304688,
      "step": 10850
    },
    {
      "epoch": 4.396934247680516,
      "grad_norm": 2.3020405769348145,
      "learning_rate": 6.034691407825737e-06,
      "loss": 0.11184153556823731,
      "step": 10900
    },
    {
      "epoch": 4.417103670835014,
      "grad_norm": 0.3617361783981323,
      "learning_rate": 5.832997176280758e-06,
      "loss": 0.12001494407653808,
      "step": 10950
    },
    {
      "epoch": 4.4372730939895115,
      "grad_norm": 3.145230531692505,
      "learning_rate": 5.6313029447357805e-06,
      "loss": 0.111842679977417,
      "step": 11000
    },
    {
      "epoch": 4.45744251714401,
      "grad_norm": 6.321645736694336,
      "learning_rate": 5.429608713190803e-06,
      "loss": 0.1446524429321289,
      "step": 11050
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 0.19845882058143616,
      "learning_rate": 5.227914481645825e-06,
      "loss": 0.10471660614013673,
      "step": 11100
    },
    {
      "epoch": 4.497781363453005,
      "grad_norm": 14.234234809875488,
      "learning_rate": 5.0262202501008476e-06,
      "loss": 0.13982812881469728,
      "step": 11150
    },
    {
      "epoch": 4.5179507866075035,
      "grad_norm": 6.142904281616211,
      "learning_rate": 4.824526018555869e-06,
      "loss": 0.14638319969177246,
      "step": 11200
    },
    {
      "epoch": 4.538120209762001,
      "grad_norm": 10.167677879333496,
      "learning_rate": 4.622831787010891e-06,
      "loss": 0.12575898170471192,
      "step": 11250
    },
    {
      "epoch": 4.558289632916498,
      "grad_norm": 0.4165036082267761,
      "learning_rate": 4.421137555465914e-06,
      "loss": 0.10566370964050292,
      "step": 11300
    },
    {
      "epoch": 4.578459056070996,
      "grad_norm": 0.9815388321876526,
      "learning_rate": 4.219443323920936e-06,
      "loss": 0.1358126163482666,
      "step": 11350
    },
    {
      "epoch": 4.598628479225495,
      "grad_norm": 3.1989729404449463,
      "learning_rate": 4.017749092375958e-06,
      "loss": 0.1418777275085449,
      "step": 11400
    },
    {
      "epoch": 4.618797902379992,
      "grad_norm": 2.143892288208008,
      "learning_rate": 3.81605486083098e-06,
      "loss": 0.13869763374328614,
      "step": 11450
    },
    {
      "epoch": 4.63896732553449,
      "grad_norm": 1.0389790534973145,
      "learning_rate": 3.6143606292860022e-06,
      "loss": 0.20401397705078125,
      "step": 11500
    },
    {
      "epoch": 4.659136748688987,
      "grad_norm": 8.121844291687012,
      "learning_rate": 3.4126663977410246e-06,
      "loss": 0.13985713005065917,
      "step": 11550
    },
    {
      "epoch": 4.679306171843486,
      "grad_norm": 18.19819450378418,
      "learning_rate": 3.210972166196047e-06,
      "loss": 0.12457592010498048,
      "step": 11600
    },
    {
      "epoch": 4.699475594997983,
      "grad_norm": 0.9452546238899231,
      "learning_rate": 3.009277934651069e-06,
      "loss": 0.14423842430114747,
      "step": 11650
    },
    {
      "epoch": 4.719645018152481,
      "grad_norm": 0.4873313307762146,
      "learning_rate": 2.807583703106091e-06,
      "loss": 0.11125797271728516,
      "step": 11700
    },
    {
      "epoch": 4.7398144413069785,
      "grad_norm": 0.23056188225746155,
      "learning_rate": 2.6058894715611135e-06,
      "loss": 0.10489352226257324,
      "step": 11750
    },
    {
      "epoch": 4.759983864461477,
      "grad_norm": 4.3498921394348145,
      "learning_rate": 2.4041952400161354e-06,
      "loss": 0.11880455017089844,
      "step": 11800
    },
    {
      "epoch": 4.780153287615974,
      "grad_norm": 6.197351932525635,
      "learning_rate": 2.2025010084711578e-06,
      "loss": 0.11629629135131836,
      "step": 11850
    },
    {
      "epoch": 4.800322710770472,
      "grad_norm": 1.4869271516799927,
      "learning_rate": 2.0008067769261797e-06,
      "loss": 0.06002850532531738,
      "step": 11900
    },
    {
      "epoch": 4.82049213392497,
      "grad_norm": 0.11449511349201202,
      "learning_rate": 1.799112545381202e-06,
      "loss": 0.09682181358337402,
      "step": 11950
    },
    {
      "epoch": 4.840661557079468,
      "grad_norm": 0.537159264087677,
      "learning_rate": 1.5974183138362242e-06,
      "loss": 0.1060289478302002,
      "step": 12000
    },
    {
      "epoch": 4.860830980233965,
      "grad_norm": 1.7972829341888428,
      "learning_rate": 1.3957240822912465e-06,
      "loss": 0.1122105884552002,
      "step": 12050
    },
    {
      "epoch": 4.881000403388463,
      "grad_norm": 0.5588096976280212,
      "learning_rate": 1.1940298507462686e-06,
      "loss": 0.20083948135375976,
      "step": 12100
    },
    {
      "epoch": 4.901169826542961,
      "grad_norm": 0.34725531935691833,
      "learning_rate": 9.92335619201291e-07,
      "loss": 0.1103892421722412,
      "step": 12150
    },
    {
      "epoch": 4.921339249697459,
      "grad_norm": 4.558995246887207,
      "learning_rate": 7.906413876563131e-07,
      "loss": 0.09908536911010742,
      "step": 12200
    },
    {
      "epoch": 4.941508672851956,
      "grad_norm": 8.438501358032227,
      "learning_rate": 5.889471561113352e-07,
      "loss": 0.13646589279174803,
      "step": 12250
    },
    {
      "epoch": 4.961678096006454,
      "grad_norm": 4.4897847175598145,
      "learning_rate": 3.872529245663574e-07,
      "loss": 0.11858689308166503,
      "step": 12300
    },
    {
      "epoch": 4.981847519160952,
      "grad_norm": 3.2453246116638184,
      "learning_rate": 1.8555869302137961e-07,
      "loss": 0.11941192626953125,
      "step": 12350
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9532667876588021,
      "eval_loss": 0.1782979518175125,
      "eval_runtime": 20.7187,
      "eval_samples_per_second": 212.755,
      "eval_steps_per_second": 13.321,
      "step": 12395
    }
  ],
  "logging_steps": 50,
  "max_steps": 12395,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 4.227186605693829e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}