{
  "best_global_step": 23000,
  "best_metric": 0.24171218276023865,
  "best_model_checkpoint": "/output/checkpoint-23000",
  "epoch": 0.3441410454593488,
  "eval_steps": 1000,
  "global_step": 23000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.995392406731844,
      "epoch": 0.00014962654150406469,
      "grad_norm": 1.6540336608886719,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.6699,
      "mean_token_accuracy": 0.8633284188807011,
      "num_tokens": 74655.0,
      "step": 10
    },
    {
      "entropy": 1.9947461523115635,
      "epoch": 0.00029925308300812937,
      "grad_norm": 1.5377693176269531,
      "learning_rate": 6.333333333333333e-07,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8691052194684744,
      "num_tokens": 148980.0,
      "step": 20
    },
    {
      "entropy": 2.00217300131917,
      "epoch": 0.0004488796245121941,
      "grad_norm": 1.5561202764511108,
      "learning_rate": 9.666666666666668e-07,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8677759364247322,
      "num_tokens": 223581.0,
      "step": 30
    },
    {
      "entropy": 1.993088324368,
      "epoch": 0.0005985061660162587,
      "grad_norm": 1.4348832368850708,
      "learning_rate": 1.3e-06,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8703481316566467,
      "num_tokens": 301292.0,
      "step": 40
    },
    {
      "entropy": 1.9850200325250626,
      "epoch": 0.0007481327075203235,
      "grad_norm": 1.510475754737854,
      "learning_rate": 1.6333333333333333e-06,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8701431334018708,
      "num_tokens": 373829.0,
      "step": 50
    },
    {
      "entropy": 2.0068949691951277,
      "epoch": 0.0008977592490243882,
      "grad_norm": 1.2555310726165771,
      "learning_rate": 1.9666666666666668e-06,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8808745600283145,
      "num_tokens": 446456.0,
      "step": 60
    },
    {
      "entropy": 1.9884536407887936,
      "epoch": 0.0010473857905284529,
      "grad_norm": 0.8961914777755737,
      "learning_rate": 2.3e-06,
      "loss": 0.54,
      "mean_token_accuracy": 0.8860756240785121,
      "num_tokens": 519100.0,
      "step": 70
    },
    {
      "entropy": 1.9966150291264058,
      "epoch": 0.0011970123320325175,
      "grad_norm": 0.5592026710510254,
      "learning_rate": 2.6333333333333337e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.9029676467180252,
      "num_tokens": 593933.0,
      "step": 80
    },
    {
      "entropy": 1.971697284281254,
      "epoch": 0.0013466388735365823,
      "grad_norm": 0.7416914105415344,
      "learning_rate": 2.966666666666667e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.9083496272563935,
      "num_tokens": 668090.0,
      "step": 90
    },
    {
      "entropy": 2.020202798396349,
      "epoch": 0.001496265415040647,
      "grad_norm": 0.5684888958930969,
      "learning_rate": 3.3e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.9050738759338856,
      "num_tokens": 742570.0,
      "step": 100
    },
    {
      "entropy": 2.0086157262325286,
      "epoch": 0.0016458919565447117,
      "grad_norm": 0.6281460523605347,
      "learning_rate": 3.633333333333334e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.911510188318789,
      "num_tokens": 817427.0,
      "step": 110
    },
    {
      "entropy": 2.0083575293421747,
      "epoch": 0.0017955184980487763,
      "grad_norm": 0.4346252679824829,
      "learning_rate": 3.966666666666667e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.9160505689680576,
      "num_tokens": 894027.0,
      "step": 120
    },
    {
      "entropy": 2.027063339948654,
      "epoch": 0.0019451450395528412,
      "grad_norm": 0.4729572832584381,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.9123359378427267,
      "num_tokens": 968082.0,
      "step": 130
    },
    {
      "entropy": 2.0357840031385424,
      "epoch": 0.0020947715810569058,
      "grad_norm": 0.4820358157157898,
      "learning_rate": 4.633333333333334e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.9201525665819645,
      "num_tokens": 1041111.0,
      "step": 140
    },
    {
      "entropy": 2.0041785940527914,
      "epoch": 0.0022443981225609704,
      "grad_norm": 0.4512932002544403,
      "learning_rate": 4.966666666666667e-06,
      "loss": 0.3719,
      "mean_token_accuracy": 0.9196753237396479,
      "num_tokens": 1114683.0,
      "step": 150
    },
    {
      "entropy": 1.986509595066309,
      "epoch": 0.002394024664065035,
      "grad_norm": 0.49413782358169556,
      "learning_rate": 5.3e-06,
      "loss": 0.3633,
      "mean_token_accuracy": 0.9218655321747065,
      "num_tokens": 1189363.0,
      "step": 160
    },
    {
      "entropy": 2.008907197415829,
      "epoch": 0.0025436512055691,
      "grad_norm": 0.48253825306892395,
      "learning_rate": 5.633333333333333e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.9201793625950814,
      "num_tokens": 1262509.0,
      "step": 170
    },
    {
      "entropy": 2.0038809821009638,
      "epoch": 0.0026932777470731646,
      "grad_norm": 0.5222899317741394,
      "learning_rate": 5.9666666666666666e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.9163873542100192,
      "num_tokens": 1335852.0,
      "step": 180
    },
    {
      "entropy": 1.9737722158432007,
      "epoch": 0.0028429042885772292,
      "grad_norm": 0.6734032034873962,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.9171450879424811,
      "num_tokens": 1410353.0,
      "step": 190
    },
    {
      "entropy": 1.9792789742350578,
      "epoch": 0.002992530830081294,
      "grad_norm": 0.4623475670814514,
      "learning_rate": 6.633333333333333e-06,
      "loss": 0.3541,
      "mean_token_accuracy": 0.9202779714018107,
      "num_tokens": 1484450.0,
      "step": 200
    },
    {
      "entropy": 1.9741366975009442,
      "epoch": 0.003142157371585359,
      "grad_norm": 0.5829670429229736,
      "learning_rate": 6.966666666666667e-06,
      "loss": 0.3482,
      "mean_token_accuracy": 0.9214173048734665,
      "num_tokens": 1558025.0,
      "step": 210
    },
    {
      "entropy": 1.9535197712481023,
      "epoch": 0.0032917839130894235,
      "grad_norm": 0.46482422947883606,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.3219,
      "mean_token_accuracy": 0.9295845787972212,
      "num_tokens": 1635960.0,
      "step": 220
    },
    {
      "entropy": 1.985553939640522,
      "epoch": 0.003441410454593488,
      "grad_norm": 0.5042750239372253,
      "learning_rate": 7.633333333333334e-06,
      "loss": 0.3468,
      "mean_token_accuracy": 0.9231617078185081,
      "num_tokens": 1712553.0,
      "step": 230
    },
    {
      "entropy": 1.99157450273633,
      "epoch": 0.0035910369960975527,
      "grad_norm": 0.5442078113555908,
      "learning_rate": 7.966666666666666e-06,
      "loss": 0.3359,
      "mean_token_accuracy": 0.9257699001580477,
      "num_tokens": 1788108.0,
      "step": 240
    },
    {
      "entropy": 2.0001078858971595,
      "epoch": 0.0037406635376016173,
      "grad_norm": 0.5726729035377502,
      "learning_rate": 8.3e-06,
      "loss": 0.3264,
      "mean_token_accuracy": 0.9232569672167301,
      "num_tokens": 1861212.0,
      "step": 250
    },
    {
      "entropy": 1.9762448623776436,
      "epoch": 0.0038902900791056823,
      "grad_norm": 0.48842018842697144,
      "learning_rate": 8.633333333333334e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.9280190449208021,
      "num_tokens": 1935627.0,
      "step": 260
    },
    {
      "entropy": 1.9863562628626823,
      "epoch": 0.004039916620609747,
      "grad_norm": 0.5623534321784973,
      "learning_rate": 8.966666666666668e-06,
      "loss": 0.3313,
      "mean_token_accuracy": 0.9267206363379955,
      "num_tokens": 2012173.0,
      "step": 270
    },
    {
      "entropy": 2.030895411223173,
      "epoch": 0.0041895431621138115,
      "grad_norm": 0.7707616090774536,
      "learning_rate": 9.3e-06,
      "loss": 0.3494,
      "mean_token_accuracy": 0.9233444124460221,
      "num_tokens": 2085239.0,
      "step": 280
    },
    {
      "entropy": 2.0427174262702463,
      "epoch": 0.004339169703617876,
      "grad_norm": 0.47786983847618103,
      "learning_rate": 9.633333333333335e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.9239508364349603,
      "num_tokens": 2158990.0,
      "step": 290
    },
    {
      "entropy": 2.0073467031121255,
      "epoch": 0.004488796245121941,
      "grad_norm": 0.49501946568489075,
      "learning_rate": 9.966666666666667e-06,
      "loss": 0.2997,
      "mean_token_accuracy": 0.932453353330493,
      "num_tokens": 2234311.0,
      "step": 300
    },
    {
      "entropy": 2.011546725779772,
      "epoch": 0.004638422786626005,
      "grad_norm": 0.5904964804649353,
      "learning_rate": 1.03e-05,
      "loss": 0.3373,
      "mean_token_accuracy": 0.9201153438538313,
      "num_tokens": 2305473.0,
      "step": 310
    },
    {
      "entropy": 1.9916895642876624,
      "epoch": 0.00478804932813007,
      "grad_norm": 0.5510156750679016,
      "learning_rate": 1.0633333333333334e-05,
      "loss": 0.3407,
      "mean_token_accuracy": 0.9223257884383201,
      "num_tokens": 2379145.0,
      "step": 320
    },
    {
      "entropy": 1.9839037261903285,
      "epoch": 0.004937675869634135,
      "grad_norm": 0.6314477920532227,
      "learning_rate": 1.0966666666666666e-05,
      "loss": 0.3069,
      "mean_token_accuracy": 0.9325135432183742,
      "num_tokens": 2452002.0,
      "step": 330
    },
    {
      "entropy": 1.9725721523165702,
      "epoch": 0.0050873024111382,
      "grad_norm": 0.6581112146377563,
      "learning_rate": 1.13e-05,
      "loss": 0.303,
      "mean_token_accuracy": 0.92990313321352,
      "num_tokens": 2526388.0,
      "step": 340
    },
    {
      "entropy": 1.9570358380675317,
      "epoch": 0.005236928952642265,
      "grad_norm": 0.6892510652542114,
      "learning_rate": 1.1633333333333334e-05,
      "loss": 0.2994,
      "mean_token_accuracy": 0.930433790013194,
      "num_tokens": 2601496.0,
      "step": 350
    },
    {
      "entropy": 1.9567357785999775,
      "epoch": 0.005386555494146329,
      "grad_norm": 0.5583544373512268,
      "learning_rate": 1.1966666666666668e-05,
      "loss": 0.3068,
      "mean_token_accuracy": 0.9284893423318863,
      "num_tokens": 2673864.0,
      "step": 360
    },
    {
      "entropy": 1.940490322560072,
      "epoch": 0.005536182035650394,
      "grad_norm": 0.7679713368415833,
      "learning_rate": 1.23e-05,
      "loss": 0.3228,
      "mean_token_accuracy": 0.9260424096137285,
      "num_tokens": 2746968.0,
      "step": 370
    },
    {
      "entropy": 1.9144606456160544,
      "epoch": 0.0056858085771544584,
      "grad_norm": 0.605851948261261,
      "learning_rate": 1.2633333333333333e-05,
      "loss": 0.3176,
      "mean_token_accuracy": 0.9315214604139328,
      "num_tokens": 2823002.0,
      "step": 380
    },
    {
      "entropy": 1.9592526257038116,
      "epoch": 0.005835435118658523,
      "grad_norm": 0.8147691488265991,
      "learning_rate": 1.2966666666666669e-05,
      "loss": 0.3048,
      "mean_token_accuracy": 0.9300720985978842,
      "num_tokens": 2899342.0,
      "step": 390
    },
    {
      "entropy": 1.9630496338009835,
      "epoch": 0.005985061660162588,
      "grad_norm": 0.6642063856124878,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.306,
      "mean_token_accuracy": 0.9282677911221981,
      "num_tokens": 2975334.0,
      "step": 400
    },
    {
      "entropy": 1.9693477630615235,
      "epoch": 0.006134688201666652,
      "grad_norm": 0.6033242344856262,
      "learning_rate": 1.3633333333333334e-05,
      "loss": 0.3024,
      "mean_token_accuracy": 0.930531233176589,
      "num_tokens": 3049045.0,
      "step": 410
    },
    {
      "entropy": 1.9517102107405662,
      "epoch": 0.006284314743170718,
      "grad_norm": 0.6966942548751831,
      "learning_rate": 1.3966666666666666e-05,
      "loss": 0.3072,
      "mean_token_accuracy": 0.9308092508465051,
      "num_tokens": 3121895.0,
      "step": 420
    },
    {
      "entropy": 1.9387120671570301,
      "epoch": 0.006433941284674782,
      "grad_norm": 0.5533284544944763,
      "learning_rate": 1.43e-05,
      "loss": 0.3175,
      "mean_token_accuracy": 0.9267605431377888,
      "num_tokens": 3197559.0,
      "step": 430
    },
    {
      "entropy": 2.006394574791193,
      "epoch": 0.006583567826178847,
      "grad_norm": 0.5841144919395447,
      "learning_rate": 1.4633333333333334e-05,
      "loss": 0.3451,
      "mean_token_accuracy": 0.9212180498987437,
      "num_tokens": 3267770.0,
      "step": 440
    },
    {
      "entropy": 1.9221089392900468,
      "epoch": 0.0067331943676829115,
      "grad_norm": 0.6260424256324768,
      "learning_rate": 1.4966666666666668e-05,
      "loss": 0.2986,
      "mean_token_accuracy": 0.9328217662870883,
      "num_tokens": 3342336.0,
      "step": 450
    },
    {
      "entropy": 1.9266556471586227,
      "epoch": 0.006882820909186976,
      "grad_norm": 0.6570745706558228,
      "learning_rate": 1.53e-05,
      "loss": 0.3153,
      "mean_token_accuracy": 0.9267579711973667,
      "num_tokens": 3414173.0,
      "step": 460
    },
    {
      "entropy": 1.9416103184223175,
      "epoch": 0.007032447450691041,
      "grad_norm": 0.6072602868080139,
      "learning_rate": 1.563333333333333e-05,
      "loss": 0.305,
      "mean_token_accuracy": 0.9291992306709289,
      "num_tokens": 3486177.0,
      "step": 470
    },
    {
      "entropy": 1.916047702729702,
      "epoch": 0.007182073992195105,
      "grad_norm": 0.645366370677948,
      "learning_rate": 1.5966666666666667e-05,
      "loss": 0.2985,
      "mean_token_accuracy": 0.9295349083840847,
      "num_tokens": 3559387.0,
      "step": 480
    },
    {
      "entropy": 1.9014546275138855,
      "epoch": 0.00733170053369917,
      "grad_norm": 0.7040843367576599,
      "learning_rate": 1.63e-05,
      "loss": 0.2982,
      "mean_token_accuracy": 0.9314001020044088,
      "num_tokens": 3630666.0,
      "step": 490
    },
    {
      "entropy": 1.9023175269365311,
      "epoch": 0.0074813270752032346,
      "grad_norm": 0.8277549147605896,
      "learning_rate": 1.6633333333333336e-05,
      "loss": 0.2889,
      "mean_token_accuracy": 0.9310158092528582,
      "num_tokens": 3707339.0,
      "step": 500
    },
    {
      "entropy": 1.8840498104691505,
      "epoch": 0.0076309536167073,
      "grad_norm": 0.659382700920105,
      "learning_rate": 1.6966666666666668e-05,
      "loss": 0.2708,
      "mean_token_accuracy": 0.9373713478446006,
      "num_tokens": 3782141.0,
      "step": 510
    },
    {
      "entropy": 1.8895556643605231,
      "epoch": 0.007780580158211365,
      "grad_norm": 0.5819388031959534,
      "learning_rate": 1.73e-05,
      "loss": 0.2964,
      "mean_token_accuracy": 0.9296660047024489,
      "num_tokens": 3858279.0,
      "step": 520
    },
    {
      "entropy": 1.9089583836495876,
      "epoch": 0.007930206699715428,
      "grad_norm": 0.6067777276039124,
      "learning_rate": 1.7633333333333336e-05,
      "loss": 0.31,
      "mean_token_accuracy": 0.9262268859893084,
      "num_tokens": 3931542.0,
      "step": 530
    },
    {
      "entropy": 1.9079712502658368,
      "epoch": 0.008079833241219494,
      "grad_norm": 0.8357219099998474,
      "learning_rate": 1.796666666666667e-05,
      "loss": 0.3032,
      "mean_token_accuracy": 0.9321066863834858,
      "num_tokens": 4006252.0,
      "step": 540
    },
    {
      "entropy": 1.8887382842600346,
      "epoch": 0.008229459782723558,
      "grad_norm": 0.7288631796836853,
      "learning_rate": 1.83e-05,
      "loss": 0.2983,
      "mean_token_accuracy": 0.9299008283764124,
      "num_tokens": 4084345.0,
      "step": 550
    },
    {
      "entropy": 1.914674286544323,
      "epoch": 0.008379086324227623,
      "grad_norm": 0.7599960565567017,
      "learning_rate": 1.8633333333333333e-05,
      "loss": 0.2988,
      "mean_token_accuracy": 0.9312361136078835,
      "num_tokens": 4159901.0,
      "step": 560
    },
    {
      "entropy": 1.9340942613780499,
      "epoch": 0.008528712865731689,
      "grad_norm": 0.6785186529159546,
      "learning_rate": 1.896666666666667e-05,
      "loss": 0.3168,
      "mean_token_accuracy": 0.9255207315087318,
      "num_tokens": 4234890.0,
      "step": 570
    },
    {
      "entropy": 1.9239650063216687,
      "epoch": 0.008678339407235752,
      "grad_norm": 0.6722201108932495,
      "learning_rate": 1.93e-05,
      "loss": 0.2878,
      "mean_token_accuracy": 0.9320832658559084,
      "num_tokens": 4309875.0,
      "step": 580
    },
    {
      "entropy": 1.9070725180208683,
      "epoch": 0.008827965948739818,
      "grad_norm": 0.7774500846862793,
      "learning_rate": 1.9633333333333334e-05,
      "loss": 0.3097,
      "mean_token_accuracy": 0.9279659233987332,
      "num_tokens": 4385491.0,
      "step": 590
    },
    {
      "entropy": 1.9078837275505065,
      "epoch": 0.008977592490243881,
      "grad_norm": 0.6776214838027954,
      "learning_rate": 1.9966666666666666e-05,
      "loss": 0.2732,
      "mean_token_accuracy": 0.9350131578743458,
      "num_tokens": 4460481.0,
      "step": 600
    },
    {
      "entropy": 1.9244660943746568,
      "epoch": 0.009127219031747947,
      "grad_norm": 0.6977888941764832,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.298,
      "mean_token_accuracy": 0.9321236774325371,
      "num_tokens": 4533639.0,
      "step": 610
    },
    {
      "entropy": 1.9258983984589577,
      "epoch": 0.00927684557325201,
      "grad_norm": 0.7099884152412415,
      "learning_rate": 2.0633333333333335e-05,
      "loss": 0.2899,
      "mean_token_accuracy": 0.9332225061953068,
      "num_tokens": 4606907.0,
      "step": 620
    },
    {
      "entropy": 1.9157255716621875,
      "epoch": 0.009426472114756076,
      "grad_norm": 0.7240561246871948,
      "learning_rate": 2.0966666666666667e-05,
      "loss": 0.296,
      "mean_token_accuracy": 0.9303838603198529,
      "num_tokens": 4681670.0,
      "step": 630
    },
    {
      "entropy": 1.9390470430254936,
      "epoch": 0.00957609865626014,
      "grad_norm": 0.69577556848526,
      "learning_rate": 2.13e-05,
      "loss": 0.314,
      "mean_token_accuracy": 0.9272349439561367,
      "num_tokens": 4754738.0,
      "step": 640
    },
    {
      "entropy": 1.9329915449023247,
      "epoch": 0.009725725197764205,
      "grad_norm": 0.7179453372955322,
      "learning_rate": 2.1633333333333332e-05,
      "loss": 0.3054,
      "mean_token_accuracy": 0.9324368629604578,
      "num_tokens": 4826393.0,
      "step": 650
    },
    {
      "entropy": 1.9232095345854758,
      "epoch": 0.00987535173926827,
      "grad_norm": 0.6640360355377197,
      "learning_rate": 2.1966666666666668e-05,
      "loss": 0.2988,
      "mean_token_accuracy": 0.9285763081163168,
      "num_tokens": 4902751.0,
      "step": 660
    },
    {
      "entropy": 1.9348515033721925,
      "epoch": 0.010024978280772335,
      "grad_norm": 0.6657624244689941,
      "learning_rate": 2.23e-05,
      "loss": 0.2645,
      "mean_token_accuracy": 0.9398685716092586,
      "num_tokens": 4978287.0,
      "step": 670
    },
    {
      "entropy": 1.8591935105621815,
      "epoch": 0.0101746048222764,
      "grad_norm": 0.5845288038253784,
      "learning_rate": 2.2633333333333336e-05,
      "loss": 0.2815,
      "mean_token_accuracy": 0.9355435144156218,
      "num_tokens": 5052167.0,
      "step": 680
    },
    {
      "entropy": 1.8645537972450257,
      "epoch": 0.010324231363780464,
      "grad_norm": 0.5923750996589661,
      "learning_rate": 2.2966666666666668e-05,
      "loss": 0.2795,
      "mean_token_accuracy": 0.9351237382739782,
      "num_tokens": 5126224.0,
      "step": 690
    },
    {
      "entropy": 1.8773439265787601,
      "epoch": 0.01047385790528453,
      "grad_norm": 0.5365191698074341,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.2791,
      "mean_token_accuracy": 0.9342083893716335,
      "num_tokens": 5199985.0,
      "step": 700
    },
    {
      "entropy": 1.8817263856530189,
      "epoch": 0.010623484446788593,
      "grad_norm": 0.8465101718902588,
      "learning_rate": 2.3633333333333336e-05,
      "loss": 0.2777,
      "mean_token_accuracy": 0.9357718873769045,
      "num_tokens": 5275876.0,
      "step": 710
    },
    {
      "entropy": 1.893296255916357,
      "epoch": 0.010773110988292658,
      "grad_norm": 0.685375988483429,
      "learning_rate": 2.396666666666667e-05,
      "loss": 0.2977,
      "mean_token_accuracy": 0.9300573889166117,
      "num_tokens": 5347997.0,
      "step": 720
    },
    {
      "entropy": 1.92828084602952,
      "epoch": 0.010922737529796722,
      "grad_norm": 0.7465219497680664,
      "learning_rate": 2.43e-05,
      "loss": 0.2966,
      "mean_token_accuracy": 0.9275187090039253,
      "num_tokens": 5418434.0,
      "step": 730
    },
    {
      "entropy": 1.8966885544359684,
      "epoch": 0.011072364071300788,
      "grad_norm": 0.5790083408355713,
      "learning_rate": 2.4633333333333334e-05,
      "loss": 0.2839,
      "mean_token_accuracy": 0.9362265005707741,
      "num_tokens": 5491242.0,
      "step": 740
    },
    {
      "entropy": 1.8974473163485528,
      "epoch": 0.011221990612804853,
      "grad_norm": 0.7603051662445068,
      "learning_rate": 2.496666666666667e-05,
      "loss": 0.2976,
      "mean_token_accuracy": 0.9297301787883043,
      "num_tokens": 5563078.0,
      "step": 750
    },
    {
      "entropy": 1.8984058760106564,
      "epoch": 0.011371617154308917,
      "grad_norm": 0.8949026465415955,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.2788,
      "mean_token_accuracy": 0.9326956555247307,
      "num_tokens": 5638433.0,
      "step": 760
    },
    {
      "entropy": 1.9002277642488479,
      "epoch": 0.011521243695812982,
      "grad_norm": 0.5953108668327332,
      "learning_rate": 2.5633333333333338e-05,
      "loss": 0.2738,
      "mean_token_accuracy": 0.9366820387542247,
      "num_tokens": 5714766.0,
      "step": 770
    },
    {
      "entropy": 1.858785030245781,
      "epoch": 0.011670870237317046,
      "grad_norm": 0.5735443234443665,
      "learning_rate": 2.5966666666666667e-05,
      "loss": 0.2745,
      "mean_token_accuracy": 0.9378697078675031,
      "num_tokens": 5786645.0,
      "step": 780
    },
    {
      "entropy": 1.8972781613469123,
      "epoch": 0.011820496778821112,
      "grad_norm": 0.6718502044677734,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.2905,
      "mean_token_accuracy": 0.9321922946721315,
      "num_tokens": 5860060.0,
      "step": 790
    },
    {
      "entropy": 1.8990380063652992,
      "epoch": 0.011970123320325175,
      "grad_norm": 0.6002764105796814,
      "learning_rate": 2.663333333333333e-05,
      "loss": 0.2633,
      "mean_token_accuracy": 0.9380236633121968,
      "num_tokens": 5931732.0,
      "step": 800
    },
    {
      "entropy": 1.8749303452670574,
      "epoch": 0.01211974986182924,
      "grad_norm": 0.6109471917152405,
      "learning_rate": 2.6966666666666667e-05,
      "loss": 0.2715,
      "mean_token_accuracy": 0.9344659682363272,
      "num_tokens": 6004329.0,
      "step": 810
    },
    {
      "entropy": 1.8789727620780468,
      "epoch": 0.012269376403333304,
      "grad_norm": 0.5260920524597168,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.2706,
      "mean_token_accuracy": 0.9346622083336115,
      "num_tokens": 6078354.0,
      "step": 820
    },
    {
      "entropy": 1.9194171257317065,
      "epoch": 0.01241900294483737,
      "grad_norm": 0.7049269080162048,
      "learning_rate": 2.7633333333333332e-05,
      "loss": 0.2844,
      "mean_token_accuracy": 0.9350420698523522,
      "num_tokens": 6152028.0,
      "step": 830
    },
    {
      "entropy": 1.943284920603037,
      "epoch": 0.012568629486341435,
      "grad_norm": 0.6106187701225281,
      "learning_rate": 2.7966666666666668e-05,
      "loss": 0.25,
      "mean_token_accuracy": 0.9411705233156681,
      "num_tokens": 6224768.0,
      "step": 840
    },
    {
      "entropy": 1.8886298567056656,
      "epoch": 0.0127182560278455,
      "grad_norm": 0.567125678062439,
      "learning_rate": 2.83e-05,
      "loss": 0.2764,
      "mean_token_accuracy": 0.9334474727511406,
      "num_tokens": 6299538.0,
      "step": 850
    },
    {
      "entropy": 1.9227832846343518,
      "epoch": 0.012867882569349565,
      "grad_norm": 0.5880188941955566,
      "learning_rate": 2.8633333333333336e-05,
      "loss": 0.2801,
      "mean_token_accuracy": 0.9350915282964707,
      "num_tokens": 6374973.0,
      "step": 860
    },
    {
      "entropy": 1.9170513726770877,
      "epoch": 0.013017509110853628,
      "grad_norm": 0.6411625146865845,
      "learning_rate": 2.8966666666666668e-05,
      "loss": 0.2692,
      "mean_token_accuracy": 0.9337792567908764,
      "num_tokens": 6449041.0,
      "step": 870
    },
    {
      "entropy": 1.921680285036564,
      "epoch": 0.013167135652357694,
      "grad_norm": 0.6360586881637573,
      "learning_rate": 2.93e-05,
      "loss": 0.2948,
      "mean_token_accuracy": 0.9307741347700358,
      "num_tokens": 6523677.0,
      "step": 880
    },
    {
      "entropy": 1.8694362312555313,
      "epoch": 0.013316762193861758,
      "grad_norm": 0.6214519143104553,
      "learning_rate": 2.9633333333333336e-05,
      "loss": 0.272,
      "mean_token_accuracy": 0.9335201870650053,
      "num_tokens": 6599020.0,
      "step": 890
    },
    {
      "entropy": 1.8842721477150917,
      "epoch": 0.013466388735365823,
      "grad_norm": 0.4771357774734497,
      "learning_rate": 2.9966666666666672e-05,
      "loss": 0.2693,
      "mean_token_accuracy": 0.9341329276561737,
      "num_tokens": 6672193.0,
      "step": 900
    },
    {
      "entropy": 1.8837077029049396,
      "epoch": 0.013616015276869887,
      "grad_norm": 0.5215792655944824,
      "learning_rate": 3.03e-05,
      "loss": 0.2697,
      "mean_token_accuracy": 0.940300104022026,
      "num_tokens": 6747137.0,
      "step": 910
    },
    {
      "entropy": 1.899841669946909,
      "epoch": 0.013765641818373952,
      "grad_norm": 0.5038917064666748,
      "learning_rate": 3.063333333333334e-05,
      "loss": 0.2836,
      "mean_token_accuracy": 0.9341947309672832,
      "num_tokens": 6822933.0,
      "step": 920
    },
    {
      "entropy": 1.8710016667842866,
      "epoch": 0.013915268359878018,
      "grad_norm": 0.5185027718544006,
      "learning_rate": 3.096666666666666e-05,
      "loss": 0.2467,
      "mean_token_accuracy": 0.9394612934440374,
      "num_tokens": 6901078.0,
      "step": 930
    },
    {
      "entropy": 1.8596931979060174,
      "epoch": 0.014064894901382081,
      "grad_norm": 0.4969352185726166,
      "learning_rate": 3.13e-05,
      "loss": 0.2719,
      "mean_token_accuracy": 0.9343028452247382,
      "num_tokens": 6974748.0,
      "step": 940
    },
    {
      "entropy": 1.849874021857977,
      "epoch": 0.014214521442886147,
      "grad_norm": 0.47033780813217163,
      "learning_rate": 3.1633333333333334e-05,
      "loss": 0.2591,
      "mean_token_accuracy": 0.9395352073013783,
      "num_tokens": 7048842.0,
      "step": 950
    },
    {
      "entropy": 1.8826984852552413,
      "epoch": 0.01436414798439021,
      "grad_norm": 0.5043299198150635,
      "learning_rate": 3.196666666666667e-05,
      "loss": 0.3006,
      "mean_token_accuracy": 0.9302482720464468,
      "num_tokens": 7121486.0,
      "step": 960
    },
    {
      "entropy": 1.890236098319292,
      "epoch": 0.014513774525894276,
      "grad_norm": 0.4198809266090393,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.2623,
      "mean_token_accuracy": 0.9404679380357266,
      "num_tokens": 7197273.0,
      "step": 970
    },
    {
      "entropy": 1.894095178693533,
      "epoch": 0.01466340106739834,
      "grad_norm": 0.5990613698959351,
      "learning_rate": 3.263333333333333e-05,
      "loss": 0.2917,
      "mean_token_accuracy": 0.9296809330582618,
      "num_tokens": 7273490.0,
      "step": 980
    },
    {
      "entropy": 1.9271334432065488,
      "epoch": 0.014813027608902405,
      "grad_norm": 0.5427032709121704,
      "learning_rate": 3.296666666666667e-05,
      "loss": 0.2784,
      "mean_token_accuracy": 0.9338910602033138,
      "num_tokens": 7348486.0,
      "step": 990
    },
    {
      "entropy": 1.9327290922403335,
      "epoch": 0.014962654150406469,
      "grad_norm": 0.5623543858528137,
      "learning_rate": 3.33e-05,
      "loss": 0.2783,
      "mean_token_accuracy": 0.9333202257752419,
      "num_tokens": 7425317.0,
      "step": 1000
    },
    {
      "epoch": 0.014962654150406469,
      "eval_entropy": 1.8753016370296478,
      "eval_loss": 0.3028193414211273,
      "eval_mean_token_accuracy": 0.9350398851037025,
      "eval_num_tokens": 7425317.0,
      "eval_runtime": 550.7977,
      "eval_samples_per_second": 36.311,
      "eval_steps_per_second": 9.078,
      "step": 1000
    },
    {
      "entropy": 1.8598886378109456,
      "epoch": 0.015112280691910535,
      "grad_norm": 0.5623306035995483,
      "learning_rate": 3.3633333333333335e-05,
      "loss": 0.2698,
      "mean_token_accuracy": 0.9328101098537445,
      "num_tokens": 7497257.0,
      "step": 1010
    },
    {
      "entropy": 1.9073072776198388,
      "epoch": 0.0152619072334146,
      "grad_norm": 0.5176863074302673,
      "learning_rate": 3.396666666666667e-05,
      "loss": 0.2647,
      "mean_token_accuracy": 0.9367997858673334,
      "num_tokens": 7570454.0,
      "step": 1020
    },
    {
      "entropy": 1.8935526065528392,
      "epoch": 0.015411533774918664,
      "grad_norm": 0.5857236385345459,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9381060626357793,
      "num_tokens": 7646111.0,
      "step": 1030
    },
    {
      "entropy": 1.8718897640705108,
      "epoch": 0.01556116031642273,
      "grad_norm": 0.5254436731338501,
      "learning_rate": 3.463333333333333e-05,
      "loss": 0.2672,
      "mean_token_accuracy": 0.9377317324280738,
      "num_tokens": 7721247.0,
      "step": 1040
    },
    {
      "entropy": 1.8877231419086455,
      "epoch": 0.015710786857926793,
      "grad_norm": 0.5419355034828186,
      "learning_rate": 3.496666666666667e-05,
      "loss": 0.2689,
      "mean_token_accuracy": 0.9384996693581342,
      "num_tokens": 7797010.0,
      "step": 1050
    },
    {
      "entropy": 1.8906154781579971,
      "epoch": 0.015860413399430857,
      "grad_norm": 0.5910248160362244,
      "learning_rate": 3.53e-05,
      "loss": 0.2948,
      "mean_token_accuracy": 0.9314150124788284,
      "num_tokens": 7867342.0,
      "step": 1060
    },
    {
      "entropy": 1.8790299490094184,
      "epoch": 0.016010039940934924,
      "grad_norm": 0.6653301119804382,
      "learning_rate": 3.563333333333334e-05,
      "loss": 0.2721,
      "mean_token_accuracy": 0.9363476645201445,
      "num_tokens": 7940511.0,
      "step": 1070
    },
    {
      "entropy": 1.871644416451454,
      "epoch": 0.016159666482438988,
      "grad_norm": 0.6168724298477173,
      "learning_rate": 3.596666666666667e-05,
      "loss": 0.2844,
      "mean_token_accuracy": 0.9328656613826751,
      "num_tokens": 8013723.0,
      "step": 1080
    },
    {
      "entropy": 1.8985835947096348,
      "epoch": 0.01630929302394305,
      "grad_norm": 0.5236951112747192,
      "learning_rate": 3.63e-05,
      "loss": 0.2632,
      "mean_token_accuracy": 0.9363023322075605,
      "num_tokens": 8090956.0,
      "step": 1090
    },
    {
      "entropy": 1.9234493792057037,
      "epoch": 0.016458919565447115,
      "grad_norm": 0.4238961935043335,
      "learning_rate": 3.6633333333333334e-05,
      "loss": 0.2727,
      "mean_token_accuracy": 0.9339867666363716,
      "num_tokens": 8167253.0,
      "step": 1100
    },
    {
      "entropy": 1.902877391129732,
      "epoch": 0.016608546106951182,
      "grad_norm": 0.4996739327907562,
      "learning_rate": 3.6966666666666666e-05,
      "loss": 0.3087,
      "mean_token_accuracy": 0.9344214573502541,
      "num_tokens": 8242602.0,
      "step": 1110
    },
    {
      "entropy": 1.925073929131031,
      "epoch": 0.016758172648455246,
      "grad_norm": 0.6041417717933655,
      "learning_rate": 3.73e-05,
      "loss": 0.2968,
      "mean_token_accuracy": 0.9308765295892953,
      "num_tokens": 8317125.0,
      "step": 1120
    },
    {
      "entropy": 1.9387078195810319,
      "epoch": 0.01690779918995931,
      "grad_norm": 0.5066329836845398,
      "learning_rate": 3.763333333333334e-05,
      "loss": 0.2736,
      "mean_token_accuracy": 0.9361588280647993,
      "num_tokens": 8390889.0,
      "step": 1130
    },
    {
      "entropy": 1.9376711264252662,
      "epoch": 0.017057425731463377,
      "grad_norm": 0.3998706340789795,
      "learning_rate": 3.796666666666667e-05,
      "loss": 0.2647,
      "mean_token_accuracy": 0.9393856279551983,
      "num_tokens": 8464229.0,
      "step": 1140
    },
    {
      "entropy": 1.9371434718370437,
      "epoch": 0.01720705227296744,
      "grad_norm": 0.4436974823474884,
      "learning_rate": 3.83e-05,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9368218403309584,
      "num_tokens": 8539060.0,
      "step": 1150
    },
    {
      "entropy": 1.9364991538226604,
      "epoch": 0.017356678814471505,
      "grad_norm": 0.4545430839061737,
      "learning_rate": 3.8633333333333335e-05,
      "loss": 0.2669,
      "mean_token_accuracy": 0.9374832130968571,
      "num_tokens": 8615520.0,
      "step": 1160
    },
    {
      "entropy": 1.9852855257689952,
      "epoch": 0.017506305355975568,
      "grad_norm": 0.6584464311599731,
      "learning_rate": 3.896666666666667e-05,
      "loss": 0.289,
      "mean_token_accuracy": 0.9299779187887907,
      "num_tokens": 8687986.0,
      "step": 1170
    },
    {
      "entropy": 1.9448260597884655,
      "epoch": 0.017655931897479635,
      "grad_norm": 0.35713228583335876,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.2565,
      "mean_token_accuracy": 0.9404401514679194,
      "num_tokens": 8765081.0,
      "step": 1180
    },
    {
      "entropy": 1.9684312880039214,
      "epoch": 0.0178055584389837,
      "grad_norm": 0.6406857967376709,
      "learning_rate": 3.963333333333333e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9406179431825876,
      "num_tokens": 8842240.0,
      "step": 1190
    },
    {
      "entropy": 1.9455515153706073,
      "epoch": 0.017955184980487763,
      "grad_norm": 0.5930728912353516,
      "learning_rate": 3.996666666666667e-05,
      "loss": 0.2605,
      "mean_token_accuracy": 0.9413910396397114,
      "num_tokens": 8915359.0,
      "step": 1200
    },
    {
      "entropy": 1.966473037749529,
      "epoch": 0.01810481152199183,
      "grad_norm": 0.47765591740608215,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.2746,
      "mean_token_accuracy": 0.9357528451830148,
      "num_tokens": 8989678.0,
      "step": 1210
    },
    {
      "entropy": 1.9709252327680589,
      "epoch": 0.018254438063495894,
      "grad_norm": 0.4226253032684326,
      "learning_rate": 4.0633333333333336e-05,
      "loss": 0.2676,
      "mean_token_accuracy": 0.9349496249109507,
      "num_tokens": 9063437.0,
      "step": 1220
    },
    {
      "entropy": 1.9574174158275128,
      "epoch": 0.018404064604999958,
      "grad_norm": 0.49076026678085327,
      "learning_rate": 4.096666666666667e-05,
      "loss": 0.2749,
      "mean_token_accuracy": 0.9361681073904038,
      "num_tokens": 9136033.0,
      "step": 1230
    },
    {
      "entropy": 1.9436740562319756,
      "epoch": 0.01855369114650402,
      "grad_norm": 0.42508575320243835,
      "learning_rate": 4.13e-05,
      "loss": 0.2722,
      "mean_token_accuracy": 0.9371613308787345,
      "num_tokens": 9210323.0,
      "step": 1240
    },
    {
      "entropy": 1.9515752837061882,
      "epoch": 0.01870331768800809,
      "grad_norm": 0.5182852745056152,
      "learning_rate": 4.1633333333333333e-05,
      "loss": 0.2765,
      "mean_token_accuracy": 0.9368175711482764,
      "num_tokens": 9283468.0,
      "step": 1250
    },
    {
      "entropy": 1.9461850970983505,
      "epoch": 0.018852944229512152,
      "grad_norm": 0.5053494572639465,
      "learning_rate": 4.196666666666667e-05,
      "loss": 0.3014,
      "mean_token_accuracy": 0.928177010267973,
      "num_tokens": 9357991.0,
      "step": 1260
    },
    {
      "entropy": 1.9470947682857513,
      "epoch": 0.019002570771016216,
      "grad_norm": 0.45178303122520447,
      "learning_rate": 4.23e-05,
      "loss": 0.2647,
      "mean_token_accuracy": 0.937558152154088,
      "num_tokens": 9433852.0,
      "step": 1270
    },
    {
      "entropy": 1.9238952748477458,
      "epoch": 0.01915219731252028,
      "grad_norm": 0.4230474531650543,
      "learning_rate": 4.263333333333334e-05,
      "loss": 0.2768,
      "mean_token_accuracy": 0.9341432381421327,
      "num_tokens": 9507989.0,
      "step": 1280
    },
    {
      "entropy": 1.9638021126389504,
      "epoch": 0.019301823854024347,
      "grad_norm": 0.45698633790016174,
      "learning_rate": 4.296666666666666e-05,
      "loss": 0.2671,
      "mean_token_accuracy": 0.9356396466493606,
      "num_tokens": 9583157.0,
      "step": 1290
    },
    {
      "entropy": 1.9190260156989098,
      "epoch": 0.01945145039552841,
      "grad_norm": 0.4040304124355316,
      "learning_rate": 4.33e-05,
      "loss": 0.267,
      "mean_token_accuracy": 0.9394377671182156,
      "num_tokens": 9656687.0,
      "step": 1300
    },
    {
      "entropy": 1.950559175759554,
      "epoch": 0.019601076937032474,
      "grad_norm": 0.3600620925426483,
      "learning_rate": 4.3633333333333335e-05,
      "loss": 0.2565,
      "mean_token_accuracy": 0.9395214602351188,
      "num_tokens": 9732763.0,
      "step": 1310
    },
    {
      "entropy": 1.9855354383587838,
      "epoch": 0.01975070347853654,
      "grad_norm": 0.5064100623130798,
      "learning_rate": 4.396666666666667e-05,
      "loss": 0.2814,
      "mean_token_accuracy": 0.9348355196416378,
      "num_tokens": 9807788.0,
      "step": 1320
    },
    {
      "entropy": 1.9128718122839927,
      "epoch": 0.019900330020040605,
      "grad_norm": 0.4409477710723877,
      "learning_rate": 4.43e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.9403836719691754,
      "num_tokens": 9882836.0,
      "step": 1330
    },
    {
      "entropy": 1.8933450989425182,
      "epoch": 0.02004995656154467,
      "grad_norm": 0.4175204932689667,
      "learning_rate": 4.463333333333334e-05,
      "loss": 0.2544,
      "mean_token_accuracy": 0.9388504240661859,
      "num_tokens": 9956281.0,
      "step": 1340
    },
    {
      "entropy": 1.9070032380521298,
      "epoch": 0.020199583103048733,
      "grad_norm": 0.4159417152404785,
      "learning_rate": 4.496666666666667e-05,
      "loss": 0.2677,
      "mean_token_accuracy": 0.9361295677721501,
      "num_tokens": 10029334.0,
      "step": 1350
    },
    {
      "entropy": 1.9101487830281259,
      "epoch": 0.0203492096445528,
      "grad_norm": 0.46401578187942505,
      "learning_rate": 4.53e-05,
      "loss": 0.2866,
      "mean_token_accuracy": 0.9300867468118668,
      "num_tokens": 10103058.0,
      "step": 1360
    },
    {
      "entropy": 1.8810361474752426,
      "epoch": 0.020498836186056864,
      "grad_norm": 0.46862661838531494,
      "learning_rate": 4.5633333333333336e-05,
      "loss": 0.2935,
      "mean_token_accuracy": 0.9306722909212113,
      "num_tokens": 10176275.0,
      "step": 1370
    },
    {
      "entropy": 1.9244099050760268,
      "epoch": 0.020648462727560928,
      "grad_norm": 0.47564825415611267,
      "learning_rate": 4.596666666666667e-05,
      "loss": 0.29,
      "mean_token_accuracy": 0.9291265770792961,
      "num_tokens": 10247137.0,
      "step": 1380
    },
    {
      "entropy": 1.9467399790883064,
      "epoch": 0.020798089269064995,
      "grad_norm": 0.4551142454147339,
      "learning_rate": 4.630000000000001e-05,
      "loss": 0.2903,
      "mean_token_accuracy": 0.9279755663126708,
      "num_tokens": 10320274.0,
      "step": 1390
    },
    {
      "entropy": 1.9126875169575215,
      "epoch": 0.02094771581056906,
      "grad_norm": 0.3794918656349182,
      "learning_rate": 4.663333333333333e-05,
      "loss": 0.2652,
      "mean_token_accuracy": 0.9367523733526468,
      "num_tokens": 10396195.0,
      "step": 1400
    },
    {
      "entropy": 1.9332519829273225,
      "epoch": 0.021097342352073122,
      "grad_norm": 0.4127694368362427,
      "learning_rate": 4.696666666666667e-05,
      "loss": 0.2725,
      "mean_token_accuracy": 0.9344959072768688,
      "num_tokens": 10471313.0,
      "step": 1410
    },
    {
      "entropy": 1.9278608597815037,
      "epoch": 0.021246968893577186,
      "grad_norm": 0.42189526557922363,
      "learning_rate": 4.73e-05,
      "loss": 0.2592,
      "mean_token_accuracy": 0.9370869193226099,
      "num_tokens": 10548164.0,
      "step": 1420
    },
    {
      "entropy": 1.9116816632449627,
      "epoch": 0.021396595435081253,
      "grad_norm": 0.4197300672531128,
      "learning_rate": 4.763333333333334e-05,
      "loss": 0.2717,
      "mean_token_accuracy": 0.9355357114225626,
      "num_tokens": 10624765.0,
      "step": 1430
    },
    {
      "entropy": 1.8857680186629295,
      "epoch": 0.021546221976585317,
      "grad_norm": 0.40343478322029114,
      "learning_rate": 4.796666666666667e-05,
      "loss": 0.2663,
      "mean_token_accuracy": 0.9368946634232997,
      "num_tokens": 10698617.0,
      "step": 1440
    },
    {
      "entropy": 1.9078956685960293,
      "epoch": 0.02169584851808938,
      "grad_norm": 0.44704073667526245,
      "learning_rate": 4.83e-05,
      "loss": 0.2718,
      "mean_token_accuracy": 0.9337268963456153,
      "num_tokens": 10768740.0,
      "step": 1450
    },
    {
      "entropy": 1.919688557833433,
      "epoch": 0.021845475059593444,
      "grad_norm": 0.499002605676651,
      "learning_rate": 4.8633333333333334e-05,
      "loss": 0.2752,
      "mean_token_accuracy": 0.935129937902093,
      "num_tokens": 10840930.0,
      "step": 1460
    },
    {
      "entropy": 1.9083768844604492,
      "epoch": 0.02199510160109751,
      "grad_norm": 0.4030619263648987,
      "learning_rate": 4.8966666666666667e-05,
      "loss": 0.2678,
      "mean_token_accuracy": 0.9360015477985144,
      "num_tokens": 10914072.0,
      "step": 1470
    },
    {
      "entropy": 1.9401203468441963,
      "epoch": 0.022144728142601575,
      "grad_norm": 0.43506520986557007,
      "learning_rate": 4.93e-05,
      "loss": 0.2649,
      "mean_token_accuracy": 0.9385438848286867,
      "num_tokens": 10987332.0,
      "step": 1480
    },
    {
      "entropy": 1.950416960567236,
      "epoch": 0.02229435468410564,
      "grad_norm": 0.37713760137557983,
      "learning_rate": 4.963333333333334e-05,
      "loss": 0.2818,
      "mean_token_accuracy": 0.9327928606420756,
      "num_tokens": 11058921.0,
      "step": 1490
    },
    {
      "entropy": 1.9302063561975955,
      "epoch": 0.022443981225609706,
      "grad_norm": 0.4021291434764862,
      "learning_rate": 4.996666666666667e-05,
      "loss": 0.261,
      "mean_token_accuracy": 0.9362004119902849,
      "num_tokens": 11131444.0,
      "step": 1500
    },
    {
      "entropy": 1.9218756295740604,
      "epoch": 0.02259360776711377,
      "grad_norm": 0.38073208928108215,
      "learning_rate": 4.9999997663602054e-05,
      "loss": 0.281,
      "mean_token_accuracy": 0.9332431003451347,
      "num_tokens": 11205193.0,
      "step": 1510
    },
    {
      "entropy": 1.944690889120102,
      "epoch": 0.022743234308617834,
      "grad_norm": 0.3809882700443268,
      "learning_rate": 4.999998958716528e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9424938000738621,
      "num_tokens": 11281679.0,
      "step": 1520
    },
    {
      "entropy": 1.897782201319933,
      "epoch": 0.022892860850121897,
      "grad_norm": 0.3762741982936859,
      "learning_rate": 4.999997574184713e-05,
      "loss": 0.2517,
      "mean_token_accuracy": 0.940287358313799,
      "num_tokens": 11355390.0,
      "step": 1530
    },
    {
      "entropy": 1.916746361553669,
      "epoch": 0.023042487391625965,
      "grad_norm": 0.4179477095603943,
      "learning_rate": 4.999995612765078e-05,
      "loss": 0.2601,
      "mean_token_accuracy": 0.9393490124493837,
      "num_tokens": 11428180.0,
      "step": 1540
    },
    {
      "entropy": 1.8912997141480445,
      "epoch": 0.02319211393313003,
      "grad_norm": 0.5718160271644592,
      "learning_rate": 4.999993074458079e-05,
      "loss": 0.2835,
      "mean_token_accuracy": 0.9320059884339571,
      "num_tokens": 11501459.0,
      "step": 1550
    },
    {
      "entropy": 1.920396737009287,
      "epoch": 0.023341740474634092,
      "grad_norm": 0.3995632231235504,
      "learning_rate": 4.9999899592643027e-05,
      "loss": 0.2581,
      "mean_token_accuracy": 0.9398284446448087,
      "num_tokens": 11575585.0,
      "step": 1560
    },
    {
      "entropy": 1.9146107286214828,
      "epoch": 0.02349136701613816,
      "grad_norm": 0.40347373485565186,
      "learning_rate": 4.9999862671844685e-05,
      "loss": 0.2494,
      "mean_token_accuracy": 0.9398542977869511,
      "num_tokens": 11650607.0,
      "step": 1570
    },
    {
      "entropy": 1.9214057572185994,
      "epoch": 0.023640993557642223,
      "grad_norm": 0.35241708159446716,
      "learning_rate": 4.999981998219429e-05,
      "loss": 0.248,
      "mean_token_accuracy": 0.9402223721146583,
      "num_tokens": 11724299.0,
      "step": 1580
    },
    {
      "entropy": 1.928932322561741,
      "epoch": 0.023790620099146287,
      "grad_norm": 0.4366796910762787,
      "learning_rate": 4.999977152370175e-05,
      "loss": 0.2676,
      "mean_token_accuracy": 0.9381843123584985,
      "num_tokens": 11799196.0,
      "step": 1590
    },
    {
      "entropy": 1.8909890569746495,
      "epoch": 0.02394024664065035,
      "grad_norm": 0.4896111786365509,
      "learning_rate": 4.999971729637824e-05,
      "loss": 0.2646,
      "mean_token_accuracy": 0.9363083928823471,
      "num_tokens": 11873266.0,
      "step": 1600
    },
    {
      "entropy": 1.8947751440107823,
      "epoch": 0.024089873182154418,
      "grad_norm": 0.48303547501564026,
      "learning_rate": 4.99996573002363e-05,
      "loss": 0.2642,
      "mean_token_accuracy": 0.9363098442554474,
      "num_tokens": 11946112.0,
      "step": 1610
    },
    {
      "entropy": 1.8935870833694934,
      "epoch": 0.02423949972365848,
      "grad_norm": 0.4124827980995178,
      "learning_rate": 4.9999591535289807e-05,
      "loss": 0.2711,
      "mean_token_accuracy": 0.9327072910964489,
      "num_tokens": 12017123.0,
      "step": 1620
    },
    {
      "entropy": 1.9289609640836716,
      "epoch": 0.024389126265162545,
      "grad_norm": 0.2973644733428955,
      "learning_rate": 4.999952000155397e-05,
      "loss": 0.2586,
      "mean_token_accuracy": 0.936373620480299,
      "num_tokens": 12095745.0,
      "step": 1630
    },
    {
      "entropy": 1.9134388372302056,
      "epoch": 0.02453875280666661,
      "grad_norm": 0.38001781702041626,
      "learning_rate": 4.999944269904532e-05,
      "loss": 0.2439,
      "mean_token_accuracy": 0.9400587182492017,
      "num_tokens": 12170842.0,
      "step": 1640
    },
    {
      "entropy": 1.8777817323803903,
      "epoch": 0.024688379348170676,
      "grad_norm": 0.3203199803829193,
      "learning_rate": 4.9999359627781744e-05,
      "loss": 0.2462,
      "mean_token_accuracy": 0.9418114703148603,
      "num_tokens": 12248060.0,
      "step": 1650
    },
    {
      "entropy": 1.8883257977664472,
      "epoch": 0.02483800588967474,
      "grad_norm": 0.4211905300617218,
      "learning_rate": 4.999927078778244e-05,
      "loss": 0.2753,
      "mean_token_accuracy": 0.9372297335416079,
      "num_tokens": 12322829.0,
      "step": 1660
    },
    {
      "entropy": 1.836207826435566,
      "epoch": 0.024987632431178804,
      "grad_norm": 0.41157951951026917,
      "learning_rate": 4.999917617906796e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9439395423978567,
      "num_tokens": 12399833.0,
      "step": 1670
    },
    {
      "entropy": 1.8546844005584717,
      "epoch": 0.02513725897268287,
      "grad_norm": 0.49052169919013977,
      "learning_rate": 4.9999075801660156e-05,
      "loss": 0.2972,
      "mean_token_accuracy": 0.9316247932612896,
      "num_tokens": 12474295.0,
      "step": 1680
    },
    {
      "entropy": 1.870160911977291,
      "epoch": 0.025286885514186935,
      "grad_norm": 0.4229196310043335,
      "learning_rate": 4.999896965558225e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.9369913544505835,
      "num_tokens": 12548840.0,
      "step": 1690
    },
    {
      "entropy": 1.865529114753008,
      "epoch": 0.025436512055691,
      "grad_norm": 0.3937903642654419,
      "learning_rate": 4.99988577408588e-05,
      "loss": 0.2514,
      "mean_token_accuracy": 0.9392479062080383,
      "num_tokens": 12626807.0,
      "step": 1700
    },
    {
      "entropy": 1.9266904883086682,
      "epoch": 0.025586138597195062,
      "grad_norm": 0.37837767601013184,
      "learning_rate": 4.999874005751567e-05,
      "loss": 0.2454,
      "mean_token_accuracy": 0.9401845403015614,
      "num_tokens": 12701250.0,
      "step": 1710
    },
    {
      "entropy": 1.9229820005595684,
      "epoch": 0.02573576513869913,
      "grad_norm": 0.38864415884017944,
      "learning_rate": 4.9998616605580056e-05,
      "loss": 0.2529,
      "mean_token_accuracy": 0.9398800730705261,
      "num_tokens": 12774216.0,
      "step": 1720
    },
    {
      "entropy": 1.9213424824178218,
      "epoch": 0.025885391680203193,
      "grad_norm": 0.44867226481437683,
      "learning_rate": 4.999848738508052e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9368224963545799,
      "num_tokens": 12849238.0,
      "step": 1730
    },
    {
      "entropy": 1.9184985421597958,
      "epoch": 0.026035018221707257,
      "grad_norm": 0.5166764259338379,
      "learning_rate": 4.999835239604694e-05,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9378155428916216,
      "num_tokens": 12926637.0,
      "step": 1740
    },
    {
      "entropy": 1.9088952593505382,
      "epoch": 0.02618464476321132,
      "grad_norm": 0.48186591267585754,
      "learning_rate": 4.999821163851053e-05,
      "loss": 0.2608,
      "mean_token_accuracy": 0.9352299775928259,
      "num_tokens": 13002156.0,
      "step": 1750
    },
    {
      "entropy": 1.9066528983414173,
      "epoch": 0.026334271304715388,
      "grad_norm": 0.4407900869846344,
      "learning_rate": 4.999806511250382e-05,
      "loss": 0.2745,
      "mean_token_accuracy": 0.9300562132149934,
      "num_tokens": 13075930.0,
      "step": 1760
    },
    {
      "entropy": 1.8909938417375087,
      "epoch": 0.02648389784621945,
      "grad_norm": 0.29750967025756836,
      "learning_rate": 4.999791281806069e-05,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9349593546241521,
      "num_tokens": 13148434.0,
      "step": 1770
    },
    {
      "entropy": 1.9058317221701144,
      "epoch": 0.026633524387723515,
      "grad_norm": 0.42107805609703064,
      "learning_rate": 4.9997754755216383e-05,
      "loss": 0.2689,
      "mean_token_accuracy": 0.935636480897665,
      "num_tokens": 13219905.0,
      "step": 1780
    },
    {
      "entropy": 1.8798053488135338,
      "epoch": 0.026783150929227582,
      "grad_norm": 0.4711362421512604,
      "learning_rate": 4.999759092400741e-05,
      "loss": 0.2549,
      "mean_token_accuracy": 0.938982268422842,
      "num_tokens": 13298194.0,
      "step": 1790
    },
    {
      "entropy": 1.874748171120882,
      "epoch": 0.026932777470731646,
      "grad_norm": 0.3945498764514923,
      "learning_rate": 4.999742132447168e-05,
      "loss": 0.2414,
      "mean_token_accuracy": 0.9416402768343687,
      "num_tokens": 13373833.0,
      "step": 1800
    },
    {
      "entropy": 1.88096881210804,
      "epoch": 0.02708240401223571,
      "grad_norm": 0.5215915441513062,
      "learning_rate": 4.999724595664838e-05,
      "loss": 0.2645,
      "mean_token_accuracy": 0.9364828310906887,
      "num_tokens": 13450606.0,
      "step": 1810
    },
    {
      "entropy": 1.8734424695372582,
      "epoch": 0.027232030553739774,
      "grad_norm": 0.34603357315063477,
      "learning_rate": 4.9997064820578087e-05,
      "loss": 0.27,
      "mean_token_accuracy": 0.9391911398619414,
      "num_tokens": 13524458.0,
      "step": 1820
    },
    {
      "entropy": 1.9060774892568588,
      "epoch": 0.02738165709524384,
      "grad_norm": 0.4155914783477783,
      "learning_rate": 4.999687791630266e-05,
      "loss": 0.2761,
      "mean_token_accuracy": 0.9316956490278244,
      "num_tokens": 13596633.0,
      "step": 1830
    },
    {
      "entropy": 1.9168854914605618,
      "epoch": 0.027531283636747905,
      "grad_norm": 0.32368355989456177,
      "learning_rate": 4.9996685243865326e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9384739082306623,
      "num_tokens": 13674632.0,
      "step": 1840
    },
    {
      "entropy": 1.9248627573251724,
      "epoch": 0.02768091017825197,
      "grad_norm": 0.38061970472335815,
      "learning_rate": 4.999648680331063e-05,
      "loss": 0.2484,
      "mean_token_accuracy": 0.9409559514373541,
      "num_tokens": 13750636.0,
      "step": 1850
    },
    {
      "entropy": 1.9219365417957306,
      "epoch": 0.027830536719756035,
      "grad_norm": 0.4516047537326813,
      "learning_rate": 4.999628259468445e-05,
      "loss": 0.2716,
      "mean_token_accuracy": 0.934255688264966,
      "num_tokens": 13823614.0,
      "step": 1860
    },
    {
      "entropy": 1.9693270392715931,
      "epoch": 0.0279801632612601,
      "grad_norm": 0.33230048418045044,
      "learning_rate": 4.9996072618034016e-05,
      "loss": 0.2763,
      "mean_token_accuracy": 0.9362888220697642,
      "num_tokens": 13893257.0,
      "step": 1870
    },
    {
      "entropy": 1.9107227720320226,
      "epoch": 0.028129789802764163,
      "grad_norm": 0.42593950033187866,
      "learning_rate": 4.999585687340787e-05,
      "loss": 0.28,
      "mean_token_accuracy": 0.9359276928007603,
      "num_tokens": 13964930.0,
      "step": 1880
    },
    {
      "entropy": 1.8682038851082325,
      "epoch": 0.028279416344268227,
      "grad_norm": 0.40382304787635803,
      "learning_rate": 4.9995635360855906e-05,
      "loss": 0.2706,
      "mean_token_accuracy": 0.9352442506700754,
      "num_tokens": 14040267.0,
      "step": 1890
    },
    {
      "entropy": 1.8770518898963928,
      "epoch": 0.028429042885772294,
      "grad_norm": 0.43403658270835876,
      "learning_rate": 4.999540808042931e-05,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9348610509186983,
      "num_tokens": 14112330.0,
      "step": 1900
    },
    {
      "entropy": 1.860290214419365,
      "epoch": 0.028578669427276358,
      "grad_norm": 0.3906209468841553,
      "learning_rate": 4.9995175032180676e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9474035806953907,
      "num_tokens": 14188107.0,
      "step": 1910
    },
    {
      "entropy": 1.8999541856348514,
      "epoch": 0.02872829596878042,
      "grad_norm": 0.37739259004592896,
      "learning_rate": 4.999493621616386e-05,
      "loss": 0.2616,
      "mean_token_accuracy": 0.9391215428709984,
      "num_tokens": 14262520.0,
      "step": 1920
    },
    {
      "entropy": 1.9088940359652042,
      "epoch": 0.028877922510284485,
      "grad_norm": 0.5249622464179993,
      "learning_rate": 4.9994691632434094e-05,
      "loss": 0.2817,
      "mean_token_accuracy": 0.9312744207680226,
      "num_tokens": 14335958.0,
      "step": 1930
    },
    {
      "entropy": 1.877857106924057,
      "epoch": 0.029027549051788552,
      "grad_norm": 0.36370429396629333,
      "learning_rate": 4.999444128104792e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9446245223283768,
      "num_tokens": 14408712.0,
      "step": 1940
    },
    {
      "entropy": 1.9114902280271053,
      "epoch": 0.029177175593292616,
      "grad_norm": 0.39494791626930237,
      "learning_rate": 4.9994185162063236e-05,
      "loss": 0.2512,
      "mean_token_accuracy": 0.939300874993205,
      "num_tokens": 14484152.0,
      "step": 1950
    },
    {
      "entropy": 1.908535370975733,
      "epoch": 0.02932680213479668,
      "grad_norm": 0.396015465259552,
      "learning_rate": 4.9993923275539254e-05,
      "loss": 0.2606,
      "mean_token_accuracy": 0.9388527695089579,
      "num_tokens": 14555223.0,
      "step": 1960
    },
    {
      "entropy": 1.888861994445324,
      "epoch": 0.029476428676300747,
      "grad_norm": 0.3579181432723999,
      "learning_rate": 4.9993655621536526e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9433815009891987,
      "num_tokens": 14630614.0,
      "step": 1970
    },
    {
      "entropy": 1.901060538738966,
      "epoch": 0.02962605521780481,
      "grad_norm": 0.378630667924881,
      "learning_rate": 4.999338220011694e-05,
      "loss": 0.2542,
      "mean_token_accuracy": 0.9410500589758157,
      "num_tokens": 14704309.0,
      "step": 1980
    },
    {
      "entropy": 1.9252351596951485,
      "epoch": 0.029775681759308874,
      "grad_norm": 0.3895276188850403,
      "learning_rate": 4.999310301134372e-05,
      "loss": 0.274,
      "mean_token_accuracy": 0.9360985320061446,
      "num_tokens": 14778498.0,
      "step": 1990
    },
    {
      "entropy": 1.9307069145143032,
      "epoch": 0.029925308300812938,
      "grad_norm": 0.383194237947464,
      "learning_rate": 4.9992818055281405e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9400186944752932,
      "num_tokens": 14854754.0,
      "step": 2000
    },
    {
      "epoch": 0.029925308300812938,
      "eval_entropy": 1.9322413902759552,
      "eval_loss": 0.2821834087371826,
      "eval_mean_token_accuracy": 0.9388002062678337,
      "eval_num_tokens": 14854754.0,
      "eval_runtime": 555.3915,
      "eval_samples_per_second": 36.011,
      "eval_steps_per_second": 9.003,
      "step": 2000
    },
    {
      "entropy": 1.9565280303359032,
      "epoch": 0.030074934842317005,
      "grad_norm": 0.41336962580680847,
      "learning_rate": 4.99925273319959e-05,
      "loss": 0.2684,
      "mean_token_accuracy": 0.937010831385851,
      "num_tokens": 14926706.0,
      "step": 2010
    },
    {
      "entropy": 1.9696230962872505,
      "epoch": 0.03022456138382107,
      "grad_norm": 0.3417767882347107,
      "learning_rate": 4.9992230841554414e-05,
      "loss": 0.2622,
      "mean_token_accuracy": 0.9378949467092752,
      "num_tokens": 15000508.0,
      "step": 2020
    },
    {
      "entropy": 1.9325991325080394,
      "epoch": 0.030374187925325133,
      "grad_norm": 0.424651563167572,
      "learning_rate": 4.9991928584025505e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9400513842701912,
      "num_tokens": 15074635.0,
      "step": 2030
    },
    {
      "entropy": 1.9370036140084266,
      "epoch": 0.0305238144668292,
      "grad_norm": 0.2983887791633606,
      "learning_rate": 4.999162055947906e-05,
      "loss": 0.2483,
      "mean_token_accuracy": 0.9417125564068556,
      "num_tokens": 15148762.0,
      "step": 2040
    },
    {
      "entropy": 1.917696239054203,
      "epoch": 0.030673441008333264,
      "grad_norm": 0.4128051996231079,
      "learning_rate": 4.9991306767986294e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9413021482527256,
      "num_tokens": 15223251.0,
      "step": 2050
    },
    {
      "entropy": 1.9703282028436662,
      "epoch": 0.030823067549837328,
      "grad_norm": 0.4117163121700287,
      "learning_rate": 4.999098720961977e-05,
      "loss": 0.2646,
      "mean_token_accuracy": 0.9360530350357295,
      "num_tokens": 15293635.0,
      "step": 2060
    },
    {
      "entropy": 1.9402942091226578,
      "epoch": 0.03097269409134139,
      "grad_norm": 0.3817594349384308,
      "learning_rate": 4.9990661884453374e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9421465035527945,
      "num_tokens": 15367325.0,
      "step": 2070
    },
    {
      "entropy": 1.9466923214495182,
      "epoch": 0.03112232063284546,
      "grad_norm": 0.36239781975746155,
      "learning_rate": 4.999033079256232e-05,
      "loss": 0.2651,
      "mean_token_accuracy": 0.9382477596402168,
      "num_tokens": 15442895.0,
      "step": 2080
    },
    {
      "entropy": 1.9442710898816586,
      "epoch": 0.03127194717434952,
      "grad_norm": 0.34867575764656067,
      "learning_rate": 4.9989993934023185e-05,
      "loss": 0.255,
      "mean_token_accuracy": 0.9413390252739191,
      "num_tokens": 15515950.0,
      "step": 2090
    },
    {
      "entropy": 1.9898070678114892,
      "epoch": 0.031421573715853586,
      "grad_norm": 0.35317516326904297,
      "learning_rate": 4.9989651308913817e-05,
      "loss": 0.2744,
      "mean_token_accuracy": 0.9344735324382782,
      "num_tokens": 15590687.0,
      "step": 2100
    },
    {
      "entropy": 1.952912725508213,
      "epoch": 0.03157120025735765,
      "grad_norm": 0.48686540126800537,
      "learning_rate": 4.998930291731347e-05,
      "loss": 0.2641,
      "mean_token_accuracy": 0.9377644136548042,
      "num_tokens": 15664685.0,
      "step": 2110
    },
    {
      "entropy": 1.9264074131846427,
      "epoch": 0.03172082679886171,
      "grad_norm": 0.366871178150177,
      "learning_rate": 4.998894875930269e-05,
      "loss": 0.2593,
      "mean_token_accuracy": 0.933463591337204,
      "num_tokens": 15736965.0,
      "step": 2120
    },
    {
      "entropy": 1.9123641096055508,
      "epoch": 0.031870453340365784,
      "grad_norm": 0.4008675217628479,
      "learning_rate": 4.9988588834963356e-05,
      "loss": 0.2501,
      "mean_token_accuracy": 0.9419662795960904,
      "num_tokens": 15809779.0,
      "step": 2130
    },
    {
      "entropy": 1.8954274870455265,
      "epoch": 0.03202007988186985,
      "grad_norm": 0.37401312589645386,
      "learning_rate": 4.99882231443787e-05,
      "loss": 0.2581,
      "mean_token_accuracy": 0.9397182360291481,
      "num_tokens": 15884044.0,
      "step": 2140
    },
    {
      "entropy": 1.8952218472957612,
      "epoch": 0.03216970642337391,
      "grad_norm": 0.35495302081108093,
      "learning_rate": 4.998785168763326e-05,
      "loss": 0.2542,
      "mean_token_accuracy": 0.9380063027143478,
      "num_tokens": 15959314.0,
      "step": 2150
    },
    {
      "entropy": 1.9137567408382892,
      "epoch": 0.032319332964877975,
      "grad_norm": 0.42583364248275757,
      "learning_rate": 4.998747446481295e-05,
      "loss": 0.2605,
      "mean_token_accuracy": 0.9385600250214339,
      "num_tokens": 16032634.0,
      "step": 2160
    },
    {
      "entropy": 1.9397812642157077,
      "epoch": 0.03246895950638204,
      "grad_norm": 0.4027331471443176,
      "learning_rate": 4.998709147600497e-05,
      "loss": 0.2642,
      "mean_token_accuracy": 0.9363725662231446,
      "num_tokens": 16107629.0,
      "step": 2170
    },
    {
      "entropy": 1.9473442129790783,
      "epoch": 0.0326185860478861,
      "grad_norm": 0.3517669439315796,
      "learning_rate": 4.998670272129787e-05,
      "loss": 0.2504,
      "mean_token_accuracy": 0.9392262551933527,
      "num_tokens": 16179227.0,
      "step": 2180
    },
    {
      "entropy": 1.930887696146965,
      "epoch": 0.03276821258939017,
      "grad_norm": 0.48436281085014343,
      "learning_rate": 4.998630820078155e-05,
      "loss": 0.2573,
      "mean_token_accuracy": 0.9367777921259404,
      "num_tokens": 16250814.0,
      "step": 2190
    },
    {
      "entropy": 1.9166440114378929,
      "epoch": 0.03291783913089423,
      "grad_norm": 0.3394937217235565,
      "learning_rate": 4.998590791454723e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9465084623545408,
      "num_tokens": 16325394.0,
      "step": 2200
    },
    {
      "entropy": 1.943915279954672,
      "epoch": 0.0330674656723983,
      "grad_norm": 0.3154884874820709,
      "learning_rate": 4.998550186268746e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9397447355091572,
      "num_tokens": 16399145.0,
      "step": 2210
    },
    {
      "entropy": 1.982957059890032,
      "epoch": 0.033217092213902365,
      "grad_norm": 0.41696280241012573,
      "learning_rate": 4.9985090045296127e-05,
      "loss": 0.247,
      "mean_token_accuracy": 0.9376496184617281,
      "num_tokens": 16472355.0,
      "step": 2220
    },
    {
      "entropy": 1.9771963335573672,
      "epoch": 0.03336671875540643,
      "grad_norm": 0.2631385922431946,
      "learning_rate": 4.998467246246845e-05,
      "loss": 0.2577,
      "mean_token_accuracy": 0.936913312971592,
      "num_tokens": 16545261.0,
      "step": 2230
    },
    {
      "entropy": 1.9904627092182636,
      "epoch": 0.03351634529691049,
      "grad_norm": 0.44014114141464233,
      "learning_rate": 4.9984249114300966e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9391039680689574,
      "num_tokens": 16617993.0,
      "step": 2240
    },
    {
      "entropy": 1.9717852793633939,
      "epoch": 0.033665971838414556,
      "grad_norm": 0.3949442207813263,
      "learning_rate": 4.998382000089159e-05,
      "loss": 0.2527,
      "mean_token_accuracy": 0.9395819250494242,
      "num_tokens": 16690541.0,
      "step": 2250
    },
    {
      "entropy": 1.9732861898839473,
      "epoch": 0.03381559837991862,
      "grad_norm": 0.3988160490989685,
      "learning_rate": 4.998338512233953e-05,
      "loss": 0.2533,
      "mean_token_accuracy": 0.9382437761873007,
      "num_tokens": 16763956.0,
      "step": 2260
    },
    {
      "entropy": 1.9649958074092866,
      "epoch": 0.03396522492142268,
      "grad_norm": 0.4198133945465088,
      "learning_rate": 4.998294447874532e-05,
      "loss": 0.2498,
      "mean_token_accuracy": 0.9421830154955387,
      "num_tokens": 16838776.0,
      "step": 2270
    },
    {
      "entropy": 1.9426559664309024,
      "epoch": 0.034114851462926754,
      "grad_norm": 0.3109755218029022,
      "learning_rate": 4.9982498070210865e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.9420979168266058,
      "num_tokens": 16912019.0,
      "step": 2280
    },
    {
      "entropy": 1.9416701167821884,
      "epoch": 0.03426447800443082,
      "grad_norm": 0.4229649603366852,
      "learning_rate": 4.998204589683938e-05,
      "loss": 0.2612,
      "mean_token_accuracy": 0.936279496178031,
      "num_tokens": 16986643.0,
      "step": 2290
    },
    {
      "entropy": 1.9345242165029048,
      "epoch": 0.03441410454593488,
      "grad_norm": 0.4005569517612457,
      "learning_rate": 4.998158795873541e-05,
      "loss": 0.257,
      "mean_token_accuracy": 0.9352407984435558,
      "num_tokens": 17060966.0,
      "step": 2300
    },
    {
      "entropy": 1.9515607498586178,
      "epoch": 0.034563731087438945,
      "grad_norm": 0.3703019320964813,
      "learning_rate": 4.998112425600483e-05,
      "loss": 0.2553,
      "mean_token_accuracy": 0.9372815027832985,
      "num_tokens": 17134994.0,
      "step": 2310
    },
    {
      "entropy": 1.924177948385477,
      "epoch": 0.03471335762894301,
      "grad_norm": 0.33253714442253113,
      "learning_rate": 4.998065478875488e-05,
      "loss": 0.2524,
      "mean_token_accuracy": 0.9396979693323374,
      "num_tokens": 17209004.0,
      "step": 2320
    },
    {
      "entropy": 1.9030640989542007,
      "epoch": 0.03486298417044707,
      "grad_norm": 0.3523918688297272,
      "learning_rate": 4.9980179557094093e-05,
      "loss": 0.2472,
      "mean_token_accuracy": 0.9437584325671196,
      "num_tokens": 17282925.0,
      "step": 2330
    },
    {
      "entropy": 1.9764552630484105,
      "epoch": 0.035012610711951136,
      "grad_norm": 0.34804388880729675,
      "learning_rate": 4.997969856113235e-05,
      "loss": 0.2635,
      "mean_token_accuracy": 0.9387603487819434,
      "num_tokens": 17356211.0,
      "step": 2340
    },
    {
      "entropy": 1.9563065327703952,
      "epoch": 0.03516223725345521,
      "grad_norm": 0.3786119818687439,
      "learning_rate": 4.997921180098087e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9470246482640505,
      "num_tokens": 17434090.0,
      "step": 2350
    },
    {
      "entropy": 1.9553963489830495,
      "epoch": 0.03531186379495927,
      "grad_norm": 0.3564468026161194,
      "learning_rate": 4.99787192767522e-05,
      "loss": 0.2673,
      "mean_token_accuracy": 0.936869566142559,
      "num_tokens": 17506406.0,
      "step": 2360
    },
    {
      "entropy": 1.9346966020762921,
      "epoch": 0.035461490336463335,
      "grad_norm": 0.4221576154232025,
      "learning_rate": 4.997822098856022e-05,
      "loss": 0.2623,
      "mean_token_accuracy": 0.9381757996976375,
      "num_tokens": 17579377.0,
      "step": 2370
    },
    {
      "entropy": 1.9512953609228134,
      "epoch": 0.0356111168779674,
      "grad_norm": 0.34667572379112244,
      "learning_rate": 4.9977716936520147e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9382246319204569,
      "num_tokens": 17654093.0,
      "step": 2380
    },
    {
      "entropy": 1.9857023894786834,
      "epoch": 0.03576074341947146,
      "grad_norm": 0.3324300944805145,
      "learning_rate": 4.997720712074852e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9393995355814695,
      "num_tokens": 17729993.0,
      "step": 2390
    },
    {
      "entropy": 1.9606926009058951,
      "epoch": 0.035910369960975526,
      "grad_norm": 0.29386043548583984,
      "learning_rate": 4.9976691541363226e-05,
      "loss": 0.2428,
      "mean_token_accuracy": 0.9429989587515593,
      "num_tokens": 17802752.0,
      "step": 2400
    },
    {
      "entropy": 1.9807053744792937,
      "epoch": 0.03605999650247959,
      "grad_norm": 0.4032626450061798,
      "learning_rate": 4.9976170198483464e-05,
      "loss": 0.2553,
      "mean_token_accuracy": 0.9345075275748969,
      "num_tokens": 17876956.0,
      "step": 2410
    },
    {
      "entropy": 1.9699526779353618,
      "epoch": 0.03620962304398366,
      "grad_norm": 0.27722030878067017,
      "learning_rate": 4.997564309222979e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9399482186883688,
      "num_tokens": 17952200.0,
      "step": 2420
    },
    {
      "entropy": 1.9373407885432243,
      "epoch": 0.036359249585487724,
      "grad_norm": 0.3711972236633301,
      "learning_rate": 4.997511022272407e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9414932928979397,
      "num_tokens": 18027926.0,
      "step": 2430
    },
    {
      "entropy": 1.9610688969492913,
      "epoch": 0.03650887612699179,
      "grad_norm": 0.3180617392063141,
      "learning_rate": 4.997457159008952e-05,
      "loss": 0.2568,
      "mean_token_accuracy": 0.940676337853074,
      "num_tokens": 18103269.0,
      "step": 2440
    },
    {
      "entropy": 1.9607133604586124,
      "epoch": 0.03665850266849585,
      "grad_norm": 0.375173419713974,
      "learning_rate": 4.997402719445068e-05,
      "loss": 0.2612,
      "mean_token_accuracy": 0.9352148834615945,
      "num_tokens": 18176303.0,
      "step": 2450
    },
    {
      "entropy": 1.9812813706696033,
      "epoch": 0.036808129209999915,
      "grad_norm": 0.43037471175193787,
      "learning_rate": 4.997347703593342e-05,
      "loss": 0.2529,
      "mean_token_accuracy": 0.9383953481912612,
      "num_tokens": 18247801.0,
      "step": 2460
    },
    {
      "entropy": 1.9305276803672313,
      "epoch": 0.03695775575150398,
      "grad_norm": 0.3978329300880432,
      "learning_rate": 4.997292111466496e-05,
      "loss": 0.2685,
      "mean_token_accuracy": 0.9387603871524334,
      "num_tokens": 18323608.0,
      "step": 2470
    },
    {
      "entropy": 1.9468570463359356,
      "epoch": 0.03710738229300804,
      "grad_norm": 0.4265293478965759,
      "learning_rate": 4.9972359430773815e-05,
      "loss": 0.2753,
      "mean_token_accuracy": 0.9357739523053169,
      "num_tokens": 18395689.0,
      "step": 2480
    },
    {
      "entropy": 1.9424455896019936,
      "epoch": 0.03725700883451211,
      "grad_norm": 0.4507357180118561,
      "learning_rate": 4.9971791984389874e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9411885067820549,
      "num_tokens": 18468220.0,
      "step": 2490
    },
    {
      "entropy": 1.923718734830618,
      "epoch": 0.03740663537601618,
      "grad_norm": 0.3139372169971466,
      "learning_rate": 4.997121877564434e-05,
      "loss": 0.2354,
      "mean_token_accuracy": 0.9458344623446464,
      "num_tokens": 18540482.0,
      "step": 2500
    },
    {
      "entropy": 1.9026800259947776,
      "epoch": 0.03755626191752024,
      "grad_norm": 0.31177815794944763,
      "learning_rate": 4.997063980466975e-05,
      "loss": 0.267,
      "mean_token_accuracy": 0.9363665070384741,
      "num_tokens": 18616288.0,
      "step": 2510
    },
    {
      "entropy": 1.9846577145159245,
      "epoch": 0.037705888459024305,
      "grad_norm": 0.2984126806259155,
      "learning_rate": 4.997005507159996e-05,
      "loss": 0.2756,
      "mean_token_accuracy": 0.9323837928473949,
      "num_tokens": 18687150.0,
      "step": 2520
    },
    {
      "entropy": 1.9855789996683597,
      "epoch": 0.03785551500052837,
      "grad_norm": 0.3554929494857788,
      "learning_rate": 4.996946457657018e-05,
      "loss": 0.2483,
      "mean_token_accuracy": 0.9386679690331221,
      "num_tokens": 18762178.0,
      "step": 2530
    },
    {
      "entropy": 1.9625399574637412,
      "epoch": 0.03800514154203243,
      "grad_norm": 0.4682481586933136,
      "learning_rate": 4.996886831971694e-05,
      "loss": 0.2522,
      "mean_token_accuracy": 0.9369754523038865,
      "num_tokens": 18838178.0,
      "step": 2540
    },
    {
      "entropy": 1.953991135954857,
      "epoch": 0.038154768083536496,
      "grad_norm": 0.39189374446868896,
      "learning_rate": 4.996826630117811e-05,
      "loss": 0.2596,
      "mean_token_accuracy": 0.933669513463974,
      "num_tokens": 18913184.0,
      "step": 2550
    },
    {
      "entropy": 1.9629682764410972,
      "epoch": 0.03830439462504056,
      "grad_norm": 0.30209338665008545,
      "learning_rate": 4.996765852109288e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.9371741250157356,
      "num_tokens": 18989742.0,
      "step": 2560
    },
    {
      "entropy": 1.9424784727394582,
      "epoch": 0.03845402116654463,
      "grad_norm": 0.3802827000617981,
      "learning_rate": 4.996704497960179e-05,
      "loss": 0.2618,
      "mean_token_accuracy": 0.9349131889641284,
      "num_tokens": 19062948.0,
      "step": 2570
    },
    {
      "entropy": 1.9612093448638916,
      "epoch": 0.038603647708048694,
      "grad_norm": 0.3516566753387451,
      "learning_rate": 4.996642567684669e-05,
      "loss": 0.2519,
      "mean_token_accuracy": 0.9395324651151895,
      "num_tokens": 19135771.0,
      "step": 2580
    },
    {
      "entropy": 1.9820048101246357,
      "epoch": 0.03875327424955276,
      "grad_norm": 0.2763564884662628,
      "learning_rate": 4.996580061297079e-05,
      "loss": 0.2584,
      "mean_token_accuracy": 0.9388410236686469,
      "num_tokens": 19210749.0,
      "step": 2590
    },
    {
      "entropy": 2.014185433834791,
      "epoch": 0.03890290079105682,
      "grad_norm": 0.39339205622673035,
      "learning_rate": 4.9965169788118596e-05,
      "loss": 0.2604,
      "mean_token_accuracy": 0.9401724830269813,
      "num_tokens": 19285305.0,
      "step": 2600
    },
    {
      "entropy": 2.058422737568617,
      "epoch": 0.039052527332560885,
      "grad_norm": 0.3765966594219208,
      "learning_rate": 4.996453320243598e-05,
      "loss": 0.2638,
      "mean_token_accuracy": 0.9368779562413693,
      "num_tokens": 19357731.0,
      "step": 2610
    },
    {
      "entropy": 2.0115641206502914,
      "epoch": 0.03920215387406495,
      "grad_norm": 0.411109060049057,
      "learning_rate": 4.996389085607012e-05,
      "loss": 0.2604,
      "mean_token_accuracy": 0.9363212723284959,
      "num_tokens": 19431994.0,
      "step": 2620
    },
    {
      "entropy": 1.9735753037035466,
      "epoch": 0.03935178041556901,
      "grad_norm": 0.444417268037796,
      "learning_rate": 4.996324274916955e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9439159881323576,
      "num_tokens": 19505137.0,
      "step": 2630
    },
    {
      "entropy": 1.9699680179357528,
      "epoch": 0.03950140695707308,
      "grad_norm": 0.36013904213905334,
      "learning_rate": 4.996258888188411e-05,
      "loss": 0.256,
      "mean_token_accuracy": 0.9401726860553026,
      "num_tokens": 19577925.0,
      "step": 2640
    },
    {
      "entropy": 1.9285006485879421,
      "epoch": 0.03965103349857715,
      "grad_norm": 0.3732374310493469,
      "learning_rate": 4.996192925436501e-05,
      "loss": 0.2607,
      "mean_token_accuracy": 0.9384304087609052,
      "num_tokens": 19652924.0,
      "step": 2650
    },
    {
      "entropy": 1.9509468123316764,
      "epoch": 0.03980066004008121,
      "grad_norm": 0.4283207654953003,
      "learning_rate": 4.996126386676474e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9397526569664478,
      "num_tokens": 19726067.0,
      "step": 2660
    },
    {
      "entropy": 1.9705830961465836,
      "epoch": 0.039950286581585275,
      "grad_norm": 0.35293203592300415,
      "learning_rate": 4.996059271923717e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9455716229975224,
      "num_tokens": 19804418.0,
      "step": 2670
    },
    {
      "entropy": 1.932836677134037,
      "epoch": 0.04009991312308934,
      "grad_norm": 0.3222312033176422,
      "learning_rate": 4.995991581193747e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9442108020186424,
      "num_tokens": 19881488.0,
      "step": 2680
    },
    {
      "entropy": 1.9573209173977375,
      "epoch": 0.0402495396645934,
      "grad_norm": 0.42405036091804504,
      "learning_rate": 4.995923314502216e-05,
      "loss": 0.2628,
      "mean_token_accuracy": 0.9384361617267132,
      "num_tokens": 19956782.0,
      "step": 2690
    },
    {
      "entropy": 1.9718046344816684,
      "epoch": 0.040399166206097466,
      "grad_norm": 0.44667866826057434,
      "learning_rate": 4.995854471864908e-05,
      "loss": 0.259,
      "mean_token_accuracy": 0.9392971463501454,
      "num_tokens": 20031403.0,
      "step": 2700
    },
    {
      "entropy": 2.0155383914709093,
      "epoch": 0.040548792747601536,
      "grad_norm": 0.3875434994697571,
      "learning_rate": 4.99578505329774e-05,
      "loss": 0.2699,
      "mean_token_accuracy": 0.9344771649688483,
      "num_tokens": 20106197.0,
      "step": 2710
    },
    {
      "entropy": 2.052448569238186,
      "epoch": 0.0406984192891056,
      "grad_norm": 0.29686251282691956,
      "learning_rate": 4.995715058816764e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9383937004953623,
      "num_tokens": 20181582.0,
      "step": 2720
    },
    {
      "entropy": 2.052899253368378,
      "epoch": 0.040848045830609664,
      "grad_norm": 0.3088366389274597,
      "learning_rate": 4.995644488438163e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.9385699488222599,
      "num_tokens": 20257998.0,
      "step": 2730
    },
    {
      "entropy": 1.984103611111641,
      "epoch": 0.04099767237211373,
      "grad_norm": 0.37731194496154785,
      "learning_rate": 4.995573342178254e-05,
      "loss": 0.2631,
      "mean_token_accuracy": 0.9371977284550667,
      "num_tokens": 20333779.0,
      "step": 2740
    },
    {
      "entropy": 1.9732327796518803,
      "epoch": 0.04114729891361779,
      "grad_norm": 0.31622132658958435,
      "learning_rate": 4.995501620053489e-05,
      "loss": 0.2542,
      "mean_token_accuracy": 0.9404299832880497,
      "num_tokens": 20406214.0,
      "step": 2750
    },
    {
      "entropy": 2.0346471764147283,
      "epoch": 0.041296925455121855,
      "grad_norm": 0.3011432886123657,
      "learning_rate": 4.9954293220804506e-05,
      "loss": 0.2626,
      "mean_token_accuracy": 0.9376473885029555,
      "num_tokens": 20478748.0,
      "step": 2760
    },
    {
      "entropy": 2.0026635088026525,
      "epoch": 0.04144655199662592,
      "grad_norm": 0.349150151014328,
      "learning_rate": 4.9953564482758545e-05,
      "loss": 0.2399,
      "mean_token_accuracy": 0.9408206082880497,
      "num_tokens": 20552214.0,
      "step": 2770
    },
    {
      "entropy": 2.019057286530733,
      "epoch": 0.04159617853812999,
      "grad_norm": 0.3964208662509918,
      "learning_rate": 4.995282998656551e-05,
      "loss": 0.2478,
      "mean_token_accuracy": 0.9412313893437385,
      "num_tokens": 20628151.0,
      "step": 2780
    },
    {
      "entropy": 1.9956378206610679,
      "epoch": 0.04174580507963405,
      "grad_norm": 0.2539881467819214,
      "learning_rate": 4.995208973239523e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9400748755782843,
      "num_tokens": 20704679.0,
      "step": 2790
    },
    {
      "entropy": 1.9954710811376573,
      "epoch": 0.04189543162113812,
      "grad_norm": 0.4153245985507965,
      "learning_rate": 4.995134372041887e-05,
      "loss": 0.2523,
      "mean_token_accuracy": 0.9425605285912753,
      "num_tokens": 20779516.0,
      "step": 2800
    },
    {
      "entropy": 1.9946840420365333,
      "epoch": 0.04204505816264218,
      "grad_norm": 1.7740190029144287,
      "learning_rate": 4.9950591950808915e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9387598268687725,
      "num_tokens": 20855255.0,
      "step": 2810
    },
    {
      "entropy": 1.9897942021489143,
      "epoch": 0.042194684704146244,
      "grad_norm": 0.35994648933410645,
      "learning_rate": 4.994983442373918e-05,
      "loss": 0.2638,
      "mean_token_accuracy": 0.9404065068811178,
      "num_tokens": 20927600.0,
      "step": 2820
    },
    {
      "entropy": 1.9824798643589019,
      "epoch": 0.04234431124565031,
      "grad_norm": 0.3997035324573517,
      "learning_rate": 4.9949071139384835e-05,
      "loss": 0.2652,
      "mean_token_accuracy": 0.9389049980789423,
      "num_tokens": 21002207.0,
      "step": 2830
    },
    {
      "entropy": 1.976809711754322,
      "epoch": 0.04249393778715437,
      "grad_norm": 0.3539193868637085,
      "learning_rate": 4.994830209792236e-05,
      "loss": 0.2556,
      "mean_token_accuracy": 0.941960509493947,
      "num_tokens": 21075338.0,
      "step": 2840
    },
    {
      "entropy": 1.9512909412384034,
      "epoch": 0.042643564328658436,
      "grad_norm": 0.3265222907066345,
      "learning_rate": 4.9947527299529556e-05,
      "loss": 0.2471,
      "mean_token_accuracy": 0.9411813333630562,
      "num_tokens": 21148296.0,
      "step": 2850
    },
    {
      "entropy": 1.9850236222147941,
      "epoch": 0.042793190870162506,
      "grad_norm": 0.4315653145313263,
      "learning_rate": 4.9946746744385585e-05,
      "loss": 0.2464,
      "mean_token_accuracy": 0.9355928212404251,
      "num_tokens": 21223892.0,
      "step": 2860
    },
    {
      "entropy": 1.995324043929577,
      "epoch": 0.04294281741166657,
      "grad_norm": 0.41415536403656006,
      "learning_rate": 4.994596043267093e-05,
      "loss": 0.2641,
      "mean_token_accuracy": 0.9354174751788378,
      "num_tokens": 21299910.0,
      "step": 2870
    },
    {
      "entropy": 1.9925806313753127,
      "epoch": 0.043092443953170634,
      "grad_norm": 0.34089967608451843,
      "learning_rate": 4.994516836456739e-05,
      "loss": 0.2502,
      "mean_token_accuracy": 0.9383448909968137,
      "num_tokens": 21372615.0,
      "step": 2880
    },
    {
      "entropy": 1.9594772972166539,
      "epoch": 0.0432420704946747,
      "grad_norm": 0.31095007061958313,
      "learning_rate": 4.994437054025811e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9437022406607867,
      "num_tokens": 21448511.0,
      "step": 2890
    },
    {
      "entropy": 1.9714134089648723,
      "epoch": 0.04339169703617876,
      "grad_norm": 0.35654157400131226,
      "learning_rate": 4.994356695992756e-05,
      "loss": 0.2408,
      "mean_token_accuracy": 0.943603714928031,
      "num_tokens": 21523201.0,
      "step": 2900
    },
    {
      "entropy": 1.9646256640553474,
      "epoch": 0.043541323577682825,
      "grad_norm": 0.27496638894081116,
      "learning_rate": 4.9942757623761535e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9433745224028826,
      "num_tokens": 21603228.0,
      "step": 2910
    },
    {
      "entropy": 1.9979079753160476,
      "epoch": 0.04369095011918689,
      "grad_norm": 0.39425766468048096,
      "learning_rate": 4.994194253194718e-05,
      "loss": 0.2472,
      "mean_token_accuracy": 0.9408998049795627,
      "num_tokens": 21675327.0,
      "step": 2920
    },
    {
      "entropy": 1.9493498086929322,
      "epoch": 0.04384057666069096,
      "grad_norm": 0.37097832560539246,
      "learning_rate": 4.9941121684672946e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.9407390601933002,
      "num_tokens": 21749131.0,
      "step": 2930
    },
    {
      "entropy": 1.9617694713175298,
      "epoch": 0.04399020320219502,
      "grad_norm": 0.3001614809036255,
      "learning_rate": 4.994029508212864e-05,
      "loss": 0.2614,
      "mean_token_accuracy": 0.9374994207173586,
      "num_tokens": 21824553.0,
      "step": 2940
    },
    {
      "entropy": 1.9666986733675003,
      "epoch": 0.04413982974369909,
      "grad_norm": 0.4024062752723694,
      "learning_rate": 4.993946272450538e-05,
      "loss": 0.2525,
      "mean_token_accuracy": 0.9387096963822842,
      "num_tokens": 21899985.0,
      "step": 2950
    },
    {
      "entropy": 1.9731488317251205,
      "epoch": 0.04428945628520315,
      "grad_norm": 0.3654042184352875,
      "learning_rate": 4.993862461199562e-05,
      "loss": 0.2541,
      "mean_token_accuracy": 0.935540396720171,
      "num_tokens": 21972813.0,
      "step": 2960
    },
    {
      "entropy": 1.925535501539707,
      "epoch": 0.044439082826707214,
      "grad_norm": 0.4073169231414795,
      "learning_rate": 4.9937780744793164e-05,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9387246772646904,
      "num_tokens": 22047859.0,
      "step": 2970
    },
    {
      "entropy": 1.910414982587099,
      "epoch": 0.04458870936821128,
      "grad_norm": 0.41044309735298157,
      "learning_rate": 4.99369311230931e-05,
      "loss": 0.2514,
      "mean_token_accuracy": 0.9373502794653177,
      "num_tokens": 22122082.0,
      "step": 2980
    },
    {
      "entropy": 1.924702125787735,
      "epoch": 0.04473833590971534,
      "grad_norm": 0.3491014540195465,
      "learning_rate": 4.99360757470919e-05,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9412907123565674,
      "num_tokens": 22195899.0,
      "step": 2990
    },
    {
      "entropy": 1.9836353830993176,
      "epoch": 0.04488796245121941,
      "grad_norm": 0.3577398955821991,
      "learning_rate": 4.993521461698733e-05,
      "loss": 0.2493,
      "mean_token_accuracy": 0.938875550404191,
      "num_tokens": 22267493.0,
      "step": 3000
    },
    {
      "epoch": 0.04488796245121941,
      "eval_entropy": 1.9614721521377563,
      "eval_loss": 0.27463504672050476,
      "eval_mean_token_accuracy": 0.9400859427690506,
      "eval_num_tokens": 22267493.0,
      "eval_runtime": 561.2812,
      "eval_samples_per_second": 35.633,
      "eval_steps_per_second": 8.908,
      "step": 3000
    },
    {
      "entropy": 1.951661965996027,
      "epoch": 0.045037588992723476,
      "grad_norm": 0.3033333122730255,
      "learning_rate": 4.993434773297849e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9390652362257242,
      "num_tokens": 22343028.0,
      "step": 3010
    },
    {
      "entropy": 1.983953095972538,
      "epoch": 0.04518721553422754,
      "grad_norm": 0.2951551675796509,
      "learning_rate": 4.993347509526585e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.942112996801734,
      "num_tokens": 22416955.0,
      "step": 3020
    },
    {
      "entropy": 1.970123067498207,
      "epoch": 0.045336842075731604,
      "grad_norm": 0.3634348213672638,
      "learning_rate": 4.993259670405115e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9429294042289257,
      "num_tokens": 22490319.0,
      "step": 3030
    },
    {
      "entropy": 1.9808405317366122,
      "epoch": 0.04548646861723567,
      "grad_norm": 0.28851082921028137,
      "learning_rate": 4.9931712559537505e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9430048204958439,
      "num_tokens": 22562914.0,
      "step": 3040
    },
    {
      "entropy": 1.9455489844083786,
      "epoch": 0.04563609515873973,
      "grad_norm": 0.2947673499584198,
      "learning_rate": 4.9930822661929326e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9434957057237625,
      "num_tokens": 22636390.0,
      "step": 3050
    },
    {
      "entropy": 1.9479838095605373,
      "epoch": 0.045785721700243795,
      "grad_norm": 0.39096105098724365,
      "learning_rate": 4.9929927011432395e-05,
      "loss": 0.263,
      "mean_token_accuracy": 0.9384656347334385,
      "num_tokens": 22711857.0,
      "step": 3060
    },
    {
      "entropy": 1.9661613628268242,
      "epoch": 0.045935348241747866,
      "grad_norm": 0.37372392416000366,
      "learning_rate": 4.99290256082538e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9406507927924395,
      "num_tokens": 22785524.0,
      "step": 3070
    },
    {
      "entropy": 1.9609551206231117,
      "epoch": 0.04608497478325193,
      "grad_norm": 0.45049989223480225,
      "learning_rate": 4.9928118452601955e-05,
      "loss": 0.2622,
      "mean_token_accuracy": 0.9391500268131494,
      "num_tokens": 22859637.0,
      "step": 3080
    },
    {
      "entropy": 1.9297973394393921,
      "epoch": 0.04623460132475599,
      "grad_norm": 0.4082679748535156,
      "learning_rate": 4.99272055446866e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9419586844742298,
      "num_tokens": 22934550.0,
      "step": 3090
    },
    {
      "entropy": 1.9703654855489732,
      "epoch": 0.04638422786626006,
      "grad_norm": 0.3543381094932556,
      "learning_rate": 4.992628688471884e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9391310147941112,
      "num_tokens": 23009118.0,
      "step": 3100
    },
    {
      "entropy": 1.9985708743333817,
      "epoch": 0.04653385440776412,
      "grad_norm": 0.40854477882385254,
      "learning_rate": 4.9925362472911055e-05,
      "loss": 0.2649,
      "mean_token_accuracy": 0.9377143040299416,
      "num_tokens": 23083327.0,
      "step": 3110
    },
    {
      "entropy": 1.9200824454426766,
      "epoch": 0.046683480949268184,
      "grad_norm": 0.3957577049732208,
      "learning_rate": 4.992443230947701e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9443641010671854,
      "num_tokens": 23155628.0,
      "step": 3120
    },
    {
      "entropy": 1.9159224435687066,
      "epoch": 0.04683310749077225,
      "grad_norm": 0.3775918185710907,
      "learning_rate": 4.9923496394631776e-05,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9411398988217116,
      "num_tokens": 23228741.0,
      "step": 3130
    },
    {
      "entropy": 1.94199625775218,
      "epoch": 0.04698273403227632,
      "grad_norm": 0.3637406826019287,
      "learning_rate": 4.9922554728591744e-05,
      "loss": 0.2462,
      "mean_token_accuracy": 0.9387629423290491,
      "num_tokens": 23305232.0,
      "step": 3140
    },
    {
      "entropy": 1.9253748051822186,
      "epoch": 0.04713236057378038,
      "grad_norm": 0.31956860423088074,
      "learning_rate": 4.992160731157464e-05,
      "loss": 0.2621,
      "mean_token_accuracy": 0.9364793367683888,
      "num_tokens": 23376590.0,
      "step": 3150
    },
    {
      "entropy": 1.9512000434100627,
      "epoch": 0.047281987115284446,
      "grad_norm": 0.4387859106063843,
      "learning_rate": 4.992065414379953e-05,
      "loss": 0.2641,
      "mean_token_accuracy": 0.9401253160089255,
      "num_tokens": 23450042.0,
      "step": 3160
    },
    {
      "entropy": 1.9227629132568835,
      "epoch": 0.04743161365678851,
      "grad_norm": 0.3563423156738281,
      "learning_rate": 4.99196952254868e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.9419966097921133,
      "num_tokens": 23524111.0,
      "step": 3170
    },
    {
      "entropy": 1.9600928090512753,
      "epoch": 0.047581240198292574,
      "grad_norm": 0.3269404172897339,
      "learning_rate": 4.991873055685817e-05,
      "loss": 0.2527,
      "mean_token_accuracy": 0.9378321200609208,
      "num_tokens": 23599342.0,
      "step": 3180
    },
    {
      "entropy": 1.9521087639033794,
      "epoch": 0.04773086673979664,
      "grad_norm": 0.3579699695110321,
      "learning_rate": 4.991776013813669e-05,
      "loss": 0.2408,
      "mean_token_accuracy": 0.9401014409959316,
      "num_tokens": 23673484.0,
      "step": 3190
    },
    {
      "entropy": 1.9315286450088025,
      "epoch": 0.0478804932813007,
      "grad_norm": 0.2909514605998993,
      "learning_rate": 4.9916783969546746e-05,
      "loss": 0.2497,
      "mean_token_accuracy": 0.9390591260045766,
      "num_tokens": 23746016.0,
      "step": 3200
    },
    {
      "entropy": 1.9777843527495862,
      "epoch": 0.048030119822804765,
      "grad_norm": 0.329152375459671,
      "learning_rate": 4.991580205131403e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.938442450389266,
      "num_tokens": 23818892.0,
      "step": 3210
    },
    {
      "entropy": 1.9458057560026645,
      "epoch": 0.048179746364308836,
      "grad_norm": 0.3980790078639984,
      "learning_rate": 4.991481438366559e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9381002299487591,
      "num_tokens": 23890554.0,
      "step": 3220
    },
    {
      "entropy": 1.9495098195970058,
      "epoch": 0.0483293729058129,
      "grad_norm": 0.3794592320919037,
      "learning_rate": 4.991382096682978e-05,
      "loss": 0.2389,
      "mean_token_accuracy": 0.942136612161994,
      "num_tokens": 23963817.0,
      "step": 3230
    },
    {
      "entropy": 1.972213963419199,
      "epoch": 0.04847899944731696,
      "grad_norm": 0.3533346354961395,
      "learning_rate": 4.9912821801036305e-05,
      "loss": 0.2611,
      "mean_token_accuracy": 0.9350441813468933,
      "num_tokens": 24039361.0,
      "step": 3240
    },
    {
      "entropy": 1.9682801753282546,
      "epoch": 0.04862862598882103,
      "grad_norm": 0.2912701368331909,
      "learning_rate": 4.991181688651619e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9433543212711811,
      "num_tokens": 24114753.0,
      "step": 3250
    },
    {
      "entropy": 1.9753825806081295,
      "epoch": 0.04877825253032509,
      "grad_norm": 0.34170612692832947,
      "learning_rate": 4.9910806223501786e-05,
      "loss": 0.2673,
      "mean_token_accuracy": 0.9385187692940236,
      "num_tokens": 24189928.0,
      "step": 3260
    },
    {
      "entropy": 1.9833362840116024,
      "epoch": 0.048927879071829154,
      "grad_norm": 0.26108744740486145,
      "learning_rate": 4.990978981222679e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.9427099876105786,
      "num_tokens": 24262151.0,
      "step": 3270
    },
    {
      "entropy": 1.981126519292593,
      "epoch": 0.04907750561333322,
      "grad_norm": 0.40851423144340515,
      "learning_rate": 4.9908767652926195e-05,
      "loss": 0.276,
      "mean_token_accuracy": 0.9340423602610827,
      "num_tokens": 24336211.0,
      "step": 3280
    },
    {
      "entropy": 1.9556941881775856,
      "epoch": 0.04922713215483729,
      "grad_norm": 0.4202451705932617,
      "learning_rate": 4.990773974583635e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9392862305045128,
      "num_tokens": 24408083.0,
      "step": 3290
    },
    {
      "entropy": 1.9119682967662812,
      "epoch": 0.04937675869634135,
      "grad_norm": 0.40605074167251587,
      "learning_rate": 4.9906706091194926e-05,
      "loss": 0.255,
      "mean_token_accuracy": 0.9372483100742102,
      "num_tokens": 24481936.0,
      "step": 3300
    },
    {
      "entropy": 1.949040424823761,
      "epoch": 0.049526385237845416,
      "grad_norm": 0.3744630217552185,
      "learning_rate": 4.990566668924092e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9400570672005415,
      "num_tokens": 24555468.0,
      "step": 3310
    },
    {
      "entropy": 1.963760080933571,
      "epoch": 0.04967601177934948,
      "grad_norm": 0.4890027344226837,
      "learning_rate": 4.990462154021467e-05,
      "loss": 0.2477,
      "mean_token_accuracy": 0.9406858198344707,
      "num_tokens": 24629865.0,
      "step": 3320
    },
    {
      "entropy": 1.9702147126197815,
      "epoch": 0.049825638320853544,
      "grad_norm": 0.3138258457183838,
      "learning_rate": 4.990357064435781e-05,
      "loss": 0.2554,
      "mean_token_accuracy": 0.9399385418742895,
      "num_tokens": 24706306.0,
      "step": 3330
    },
    {
      "entropy": 1.9547279819846153,
      "epoch": 0.04997526486235761,
      "grad_norm": 0.3369857370853424,
      "learning_rate": 4.9902514001913356e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.9401251390576363,
      "num_tokens": 24780087.0,
      "step": 3340
    },
    {
      "entropy": 1.9674689784646033,
      "epoch": 0.05012489140386167,
      "grad_norm": 0.3867260813713074,
      "learning_rate": 4.99014516131256e-05,
      "loss": 0.2689,
      "mean_token_accuracy": 0.9365192674100399,
      "num_tokens": 24852424.0,
      "step": 3350
    },
    {
      "entropy": 1.949245223402977,
      "epoch": 0.05027451794536574,
      "grad_norm": 0.3727648854255676,
      "learning_rate": 4.990038347824019e-05,
      "loss": 0.2427,
      "mean_token_accuracy": 0.9440439358353615,
      "num_tokens": 24928413.0,
      "step": 3360
    },
    {
      "entropy": 1.9658662863075733,
      "epoch": 0.050424144486869805,
      "grad_norm": 0.23317165672779083,
      "learning_rate": 4.98993095975041e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9429517719894648,
      "num_tokens": 25002984.0,
      "step": 3370
    },
    {
      "entropy": 1.977156399190426,
      "epoch": 0.05057377102837387,
      "grad_norm": 0.3376786410808563,
      "learning_rate": 4.989822997116564e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9428504556417465,
      "num_tokens": 25075234.0,
      "step": 3380
    },
    {
      "entropy": 1.9664802446961402,
      "epoch": 0.05072339756987793,
      "grad_norm": 0.3119570314884186,
      "learning_rate": 4.989714459947443e-05,
      "loss": 0.2412,
      "mean_token_accuracy": 0.9413769237697125,
      "num_tokens": 25150203.0,
      "step": 3390
    },
    {
      "entropy": 1.980876636505127,
      "epoch": 0.050873024111382,
      "grad_norm": 0.4399169683456421,
      "learning_rate": 4.989605348268141e-05,
      "loss": 0.2553,
      "mean_token_accuracy": 0.9360529657453298,
      "num_tokens": 25223559.0,
      "step": 3400
    },
    {
      "entropy": 2.018413391709328,
      "epoch": 0.05102265065288606,
      "grad_norm": 0.3952925503253937,
      "learning_rate": 4.9894956621038894e-05,
      "loss": 0.2555,
      "mean_token_accuracy": 0.9352204851806164,
      "num_tokens": 25296682.0,
      "step": 3410
    },
    {
      "entropy": 2.028859171271324,
      "epoch": 0.051172277194390124,
      "grad_norm": 0.3790878653526306,
      "learning_rate": 4.9893854014800486e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9425794787704944,
      "num_tokens": 25369214.0,
      "step": 3420
    },
    {
      "entropy": 1.9781946122646332,
      "epoch": 0.051321903735894195,
      "grad_norm": 0.2900044023990631,
      "learning_rate": 4.989274566422113e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9459357675164938,
      "num_tokens": 25446595.0,
      "step": 3430
    },
    {
      "entropy": 1.9665522664785384,
      "epoch": 0.05147153027739826,
      "grad_norm": 0.4413408935070038,
      "learning_rate": 4.989163156955708e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9435696020722389,
      "num_tokens": 25520219.0,
      "step": 3440
    },
    {
      "entropy": 1.908288972824812,
      "epoch": 0.05162115681890232,
      "grad_norm": 0.3438352644443512,
      "learning_rate": 4.989051173106596e-05,
      "loss": 0.2401,
      "mean_token_accuracy": 0.940539662912488,
      "num_tokens": 25595233.0,
      "step": 3450
    },
    {
      "entropy": 1.8754162169992923,
      "epoch": 0.051770783360406386,
      "grad_norm": 0.359891414642334,
      "learning_rate": 4.988938614900668e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9396415658295154,
      "num_tokens": 25669657.0,
      "step": 3460
    },
    {
      "entropy": 1.9030602879822254,
      "epoch": 0.05192040990191045,
      "grad_norm": 0.2997036576271057,
      "learning_rate": 4.988825482363949e-05,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9442131035029888,
      "num_tokens": 25747011.0,
      "step": 3470
    },
    {
      "entropy": 1.9490637734532357,
      "epoch": 0.052070036443414514,
      "grad_norm": 0.3506845235824585,
      "learning_rate": 4.9887117755225986e-05,
      "loss": 0.2549,
      "mean_token_accuracy": 0.937953794375062,
      "num_tokens": 25825150.0,
      "step": 3480
    },
    {
      "entropy": 1.97315364331007,
      "epoch": 0.05221966298491858,
      "grad_norm": 0.3471989333629608,
      "learning_rate": 4.988597494402908e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9415992636233568,
      "num_tokens": 25898431.0,
      "step": 3490
    },
    {
      "entropy": 1.9594019860029221,
      "epoch": 0.05236928952642264,
      "grad_norm": 0.3337489068508148,
      "learning_rate": 4.9884826390313e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9380198605358601,
      "num_tokens": 25972525.0,
      "step": 3500
    },
    {
      "entropy": 1.9328558690845967,
      "epoch": 0.05251891606792671,
      "grad_norm": 0.33132368326187134,
      "learning_rate": 4.988367209434332e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9402026195079088,
      "num_tokens": 26045348.0,
      "step": 3510
    },
    {
      "entropy": 1.9625490367412568,
      "epoch": 0.052668542609430775,
      "grad_norm": 0.3778687119483948,
      "learning_rate": 4.9882512056386924e-05,
      "loss": 0.2397,
      "mean_token_accuracy": 0.9408181730657816,
      "num_tokens": 26118307.0,
      "step": 3520
    },
    {
      "entropy": 1.953147780895233,
      "epoch": 0.05281816915093484,
      "grad_norm": 0.2973726689815521,
      "learning_rate": 4.9881346276712036e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9435360737144947,
      "num_tokens": 26190337.0,
      "step": 3530
    },
    {
      "entropy": 1.9584115371108055,
      "epoch": 0.0529677956924389,
      "grad_norm": 0.2827855050563812,
      "learning_rate": 4.988017475558821e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9449221797287464,
      "num_tokens": 26264793.0,
      "step": 3540
    },
    {
      "entropy": 1.9118625171482564,
      "epoch": 0.05311742223394297,
      "grad_norm": 0.41689079999923706,
      "learning_rate": 4.987899749328632e-05,
      "loss": 0.2587,
      "mean_token_accuracy": 0.936268400400877,
      "num_tokens": 26341464.0,
      "step": 3550
    },
    {
      "entropy": 1.916252052038908,
      "epoch": 0.05326704877544703,
      "grad_norm": 0.4027631878852844,
      "learning_rate": 4.9877814490078576e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9422241192311048,
      "num_tokens": 26415016.0,
      "step": 3560
    },
    {
      "entropy": 1.9182971835136413,
      "epoch": 0.053416675316951094,
      "grad_norm": 0.31365469098091125,
      "learning_rate": 4.9876625746238504e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9404305092990398,
      "num_tokens": 26488861.0,
      "step": 3570
    },
    {
      "entropy": 1.9366491749882697,
      "epoch": 0.053566301858455165,
      "grad_norm": 0.3853353261947632,
      "learning_rate": 4.987543126204096e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9417268808931112,
      "num_tokens": 26562806.0,
      "step": 3580
    },
    {
      "entropy": 1.8875850826501845,
      "epoch": 0.05371592839995923,
      "grad_norm": 0.33684220910072327,
      "learning_rate": 4.987423103776213e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.940117359906435,
      "num_tokens": 26639333.0,
      "step": 3590
    },
    {
      "entropy": 1.900030543655157,
      "epoch": 0.05386555494146329,
      "grad_norm": 0.469058632850647,
      "learning_rate": 4.987302507367953e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9396737739443779,
      "num_tokens": 26712277.0,
      "step": 3600
    },
    {
      "entropy": 1.898341841250658,
      "epoch": 0.054015181482967356,
      "grad_norm": 0.33390530943870544,
      "learning_rate": 4.9871813370071996e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9381330009549856,
      "num_tokens": 26785683.0,
      "step": 3610
    },
    {
      "entropy": 1.9147989459335804,
      "epoch": 0.05416480802447142,
      "grad_norm": 0.3603259325027466,
      "learning_rate": 4.987059592721971e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9395476449280977,
      "num_tokens": 26861265.0,
      "step": 3620
    },
    {
      "entropy": 1.9191819220781325,
      "epoch": 0.054314434565975483,
      "grad_norm": 0.362998366355896,
      "learning_rate": 4.986937274540415e-05,
      "loss": 0.2767,
      "mean_token_accuracy": 0.9337626721709966,
      "num_tokens": 26932032.0,
      "step": 3630
    },
    {
      "entropy": 1.9244706593453884,
      "epoch": 0.05446406110747955,
      "grad_norm": 0.3459271788597107,
      "learning_rate": 4.986814382490815e-05,
      "loss": 0.246,
      "mean_token_accuracy": 0.9372870121151209,
      "num_tokens": 27003386.0,
      "step": 3640
    },
    {
      "entropy": 1.8514331690967083,
      "epoch": 0.05461368764898362,
      "grad_norm": 0.3830394148826599,
      "learning_rate": 4.986690916601584e-05,
      "loss": 0.2451,
      "mean_token_accuracy": 0.9406349491328001,
      "num_tokens": 27075528.0,
      "step": 3650
    },
    {
      "entropy": 1.8538854002952576,
      "epoch": 0.05476331419048768,
      "grad_norm": 0.34492596983909607,
      "learning_rate": 4.986566876901272e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9423907049000263,
      "num_tokens": 27146246.0,
      "step": 3660
    },
    {
      "entropy": 1.8943463444709778,
      "epoch": 0.054912940731991745,
      "grad_norm": 0.3893425762653351,
      "learning_rate": 4.986442263418557e-05,
      "loss": 0.2591,
      "mean_token_accuracy": 0.9398513667285442,
      "num_tokens": 27219352.0,
      "step": 3670
    },
    {
      "entropy": 1.90275426954031,
      "epoch": 0.05506256727349581,
      "grad_norm": 0.3186204433441162,
      "learning_rate": 4.986317076182252e-05,
      "loss": 0.241,
      "mean_token_accuracy": 0.9398094248026609,
      "num_tokens": 27292169.0,
      "step": 3680
    },
    {
      "entropy": 1.89295661970973,
      "epoch": 0.05521219381499987,
      "grad_norm": 0.3565833270549774,
      "learning_rate": 4.986191315221305e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9422244787216186,
      "num_tokens": 27367443.0,
      "step": 3690
    },
    {
      "entropy": 1.8791549310088158,
      "epoch": 0.05536182035650394,
      "grad_norm": 0.45371511578559875,
      "learning_rate": 4.9860649805647916e-05,
      "loss": 0.241,
      "mean_token_accuracy": 0.9387322798371315,
      "num_tokens": 27442069.0,
      "step": 3700
    },
    {
      "entropy": 1.9195011377334594,
      "epoch": 0.055511446898008,
      "grad_norm": 0.3225381672382355,
      "learning_rate": 4.985938072241923e-05,
      "loss": 0.2539,
      "mean_token_accuracy": 0.9369510311633349,
      "num_tokens": 27515881.0,
      "step": 3710
    },
    {
      "entropy": 1.9253335848450661,
      "epoch": 0.05566107343951207,
      "grad_norm": 0.5452536940574646,
      "learning_rate": 4.985810590282044e-05,
      "loss": 0.2348,
      "mean_token_accuracy": 0.9448560155928135,
      "num_tokens": 27592786.0,
      "step": 3720
    },
    {
      "entropy": 1.9253989443182946,
      "epoch": 0.055810699981016135,
      "grad_norm": 0.3136405348777771,
      "learning_rate": 4.985682534714629e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.940189252793789,
      "num_tokens": 27666386.0,
      "step": 3730
    },
    {
      "entropy": 1.9281807832419873,
      "epoch": 0.0559603265225202,
      "grad_norm": 0.41742774844169617,
      "learning_rate": 4.985553905569288e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9419995874166489,
      "num_tokens": 27742428.0,
      "step": 3740
    },
    {
      "entropy": 1.9063849180936814,
      "epoch": 0.05610995306402426,
      "grad_norm": 0.31059181690216064,
      "learning_rate": 4.985424702875761e-05,
      "loss": 0.2408,
      "mean_token_accuracy": 0.940650561451912,
      "num_tokens": 27815570.0,
      "step": 3750
    },
    {
      "entropy": 1.8901608169078827,
      "epoch": 0.056259579605528326,
      "grad_norm": 0.4291262924671173,
      "learning_rate": 4.9852949266639235e-05,
      "loss": 0.2563,
      "mean_token_accuracy": 0.9391410622745753,
      "num_tokens": 27889659.0,
      "step": 3760
    },
    {
      "entropy": 1.9381004966795445,
      "epoch": 0.05640920614703239,
      "grad_norm": 0.3584326207637787,
      "learning_rate": 4.985164576963782e-05,
      "loss": 0.2587,
      "mean_token_accuracy": 0.9343846295028925,
      "num_tokens": 27961781.0,
      "step": 3770
    },
    {
      "entropy": 1.916378176957369,
      "epoch": 0.05655883268853645,
      "grad_norm": 0.37283962965011597,
      "learning_rate": 4.985033653805474e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9426307030022144,
      "num_tokens": 28036183.0,
      "step": 3780
    },
    {
      "entropy": 1.8985804244875908,
      "epoch": 0.056708459230040524,
      "grad_norm": 0.3972907066345215,
      "learning_rate": 4.9849021572192725e-05,
      "loss": 0.2573,
      "mean_token_accuracy": 0.9418021123856306,
      "num_tokens": 28109572.0,
      "step": 3790
    },
    {
      "entropy": 1.9059590615332127,
      "epoch": 0.05685808577154459,
      "grad_norm": 0.3257286250591278,
      "learning_rate": 4.984770087235582e-05,
      "loss": 0.2573,
      "mean_token_accuracy": 0.939378260076046,
      "num_tokens": 28183142.0,
      "step": 3800
    },
    {
      "entropy": 1.9020070657134056,
      "epoch": 0.05700771231304865,
      "grad_norm": 0.39726758003234863,
      "learning_rate": 4.984637443884939e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9379778765141964,
      "num_tokens": 28257370.0,
      "step": 3810
    },
    {
      "entropy": 1.840514612942934,
      "epoch": 0.057157338854552715,
      "grad_norm": 0.32359281182289124,
      "learning_rate": 4.984504227198013e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9432802062481642,
      "num_tokens": 28332368.0,
      "step": 3820
    },
    {
      "entropy": 1.869798631966114,
      "epoch": 0.05730696539605678,
      "grad_norm": 0.31587567925453186,
      "learning_rate": 4.984370437205607e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9402122475206852,
      "num_tokens": 28407100.0,
      "step": 3830
    },
    {
      "entropy": 1.8917766630649566,
      "epoch": 0.05745659193756084,
      "grad_norm": 0.40083804726600647,
      "learning_rate": 4.984236073938653e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9402970679104328,
      "num_tokens": 28480749.0,
      "step": 3840
    },
    {
      "entropy": 1.8789951778948306,
      "epoch": 0.057606218479064907,
      "grad_norm": 0.3196074366569519,
      "learning_rate": 4.984101137428221e-05,
      "loss": 0.2412,
      "mean_token_accuracy": 0.9394728094339371,
      "num_tokens": 28554765.0,
      "step": 3850
    },
    {
      "entropy": 1.8740115396678447,
      "epoch": 0.05775584502056897,
      "grad_norm": 0.3753882348537445,
      "learning_rate": 4.983965627705509e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9367282748222351,
      "num_tokens": 28629446.0,
      "step": 3860
    },
    {
      "entropy": 1.8869989313185216,
      "epoch": 0.05790547156207304,
      "grad_norm": 0.3814742863178253,
      "learning_rate": 4.983829544801851e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9452599551528692,
      "num_tokens": 28703311.0,
      "step": 3870
    },
    {
      "entropy": 1.8878202758729459,
      "epoch": 0.058055098103577105,
      "grad_norm": 0.3436293303966522,
      "learning_rate": 4.9836928887487106e-05,
      "loss": 0.2533,
      "mean_token_accuracy": 0.9412954740226269,
      "num_tokens": 28776972.0,
      "step": 3880
    },
    {
      "entropy": 1.8694807685911656,
      "epoch": 0.05820472464508117,
      "grad_norm": 0.389157235622406,
      "learning_rate": 4.9835556595776847e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9456676706671715,
      "num_tokens": 28850053.0,
      "step": 3890
    },
    {
      "entropy": 1.87982322499156,
      "epoch": 0.05835435118658523,
      "grad_norm": 0.31380313634872437,
      "learning_rate": 4.983417857320504e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9432132378220558,
      "num_tokens": 28922425.0,
      "step": 3900
    },
    {
      "entropy": 1.8590649612247945,
      "epoch": 0.058503977728089296,
      "grad_norm": 0.3637663424015045,
      "learning_rate": 4.983279482009031e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9404452182352543,
      "num_tokens": 28995340.0,
      "step": 3910
    },
    {
      "entropy": 1.8628978118300439,
      "epoch": 0.05865360426959336,
      "grad_norm": 0.35162267088890076,
      "learning_rate": 4.983140533675259e-05,
      "loss": 0.2446,
      "mean_token_accuracy": 0.9394167732447386,
      "num_tokens": 29069607.0,
      "step": 3920
    },
    {
      "entropy": 1.8369179308414458,
      "epoch": 0.05880323081109742,
      "grad_norm": 0.31461086869239807,
      "learning_rate": 4.983001012351317e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9363376252353192,
      "num_tokens": 29142999.0,
      "step": 3930
    },
    {
      "entropy": 1.8379734806716441,
      "epoch": 0.058952857352601494,
      "grad_norm": 0.3496672213077545,
      "learning_rate": 4.982860918069464e-05,
      "loss": 0.2452,
      "mean_token_accuracy": 0.9395245734602213,
      "num_tokens": 29214922.0,
      "step": 3940
    },
    {
      "entropy": 1.8410824812948703,
      "epoch": 0.05910248389410556,
      "grad_norm": 0.347578763961792,
      "learning_rate": 4.9827202508620925e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9396951891481876,
      "num_tokens": 29286862.0,
      "step": 3950
    },
    {
      "entropy": 1.8700770318508149,
      "epoch": 0.05925211043560962,
      "grad_norm": 0.35089391469955444,
      "learning_rate": 4.982579010761727e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9438132122159004,
      "num_tokens": 29360045.0,
      "step": 3960
    },
    {
      "entropy": 1.8578873306512833,
      "epoch": 0.059401736977113685,
      "grad_norm": 0.2721414268016815,
      "learning_rate": 4.982437197801025e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9455706175416708,
      "num_tokens": 29436980.0,
      "step": 3970
    },
    {
      "entropy": 1.9087005220353603,
      "epoch": 0.05955136351861775,
      "grad_norm": 0.35487714409828186,
      "learning_rate": 4.982294812012776e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9426407791674137,
      "num_tokens": 29511402.0,
      "step": 3980
    },
    {
      "entropy": 1.8990809462964535,
      "epoch": 0.05970099006012181,
      "grad_norm": 0.3005237579345703,
      "learning_rate": 4.9821518534299025e-05,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9435649175196886,
      "num_tokens": 29584291.0,
      "step": 3990
    },
    {
      "entropy": 1.906662954390049,
      "epoch": 0.059850616601625876,
      "grad_norm": 0.321585088968277,
      "learning_rate": 4.98200832208546e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9431164238601923,
      "num_tokens": 29660513.0,
      "step": 4000
    },
    {
      "epoch": 0.059850616601625876,
      "eval_entropy": 1.9211029508113862,
      "eval_loss": 0.2693643271923065,
      "eval_mean_token_accuracy": 0.9410648290872574,
      "eval_num_tokens": 29660513.0,
      "eval_runtime": 561.0424,
      "eval_samples_per_second": 35.648,
      "eval_steps_per_second": 8.912,
      "step": 4000
    },
    {
      "entropy": 1.9580626897513866,
      "epoch": 0.06000024314312995,
      "grad_norm": 0.33906257152557373,
      "learning_rate": 4.9818642180126326e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9397790495306253,
      "num_tokens": 29732049.0,
      "step": 4010
    },
    {
      "entropy": 1.9481628373265267,
      "epoch": 0.06014986968463401,
      "grad_norm": 0.3428041636943817,
      "learning_rate": 4.981719541244742e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.9414982065558434,
      "num_tokens": 29806650.0,
      "step": 4020
    },
    {
      "entropy": 1.9589499995112418,
      "epoch": 0.060299496226138075,
      "grad_norm": 0.3266701400279999,
      "learning_rate": 4.9815742918152395e-05,
      "loss": 0.2333,
      "mean_token_accuracy": 0.9427565354853868,
      "num_tokens": 29880045.0,
      "step": 4030
    },
    {
      "entropy": 2.0066395208239554,
      "epoch": 0.06044912276764214,
      "grad_norm": 0.32896706461906433,
      "learning_rate": 4.98142846975771e-05,
      "loss": 0.2527,
      "mean_token_accuracy": 0.938217180967331,
      "num_tokens": 29955598.0,
      "step": 4040
    },
    {
      "entropy": 1.9241844139993192,
      "epoch": 0.0605987493091462,
      "grad_norm": 0.3056383728981018,
      "learning_rate": 4.981282075105868e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9437720611691475,
      "num_tokens": 30032127.0,
      "step": 4050
    },
    {
      "entropy": 1.944128804653883,
      "epoch": 0.060748375850650266,
      "grad_norm": 0.3432963788509369,
      "learning_rate": 4.9811351078935646e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.939568443968892,
      "num_tokens": 30107810.0,
      "step": 4060
    },
    {
      "entropy": 1.9063711009919644,
      "epoch": 0.06089800239215433,
      "grad_norm": 0.3043891489505768,
      "learning_rate": 4.980987568154781e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9406748246401548,
      "num_tokens": 30182750.0,
      "step": 4070
    },
    {
      "entropy": 1.9449288852512836,
      "epoch": 0.0610476289336584,
      "grad_norm": 0.4303438067436218,
      "learning_rate": 4.980839455923631e-05,
      "loss": 0.2493,
      "mean_token_accuracy": 0.939748951792717,
      "num_tokens": 30254959.0,
      "step": 4080
    },
    {
      "entropy": 1.9096197232604026,
      "epoch": 0.061197255475162464,
      "grad_norm": 0.27646568417549133,
      "learning_rate": 4.98069077123436e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9440571941435337,
      "num_tokens": 30329785.0,
      "step": 4090
    },
    {
      "entropy": 1.8673910208046436,
      "epoch": 0.06134688201666653,
      "grad_norm": 0.38741955161094666,
      "learning_rate": 4.9805415141213474e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9482359036803245,
      "num_tokens": 30402091.0,
      "step": 4100
    },
    {
      "entropy": 1.9185664035379886,
      "epoch": 0.06149650855817059,
      "grad_norm": 0.3976614773273468,
      "learning_rate": 4.9803916846191034e-05,
      "loss": 0.2589,
      "mean_token_accuracy": 0.9383110739290714,
      "num_tokens": 30474609.0,
      "step": 4110
    },
    {
      "entropy": 1.922715900093317,
      "epoch": 0.061646135099674655,
      "grad_norm": 0.3446398079395294,
      "learning_rate": 4.980241282762272e-05,
      "loss": 0.2678,
      "mean_token_accuracy": 0.9369580764323473,
      "num_tokens": 30546762.0,
      "step": 4120
    },
    {
      "entropy": 1.9000771336257458,
      "epoch": 0.06179576164117872,
      "grad_norm": 0.4441876709461212,
      "learning_rate": 4.980090308585628e-05,
      "loss": 0.2552,
      "mean_token_accuracy": 0.9365449137985706,
      "num_tokens": 30619186.0,
      "step": 4130
    },
    {
      "entropy": 1.8724062703549862,
      "epoch": 0.06194538818268278,
      "grad_norm": 0.3932048976421356,
      "learning_rate": 4.979938762124079e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9402556151151658,
      "num_tokens": 30691989.0,
      "step": 4140
    },
    {
      "entropy": 1.9101918317377566,
      "epoch": 0.06209501472418685,
      "grad_norm": 0.2758844494819641,
      "learning_rate": 4.9797866434126666e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9434348590672016,
      "num_tokens": 30767718.0,
      "step": 4150
    },
    {
      "entropy": 1.9223548412322997,
      "epoch": 0.06224464126569092,
      "grad_norm": 0.2812887132167816,
      "learning_rate": 4.979633952486563e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9450450986623764,
      "num_tokens": 30843681.0,
      "step": 4160
    },
    {
      "entropy": 1.9239393167197705,
      "epoch": 0.06239426780719498,
      "grad_norm": 0.2662782669067383,
      "learning_rate": 4.979480689381072e-05,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9442126482725144,
      "num_tokens": 30918853.0,
      "step": 4170
    },
    {
      "entropy": 1.9172519482672215,
      "epoch": 0.06254389434869904,
      "grad_norm": 0.303029328584671,
      "learning_rate": 4.979326854131631e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.940716502815485,
      "num_tokens": 30993080.0,
      "step": 4180
    },
    {
      "entropy": 1.9151354059576988,
      "epoch": 0.06269352089020311,
      "grad_norm": 0.4333644509315491,
      "learning_rate": 4.979172446773812e-05,
      "loss": 0.2461,
      "mean_token_accuracy": 0.9411383211612702,
      "num_tokens": 31068735.0,
      "step": 4190
    },
    {
      "entropy": 1.9604038335382938,
      "epoch": 0.06284314743170717,
      "grad_norm": 0.4097864329814911,
      "learning_rate": 4.979017467343312e-05,
      "loss": 0.2506,
      "mean_token_accuracy": 0.9397956416010856,
      "num_tokens": 31141162.0,
      "step": 4200
    },
    {
      "entropy": 1.8810633726418018,
      "epoch": 0.06299277397321124,
      "grad_norm": 0.3743186295032501,
      "learning_rate": 4.978861915875969e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9427598867565393,
      "num_tokens": 31213978.0,
      "step": 4210
    },
    {
      "entropy": 1.8824913956224918,
      "epoch": 0.0631424005147153,
      "grad_norm": 0.2955774962902069,
      "learning_rate": 4.9787057924077484e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9435857642441988,
      "num_tokens": 31288342.0,
      "step": 4220
    },
    {
      "entropy": 1.8538796797394752,
      "epoch": 0.06329202705621936,
      "grad_norm": 0.3167761564254761,
      "learning_rate": 4.978549096974748e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9429129138588905,
      "num_tokens": 31361488.0,
      "step": 4230
    },
    {
      "entropy": 1.881327047944069,
      "epoch": 0.06344165359772343,
      "grad_norm": 0.40334591269493103,
      "learning_rate": 4.9783918296131973e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9413369856774807,
      "num_tokens": 31434544.0,
      "step": 4240
    },
    {
      "entropy": 1.8709825359284877,
      "epoch": 0.06359128013922749,
      "grad_norm": 0.27626362442970276,
      "learning_rate": 4.978233990359463e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9432210780680179,
      "num_tokens": 31507876.0,
      "step": 4250
    },
    {
      "entropy": 1.9080727517604827,
      "epoch": 0.06374090668073157,
      "grad_norm": 0.3503853380680084,
      "learning_rate": 4.978075579250037e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9407722063362598,
      "num_tokens": 31583559.0,
      "step": 4260
    },
    {
      "entropy": 1.8896555729210376,
      "epoch": 0.06389053322223563,
      "grad_norm": 0.34530365467071533,
      "learning_rate": 4.977916596321547e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9415797404944897,
      "num_tokens": 31657741.0,
      "step": 4270
    },
    {
      "entropy": 1.9040231756865977,
      "epoch": 0.0640401597637397,
      "grad_norm": 0.44108060002326965,
      "learning_rate": 4.977757041610756e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.9391418494284153,
      "num_tokens": 31730918.0,
      "step": 4280
    },
    {
      "entropy": 1.8855786994099617,
      "epoch": 0.06418978630524376,
      "grad_norm": 0.3584171533584595,
      "learning_rate": 4.977596915154551e-05,
      "loss": 0.2665,
      "mean_token_accuracy": 0.9348370790481567,
      "num_tokens": 31803315.0,
      "step": 4290
    },
    {
      "entropy": 1.9131797283887864,
      "epoch": 0.06433941284674782,
      "grad_norm": 0.3626730740070343,
      "learning_rate": 4.97743621698996e-05,
      "loss": 0.25,
      "mean_token_accuracy": 0.9406764697283506,
      "num_tokens": 31876156.0,
      "step": 4300
    },
    {
      "entropy": 1.9199877955019473,
      "epoch": 0.06448903938825189,
      "grad_norm": 0.27058082818984985,
      "learning_rate": 4.977274947154138e-05,
      "loss": 0.244,
      "mean_token_accuracy": 0.9423986829817295,
      "num_tokens": 31950991.0,
      "step": 4310
    },
    {
      "entropy": 1.8883497402071954,
      "epoch": 0.06463866592975595,
      "grad_norm": 0.3384176194667816,
      "learning_rate": 4.9771131056843724e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9450034677982331,
      "num_tokens": 32026333.0,
      "step": 4320
    },
    {
      "entropy": 1.9000013411045074,
      "epoch": 0.06478829247126001,
      "grad_norm": 0.3558318614959717,
      "learning_rate": 4.976950692618085e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9452659148722887,
      "num_tokens": 32102885.0,
      "step": 4330
    },
    {
      "entropy": 1.9208225816488267,
      "epoch": 0.06493791901276408,
      "grad_norm": 0.3226587176322937,
      "learning_rate": 4.9767877079928284e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9421998880803585,
      "num_tokens": 32174416.0,
      "step": 4340
    },
    {
      "entropy": 1.8675618380308152,
      "epoch": 0.06508754555426814,
      "grad_norm": 0.3689481317996979,
      "learning_rate": 4.9766241518462874e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.947725311666727,
      "num_tokens": 32247877.0,
      "step": 4350
    },
    {
      "entropy": 1.8768537603318691,
      "epoch": 0.0652371720957722,
      "grad_norm": 0.2900848984718323,
      "learning_rate": 4.97646002421628e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9417319141328335,
      "num_tokens": 32321876.0,
      "step": 4360
    },
    {
      "entropy": 1.907610397040844,
      "epoch": 0.06538679863727627,
      "grad_norm": 0.3418882489204407,
      "learning_rate": 4.976295325140753e-05,
      "loss": 0.2487,
      "mean_token_accuracy": 0.9372582565993071,
      "num_tokens": 32394379.0,
      "step": 4370
    },
    {
      "entropy": 1.9123635806143284,
      "epoch": 0.06553642517878033,
      "grad_norm": 0.43628475069999695,
      "learning_rate": 4.97613005465779e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9426358591765165,
      "num_tokens": 32467383.0,
      "step": 4380
    },
    {
      "entropy": 1.8834973566234112,
      "epoch": 0.0656860517202844,
      "grad_norm": 0.2998046875,
      "learning_rate": 4.975964212805605e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9394220065325498,
      "num_tokens": 32542653.0,
      "step": 4390
    },
    {
      "entropy": 1.8860047206282615,
      "epoch": 0.06583567826178846,
      "grad_norm": 0.3834998607635498,
      "learning_rate": 4.975797799622541e-05,
      "loss": 0.2393,
      "mean_token_accuracy": 0.9441626362502575,
      "num_tokens": 32616382.0,
      "step": 4400
    },
    {
      "entropy": 1.9164023026823997,
      "epoch": 0.06598530480329254,
      "grad_norm": 0.37270620465278625,
      "learning_rate": 4.9756308151470774e-05,
      "loss": 0.2462,
      "mean_token_accuracy": 0.9410018999129534,
      "num_tokens": 32691142.0,
      "step": 4410
    },
    {
      "entropy": 1.933480354398489,
      "epoch": 0.0661349313447966,
      "grad_norm": 0.3761358857154846,
      "learning_rate": 4.975463259417824e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9404208719730377,
      "num_tokens": 32767403.0,
      "step": 4420
    },
    {
      "entropy": 1.913258145749569,
      "epoch": 0.06628455788630067,
      "grad_norm": 0.337805837392807,
      "learning_rate": 4.9752951324735216e-05,
      "loss": 0.207,
      "mean_token_accuracy": 0.9482471570372581,
      "num_tokens": 32843360.0,
      "step": 4430
    },
    {
      "entropy": 1.9323188811540604,
      "epoch": 0.06643418442780473,
      "grad_norm": 0.3239274024963379,
      "learning_rate": 4.975126434353045e-05,
      "loss": 0.2511,
      "mean_token_accuracy": 0.9389578256756067,
      "num_tokens": 32919162.0,
      "step": 4440
    },
    {
      "entropy": 1.9385346911847592,
      "epoch": 0.0665838109693088,
      "grad_norm": 0.3355640470981598,
      "learning_rate": 4.9749571650954005e-05,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9428307611495257,
      "num_tokens": 32992829.0,
      "step": 4450
    },
    {
      "entropy": 1.9614437215030194,
      "epoch": 0.06673343751081286,
      "grad_norm": 0.44857582449913025,
      "learning_rate": 4.974787324739726e-05,
      "loss": 0.2411,
      "mean_token_accuracy": 0.9421608041971922,
      "num_tokens": 33063288.0,
      "step": 4460
    },
    {
      "entropy": 1.946429282426834,
      "epoch": 0.06688306405231692,
      "grad_norm": 0.35896041989326477,
      "learning_rate": 4.974616913325291e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9426769755780697,
      "num_tokens": 33136225.0,
      "step": 4470
    },
    {
      "entropy": 1.9726510286331176,
      "epoch": 0.06703269059382098,
      "grad_norm": 0.3772297501564026,
      "learning_rate": 4.974445930891498e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9419089809060097,
      "num_tokens": 33209274.0,
      "step": 4480
    },
    {
      "entropy": 1.9571030288934708,
      "epoch": 0.06718231713532505,
      "grad_norm": 0.29559776186943054,
      "learning_rate": 4.974274377477881e-05,
      "loss": 0.2485,
      "mean_token_accuracy": 0.9384094506502152,
      "num_tokens": 33286370.0,
      "step": 4490
    },
    {
      "entropy": 1.9689345449209212,
      "epoch": 0.06733194367682911,
      "grad_norm": 0.3302841782569885,
      "learning_rate": 4.9741022531241064e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9405658189207315,
      "num_tokens": 33359642.0,
      "step": 4500
    },
    {
      "entropy": 1.9485792152583599,
      "epoch": 0.06748157021833318,
      "grad_norm": 0.342250794172287,
      "learning_rate": 4.9739295578699726e-05,
      "loss": 0.2447,
      "mean_token_accuracy": 0.94135698415339,
      "num_tokens": 33432389.0,
      "step": 4510
    },
    {
      "entropy": 1.976430320739746,
      "epoch": 0.06763119675983724,
      "grad_norm": 0.3143739402294159,
      "learning_rate": 4.973756291755409e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9420891400426626,
      "num_tokens": 33506357.0,
      "step": 4520
    },
    {
      "entropy": 1.9461173094809054,
      "epoch": 0.0677808233013413,
      "grad_norm": 0.31003648042678833,
      "learning_rate": 4.9735824548204804e-05,
      "loss": 0.2345,
      "mean_token_accuracy": 0.943375562503934,
      "num_tokens": 33576244.0,
      "step": 4530
    },
    {
      "entropy": 1.9255486294627189,
      "epoch": 0.06793044984284537,
      "grad_norm": 0.32437863945961,
      "learning_rate": 4.9734080471053766e-05,
      "loss": 0.2559,
      "mean_token_accuracy": 0.936123200878501,
      "num_tokens": 33648987.0,
      "step": 4540
    },
    {
      "entropy": 1.926064107567072,
      "epoch": 0.06808007638434944,
      "grad_norm": 0.27620166540145874,
      "learning_rate": 4.973233068650427e-05,
      "loss": 0.2531,
      "mean_token_accuracy": 0.9400552168488503,
      "num_tokens": 33722361.0,
      "step": 4550
    },
    {
      "entropy": 1.954181769490242,
      "epoch": 0.06822970292585351,
      "grad_norm": 0.374869167804718,
      "learning_rate": 4.97305751949609e-05,
      "loss": 0.2529,
      "mean_token_accuracy": 0.9397334508597851,
      "num_tokens": 33796771.0,
      "step": 4560
    },
    {
      "entropy": 1.9658322371542454,
      "epoch": 0.06837932946735757,
      "grad_norm": 0.3029889762401581,
      "learning_rate": 4.972881399682953e-05,
      "loss": 0.2503,
      "mean_token_accuracy": 0.9371084406971931,
      "num_tokens": 33868061.0,
      "step": 4570
    },
    {
      "entropy": 1.9571785159409045,
      "epoch": 0.06852895600886164,
      "grad_norm": 0.3619020879268646,
      "learning_rate": 4.9727047092517406e-05,
      "loss": 0.2598,
      "mean_token_accuracy": 0.9366617169231176,
      "num_tokens": 33940960.0,
      "step": 4580
    },
    {
      "entropy": 1.9944537423551083,
      "epoch": 0.0686785825503657,
      "grad_norm": 0.3442176878452301,
      "learning_rate": 4.972527448243305e-05,
      "loss": 0.2579,
      "mean_token_accuracy": 0.9392837449908257,
      "num_tokens": 34015412.0,
      "step": 4590
    },
    {
      "entropy": 1.9306264266371727,
      "epoch": 0.06882820909186976,
      "grad_norm": 0.3249492645263672,
      "learning_rate": 4.972349616698633e-05,
      "loss": 0.2603,
      "mean_token_accuracy": 0.9352037813514471,
      "num_tokens": 34087841.0,
      "step": 4600
    },
    {
      "entropy": 1.9090164631605149,
      "epoch": 0.06897783563337383,
      "grad_norm": 0.32466474175453186,
      "learning_rate": 4.972171214658843e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9433791469782591,
      "num_tokens": 34160949.0,
      "step": 4610
    },
    {
      "entropy": 1.9303502015769483,
      "epoch": 0.06912746217487789,
      "grad_norm": 0.3015376031398773,
      "learning_rate": 4.9719922421651834e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9439161852002144,
      "num_tokens": 34236685.0,
      "step": 4620
    },
    {
      "entropy": 1.9393461249768733,
      "epoch": 0.06927708871638195,
      "grad_norm": 0.3473767936229706,
      "learning_rate": 4.971812699259037e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9448415026068687,
      "num_tokens": 34311219.0,
      "step": 4630
    },
    {
      "entropy": 1.9330279000103474,
      "epoch": 0.06942671525788602,
      "grad_norm": 0.32055577635765076,
      "learning_rate": 4.9716325859819166e-05,
      "loss": 0.2538,
      "mean_token_accuracy": 0.9369359739124775,
      "num_tokens": 34383649.0,
      "step": 4640
    },
    {
      "entropy": 1.944619643688202,
      "epoch": 0.06957634179939008,
      "grad_norm": 0.3436879515647888,
      "learning_rate": 4.971451902375468e-05,
      "loss": 0.2417,
      "mean_token_accuracy": 0.9436385996639729,
      "num_tokens": 34458687.0,
      "step": 4650
    },
    {
      "entropy": 1.918549082428217,
      "epoch": 0.06972596834089415,
      "grad_norm": 0.31978827714920044,
      "learning_rate": 4.9712706484814675e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9425349283963442,
      "num_tokens": 34534267.0,
      "step": 4660
    },
    {
      "entropy": 1.9291232705116272,
      "epoch": 0.06987559488239821,
      "grad_norm": 0.35346803069114685,
      "learning_rate": 4.9710888243418254e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9415120765566826,
      "num_tokens": 34608941.0,
      "step": 4670
    },
    {
      "entropy": 1.9249104529619216,
      "epoch": 0.07002522142390227,
      "grad_norm": 0.33946138620376587,
      "learning_rate": 4.970906429998582e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9414967425167561,
      "num_tokens": 34683019.0,
      "step": 4680
    },
    {
      "entropy": 1.9437850728631019,
      "epoch": 0.07017484796540634,
      "grad_norm": 0.2985547184944153,
      "learning_rate": 4.970723465493911e-05,
      "loss": 0.2291,
      "mean_token_accuracy": 0.9441001638770103,
      "num_tokens": 34755852.0,
      "step": 4690
    },
    {
      "entropy": 1.9494875200092792,
      "epoch": 0.07032447450691041,
      "grad_norm": 0.30250322818756104,
      "learning_rate": 4.9705399308701155e-05,
      "loss": 0.2547,
      "mean_token_accuracy": 0.9400616712868214,
      "num_tokens": 34826628.0,
      "step": 4700
    },
    {
      "entropy": 1.93246788084507,
      "epoch": 0.07047410104841448,
      "grad_norm": 0.30346161127090454,
      "learning_rate": 4.970355826169633e-05,
      "loss": 0.2509,
      "mean_token_accuracy": 0.941434844583273,
      "num_tokens": 34903610.0,
      "step": 4710
    },
    {
      "entropy": 1.979282123595476,
      "epoch": 0.07062372758991854,
      "grad_norm": 0.32545796036720276,
      "learning_rate": 4.9701711514350324e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9423569910228252,
      "num_tokens": 34976043.0,
      "step": 4720
    },
    {
      "entropy": 1.9976098485291005,
      "epoch": 0.0707733541314226,
      "grad_norm": 0.29499551653862,
      "learning_rate": 4.969985906709012e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.9449848659336567,
      "num_tokens": 35051988.0,
      "step": 4730
    },
    {
      "entropy": 1.9830034896731377,
      "epoch": 0.07092298067292667,
      "grad_norm": 0.3491571545600891,
      "learning_rate": 4.969800092034406e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9433895703405142,
      "num_tokens": 35125825.0,
      "step": 4740
    },
    {
      "entropy": 1.9787038072943688,
      "epoch": 0.07107260721443073,
      "grad_norm": 0.3343149721622467,
      "learning_rate": 4.969613707454176e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.9429921254515647,
      "num_tokens": 35202607.0,
      "step": 4750
    },
    {
      "entropy": 1.9830503724515438,
      "epoch": 0.0712222337559348,
      "grad_norm": 0.3656123876571655,
      "learning_rate": 4.969426753011418e-05,
      "loss": 0.2473,
      "mean_token_accuracy": 0.9363072425127029,
      "num_tokens": 35275357.0,
      "step": 4760
    },
    {
      "entropy": 1.9816188909113408,
      "epoch": 0.07137186029743886,
      "grad_norm": 0.39292991161346436,
      "learning_rate": 4.96923922874936e-05,
      "loss": 0.2423,
      "mean_token_accuracy": 0.9393453326076269,
      "num_tokens": 35350908.0,
      "step": 4770
    },
    {
      "entropy": 1.9509979344904422,
      "epoch": 0.07152148683894292,
      "grad_norm": 0.3487650454044342,
      "learning_rate": 4.969051134711359e-05,
      "loss": 0.2204,
      "mean_token_accuracy": 0.9481784246861935,
      "num_tokens": 35425571.0,
      "step": 4780
    },
    {
      "entropy": 1.9890504203736783,
      "epoch": 0.07167111338044699,
      "grad_norm": 0.31558161973953247,
      "learning_rate": 4.968862470940908e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9430839359760285,
      "num_tokens": 35499792.0,
      "step": 4790
    },
    {
      "entropy": 1.9698636054992675,
      "epoch": 0.07182073992195105,
      "grad_norm": 0.328294038772583,
      "learning_rate": 4.968673237481629e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9435634694993495,
      "num_tokens": 35573835.0,
      "step": 4800
    },
    {
      "entropy": 1.952013086527586,
      "epoch": 0.07197036646345512,
      "grad_norm": 0.35859546065330505,
      "learning_rate": 4.968483434377275e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9411528706550598,
      "num_tokens": 35648542.0,
      "step": 4810
    },
    {
      "entropy": 1.9332124888896942,
      "epoch": 0.07211999300495918,
      "grad_norm": 0.4262755811214447,
      "learning_rate": 4.968293061671732e-05,
      "loss": 0.246,
      "mean_token_accuracy": 0.9411147557199001,
      "num_tokens": 35720285.0,
      "step": 4820
    },
    {
      "entropy": 1.9313638255000114,
      "epoch": 0.07226961954646324,
      "grad_norm": 0.3727905750274658,
      "learning_rate": 4.9681021194090185e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9433243900537491,
      "num_tokens": 35793312.0,
      "step": 4830
    },
    {
      "entropy": 1.913079208135605,
      "epoch": 0.07241924608796732,
      "grad_norm": 0.3264462351799011,
      "learning_rate": 4.9679106076332834e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9430547960102558,
      "num_tokens": 35865904.0,
      "step": 4840
    },
    {
      "entropy": 1.9382446803152562,
      "epoch": 0.07256887262947138,
      "grad_norm": 0.3025546669960022,
      "learning_rate": 4.9677185263888065e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9424657423049212,
      "num_tokens": 35937086.0,
      "step": 4850
    },
    {
      "entropy": 1.9100962474942207,
      "epoch": 0.07271849917097545,
      "grad_norm": 0.32135671377182007,
      "learning_rate": 4.967525875720002e-05,
      "loss": 0.2345,
      "mean_token_accuracy": 0.9452652603387832,
      "num_tokens": 36010741.0,
      "step": 4860
    },
    {
      "entropy": 1.961641325056553,
      "epoch": 0.07286812571247951,
      "grad_norm": 0.34906744956970215,
      "learning_rate": 4.9673326556714135e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9447276696562767,
      "num_tokens": 36085219.0,
      "step": 4870
    },
    {
      "entropy": 1.9557662114501,
      "epoch": 0.07301775225398358,
      "grad_norm": 0.2994137108325958,
      "learning_rate": 4.967138866287717e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9427169468253851,
      "num_tokens": 36159565.0,
      "step": 4880
    },
    {
      "entropy": 1.991092646867037,
      "epoch": 0.07316737879548764,
      "grad_norm": 0.26531320810317993,
      "learning_rate": 4.96694450761372e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9408191055059433,
      "num_tokens": 36235861.0,
      "step": 4890
    },
    {
      "entropy": 1.9640290170907975,
      "epoch": 0.0733170053369917,
      "grad_norm": 0.353287011384964,
      "learning_rate": 4.966749579694362e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9409853644669056,
      "num_tokens": 36309936.0,
      "step": 4900
    },
    {
      "entropy": 1.9747645497322082,
      "epoch": 0.07346663187849577,
      "grad_norm": 0.31028926372528076,
      "learning_rate": 4.966554082574713e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9449014842510224,
      "num_tokens": 36382615.0,
      "step": 4910
    },
    {
      "entropy": 1.9454577811062337,
      "epoch": 0.07361625841999983,
      "grad_norm": 0.3836038112640381,
      "learning_rate": 4.9663580162999756e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9422901652753353,
      "num_tokens": 36456532.0,
      "step": 4920
    },
    {
      "entropy": 1.9163921058177948,
      "epoch": 0.0737658849615039,
      "grad_norm": 0.3939346373081207,
      "learning_rate": 4.966161380915485e-05,
      "loss": 0.2519,
      "mean_token_accuracy": 0.938479682803154,
      "num_tokens": 36529939.0,
      "step": 4930
    },
    {
      "entropy": 1.9082995057106018,
      "epoch": 0.07391551150300796,
      "grad_norm": 0.2836814820766449,
      "learning_rate": 4.965964176466705e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9436579316854476,
      "num_tokens": 36605787.0,
      "step": 4940
    },
    {
      "entropy": 1.9427552871406077,
      "epoch": 0.07406513804451202,
      "grad_norm": 0.34843355417251587,
      "learning_rate": 4.965766402999234e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9426591526716948,
      "num_tokens": 36677430.0,
      "step": 4950
    },
    {
      "entropy": 1.9466373950242997,
      "epoch": 0.07421476458601609,
      "grad_norm": 0.3679488003253937,
      "learning_rate": 4.9655680605588006e-05,
      "loss": 0.2479,
      "mean_token_accuracy": 0.9395054552704096,
      "num_tokens": 36749568.0,
      "step": 4960
    },
    {
      "entropy": 1.9394116386771203,
      "epoch": 0.07436439112752015,
      "grad_norm": 0.4099727272987366,
      "learning_rate": 4.965369149191265e-05,
      "loss": 0.2449,
      "mean_token_accuracy": 0.9420007590204478,
      "num_tokens": 36821180.0,
      "step": 4970
    },
    {
      "entropy": 1.97816614061594,
      "epoch": 0.07451401766902423,
      "grad_norm": 0.4656299352645874,
      "learning_rate": 4.9651696689426184e-05,
      "loss": 0.2708,
      "mean_token_accuracy": 0.9330280151218175,
      "num_tokens": 36893246.0,
      "step": 4980
    },
    {
      "entropy": 2.0031517200171947,
      "epoch": 0.07466364421052829,
      "grad_norm": 0.37220248579978943,
      "learning_rate": 4.964969619858986e-05,
      "loss": 0.2537,
      "mean_token_accuracy": 0.9383817665278912,
      "num_tokens": 36967194.0,
      "step": 4990
    },
    {
      "entropy": 1.942775446921587,
      "epoch": 0.07481327075203235,
      "grad_norm": 0.40330421924591064,
      "learning_rate": 4.96476900198662e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9421227253973484,
      "num_tokens": 37042692.0,
      "step": 5000
    },
    {
      "epoch": 0.07481327075203235,
      "eval_entropy": 1.9384293358087539,
      "eval_loss": 0.26564010977745056,
      "eval_mean_token_accuracy": 0.941402601313591,
      "eval_num_tokens": 37042692.0,
      "eval_runtime": 558.9942,
      "eval_samples_per_second": 35.779,
      "eval_steps_per_second": 8.945,
      "step": 5000
    },
    {
      "entropy": 1.9523049749433994,
      "epoch": 0.07496289729353642,
      "grad_norm": 0.37337055802345276,
      "learning_rate": 4.964567815371908e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.9407647088170051,
      "num_tokens": 37114357.0,
      "step": 5010
    },
    {
      "entropy": 1.9687109090387822,
      "epoch": 0.07511252383504048,
      "grad_norm": 0.3207288086414337,
      "learning_rate": 4.964366060061369e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.9407301295548678,
      "num_tokens": 37188298.0,
      "step": 5020
    },
    {
      "entropy": 1.9678151339292527,
      "epoch": 0.07526215037654455,
      "grad_norm": 0.31629714369773865,
      "learning_rate": 4.964163736101652e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9395524870604277,
      "num_tokens": 37262808.0,
      "step": 5030
    },
    {
      "entropy": 1.954004179686308,
      "epoch": 0.07541177691804861,
      "grad_norm": 0.31784629821777344,
      "learning_rate": 4.963960843539537e-05,
      "loss": 0.2534,
      "mean_token_accuracy": 0.9376971654593944,
      "num_tokens": 37336289.0,
      "step": 5040
    },
    {
      "entropy": 1.9743161544203758,
      "epoch": 0.07556140345955267,
      "grad_norm": 0.3836459219455719,
      "learning_rate": 4.9637573824219375e-05,
      "loss": 0.2412,
      "mean_token_accuracy": 0.9393380101770162,
      "num_tokens": 37411951.0,
      "step": 5050
    },
    {
      "entropy": 1.9894124783575535,
      "epoch": 0.07571103000105674,
      "grad_norm": 0.3423228859901428,
      "learning_rate": 4.963553352795896e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9398914355784654,
      "num_tokens": 37485616.0,
      "step": 5060
    },
    {
      "entropy": 1.9735226832330226,
      "epoch": 0.0758606565425608,
      "grad_norm": 0.3276052474975586,
      "learning_rate": 4.963348754708589e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9390364356338978,
      "num_tokens": 37559606.0,
      "step": 5070
    },
    {
      "entropy": 1.9659298650920392,
      "epoch": 0.07601028308406486,
      "grad_norm": 0.41879716515541077,
      "learning_rate": 4.963143588207322e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9455731365829706,
      "num_tokens": 37634845.0,
      "step": 5080
    },
    {
      "entropy": 1.9751592963933944,
      "epoch": 0.07615990962556893,
      "grad_norm": 0.3498132526874542,
      "learning_rate": 4.9629378533395345e-05,
      "loss": 0.247,
      "mean_token_accuracy": 0.9400275968015194,
      "num_tokens": 37708543.0,
      "step": 5090
    },
    {
      "entropy": 1.9732576176524161,
      "epoch": 0.07630953616707299,
      "grad_norm": 0.3235454261302948,
      "learning_rate": 4.962731550152795e-05,
      "loss": 0.2417,
      "mean_token_accuracy": 0.9391399990767241,
      "num_tokens": 37783198.0,
      "step": 5100
    },
    {
      "entropy": 2.002301864326,
      "epoch": 0.07645916270857706,
      "grad_norm": 0.3621864318847656,
      "learning_rate": 4.962524678694804e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9401468854397536,
      "num_tokens": 37856991.0,
      "step": 5110
    },
    {
      "entropy": 1.989931981265545,
      "epoch": 0.07660878925008112,
      "grad_norm": 0.3116135895252228,
      "learning_rate": 4.962317239013396e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9416136592626572,
      "num_tokens": 37929877.0,
      "step": 5120
    },
    {
      "entropy": 2.010821530967951,
      "epoch": 0.0767584157915852,
      "grad_norm": 0.3892084062099457,
      "learning_rate": 4.962109231156534e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9446036618202924,
      "num_tokens": 38000683.0,
      "step": 5130
    },
    {
      "entropy": 1.9883898489177227,
      "epoch": 0.07690804233308926,
      "grad_norm": 0.3077300190925598,
      "learning_rate": 4.961900655172311e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9417524494230747,
      "num_tokens": 38071699.0,
      "step": 5140
    },
    {
      "entropy": 2.0000866375863553,
      "epoch": 0.07705766887459332,
      "grad_norm": 0.3600141704082489,
      "learning_rate": 4.961691511108956e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9416221875697375,
      "num_tokens": 38146040.0,
      "step": 5150
    },
    {
      "entropy": 1.9977682903409004,
      "epoch": 0.07720729541609739,
      "grad_norm": 0.33522486686706543,
      "learning_rate": 4.9614817990148255e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.9443767491728068,
      "num_tokens": 38218277.0,
      "step": 5160
    },
    {
      "entropy": 1.9936495140194892,
      "epoch": 0.07735692195760145,
      "grad_norm": 0.332469642162323,
      "learning_rate": 4.961271518938409e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9428583584725857,
      "num_tokens": 38291332.0,
      "step": 5170
    },
    {
      "entropy": 1.9607647128403187,
      "epoch": 0.07750654849910552,
      "grad_norm": 0.34432587027549744,
      "learning_rate": 4.961060670928328e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.9383972018957139,
      "num_tokens": 38364775.0,
      "step": 5180
    },
    {
      "entropy": 1.9823058605194093,
      "epoch": 0.07765617504060958,
      "grad_norm": 0.33515360951423645,
      "learning_rate": 4.960849255033333e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9448011297732591,
      "num_tokens": 38437937.0,
      "step": 5190
    },
    {
      "entropy": 1.9763479329645635,
      "epoch": 0.07780580158211364,
      "grad_norm": 0.29226094484329224,
      "learning_rate": 4.960637271302307e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9391347020864487,
      "num_tokens": 38512553.0,
      "step": 5200
    },
    {
      "entropy": 1.9511084653437138,
      "epoch": 0.0779554281236177,
      "grad_norm": 0.361581027507782,
      "learning_rate": 4.9604247197842655e-05,
      "loss": 0.2326,
      "mean_token_accuracy": 0.9453748978674412,
      "num_tokens": 38585387.0,
      "step": 5210
    },
    {
      "entropy": 1.9513887643814087,
      "epoch": 0.07810505466512177,
      "grad_norm": 0.35784319043159485,
      "learning_rate": 4.960211600528353e-05,
      "loss": 0.2395,
      "mean_token_accuracy": 0.9431748818606138,
      "num_tokens": 38661364.0,
      "step": 5220
    },
    {
      "entropy": 1.952230542898178,
      "epoch": 0.07825468120662583,
      "grad_norm": 0.34889718890190125,
      "learning_rate": 4.959997913583847e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9445189233869314,
      "num_tokens": 38733358.0,
      "step": 5230
    },
    {
      "entropy": 1.9217127703130246,
      "epoch": 0.0784043077481299,
      "grad_norm": 0.36322370171546936,
      "learning_rate": 4.9597836590001565e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9428963247686625,
      "num_tokens": 38808135.0,
      "step": 5240
    },
    {
      "entropy": 1.9605079390108586,
      "epoch": 0.07855393428963396,
      "grad_norm": 0.34608060121536255,
      "learning_rate": 4.95956883682682e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9405039887875318,
      "num_tokens": 38882958.0,
      "step": 5250
    },
    {
      "entropy": 1.9740474626421929,
      "epoch": 0.07870356083113803,
      "grad_norm": 0.35976243019104004,
      "learning_rate": 4.959353447113508e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9449405457824469,
      "num_tokens": 38956921.0,
      "step": 5260
    },
    {
      "entropy": 1.9484592780470849,
      "epoch": 0.0788531873726421,
      "grad_norm": 0.26556944847106934,
      "learning_rate": 4.959137489910023e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.943807503581047,
      "num_tokens": 39031270.0,
      "step": 5270
    },
    {
      "entropy": 2.014249498397112,
      "epoch": 0.07900281391414617,
      "grad_norm": 0.46005508303642273,
      "learning_rate": 4.958920965266298e-05,
      "loss": 0.2594,
      "mean_token_accuracy": 0.9358900990337133,
      "num_tokens": 39101910.0,
      "step": 5280
    },
    {
      "entropy": 2.0210400484502316,
      "epoch": 0.07915244045565023,
      "grad_norm": 0.33988428115844727,
      "learning_rate": 4.958703873232397e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9395722724497318,
      "num_tokens": 39176595.0,
      "step": 5290
    },
    {
      "entropy": 2.0097762353718283,
      "epoch": 0.0793020669971543,
      "grad_norm": 0.39646950364112854,
      "learning_rate": 4.958486213858517e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.937994945794344,
      "num_tokens": 39251376.0,
      "step": 5300
    },
    {
      "entropy": 2.003940711170435,
      "epoch": 0.07945169353865836,
      "grad_norm": 0.3544342517852783,
      "learning_rate": 4.958267987194983e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9402030665427447,
      "num_tokens": 39324442.0,
      "step": 5310
    },
    {
      "entropy": 2.0033877596259115,
      "epoch": 0.07960132008016242,
      "grad_norm": 0.31247514486312866,
      "learning_rate": 4.9580491932922535e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9386104423552751,
      "num_tokens": 39395823.0,
      "step": 5320
    },
    {
      "entropy": 2.0228067748248577,
      "epoch": 0.07975094662166649,
      "grad_norm": 0.37067052721977234,
      "learning_rate": 4.957829832200918e-05,
      "loss": 0.251,
      "mean_token_accuracy": 0.9381369393318891,
      "num_tokens": 39468456.0,
      "step": 5330
    },
    {
      "entropy": 2.0012464843690396,
      "epoch": 0.07990057316317055,
      "grad_norm": 0.3133704364299774,
      "learning_rate": 4.957609903971695e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9442806579172611,
      "num_tokens": 39543603.0,
      "step": 5340
    },
    {
      "entropy": 2.020212238281965,
      "epoch": 0.08005019970467461,
      "grad_norm": 0.3086543381214142,
      "learning_rate": 4.957389408655439e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9423615679144859,
      "num_tokens": 39620108.0,
      "step": 5350
    },
    {
      "entropy": 2.020922037214041,
      "epoch": 0.08019982624617868,
      "grad_norm": 0.3234139680862427,
      "learning_rate": 4.957168346303129e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.937639844045043,
      "num_tokens": 39693639.0,
      "step": 5360
    },
    {
      "entropy": 2.026205163449049,
      "epoch": 0.08034945278768274,
      "grad_norm": 0.3466288447380066,
      "learning_rate": 4.9569467169658805e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.946575828641653,
      "num_tokens": 39769108.0,
      "step": 5370
    },
    {
      "entropy": 2.0180299304425717,
      "epoch": 0.0804990793291868,
      "grad_norm": 0.37291425466537476,
      "learning_rate": 4.956724520694938e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9410816390067339,
      "num_tokens": 39841327.0,
      "step": 5380
    },
    {
      "entropy": 1.9986849114298821,
      "epoch": 0.08064870587069087,
      "grad_norm": 0.3366378843784332,
      "learning_rate": 4.9565017575416766e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.9397193849086761,
      "num_tokens": 39913557.0,
      "step": 5390
    },
    {
      "entropy": 1.9762165531516076,
      "epoch": 0.08079833241219493,
      "grad_norm": 0.3124225437641144,
      "learning_rate": 4.956278427557603e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9442578006535769,
      "num_tokens": 39988785.0,
      "step": 5400
    },
    {
      "entropy": 2.019265276193619,
      "epoch": 0.080947958953699,
      "grad_norm": 0.42172354459762573,
      "learning_rate": 4.956054530794357e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.9396452262997628,
      "num_tokens": 40062968.0,
      "step": 5410
    },
    {
      "entropy": 2.013544649630785,
      "epoch": 0.08109758549520307,
      "grad_norm": 0.3214823305606842,
      "learning_rate": 4.955830067303705e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9432305067777633,
      "num_tokens": 40136887.0,
      "step": 5420
    },
    {
      "entropy": 2.024566163122654,
      "epoch": 0.08124721203670714,
      "grad_norm": 0.3532926142215729,
      "learning_rate": 4.955605037137548e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.942586112394929,
      "num_tokens": 40209712.0,
      "step": 5430
    },
    {
      "entropy": 2.0255594074726107,
      "epoch": 0.0813968385782112,
      "grad_norm": 0.41796180605888367,
      "learning_rate": 4.955379440347917e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9389738377183676,
      "num_tokens": 40286893.0,
      "step": 5440
    },
    {
      "entropy": 2.0882167518138885,
      "epoch": 0.08154646511971526,
      "grad_norm": 0.34808599948883057,
      "learning_rate": 4.955153276986973e-05,
      "loss": 0.236,
      "mean_token_accuracy": 0.9430191822350025,
      "num_tokens": 40361321.0,
      "step": 5450
    },
    {
      "entropy": 2.1006139114499094,
      "epoch": 0.08169609166121933,
      "grad_norm": 0.30729398131370544,
      "learning_rate": 4.954926547107012e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.943875466287136,
      "num_tokens": 40435218.0,
      "step": 5460
    },
    {
      "entropy": 2.0789404205977915,
      "epoch": 0.08184571820272339,
      "grad_norm": 0.325589656829834,
      "learning_rate": 4.954699250760455e-05,
      "loss": 0.2497,
      "mean_token_accuracy": 0.9384634628891945,
      "num_tokens": 40508045.0,
      "step": 5470
    },
    {
      "entropy": 2.0723202802240848,
      "epoch": 0.08199534474422746,
      "grad_norm": 0.3058010935783386,
      "learning_rate": 4.9544713879998564e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9395518653094769,
      "num_tokens": 40582044.0,
      "step": 5480
    },
    {
      "entropy": 2.0372195966541766,
      "epoch": 0.08214497128573152,
      "grad_norm": 0.3490447998046875,
      "learning_rate": 4.954242958877905e-05,
      "loss": 0.2474,
      "mean_token_accuracy": 0.9405687320977449,
      "num_tokens": 40656317.0,
      "step": 5490
    },
    {
      "entropy": 2.062868182361126,
      "epoch": 0.08229459782723558,
      "grad_norm": 0.3289259970188141,
      "learning_rate": 4.954013963447416e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9369831923395395,
      "num_tokens": 40729995.0,
      "step": 5500
    },
    {
      "entropy": 2.048599060624838,
      "epoch": 0.08244422436873965,
      "grad_norm": 0.4003506898880005,
      "learning_rate": 4.953784401761337e-05,
      "loss": 0.262,
      "mean_token_accuracy": 0.9353604279458523,
      "num_tokens": 40801691.0,
      "step": 5510
    },
    {
      "entropy": 2.0466851241886617,
      "epoch": 0.08259385091024371,
      "grad_norm": 0.41236403584480286,
      "learning_rate": 4.953554273872748e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9421020530164241,
      "num_tokens": 40874001.0,
      "step": 5520
    },
    {
      "entropy": 2.0186397559940814,
      "epoch": 0.08274347745174777,
      "grad_norm": 0.344412624835968,
      "learning_rate": 4.9533235798348574e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9436346687376499,
      "num_tokens": 40947083.0,
      "step": 5530
    },
    {
      "entropy": 1.9933625869452953,
      "epoch": 0.08289310399325184,
      "grad_norm": 0.31085333228111267,
      "learning_rate": 4.953092319701007e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9434937864542008,
      "num_tokens": 41027142.0,
      "step": 5540
    },
    {
      "entropy": 2.0093927189707754,
      "epoch": 0.0830427305347559,
      "grad_norm": 0.3440219759941101,
      "learning_rate": 4.952860493524667e-05,
      "loss": 0.2502,
      "mean_token_accuracy": 0.9372138548642397,
      "num_tokens": 41098733.0,
      "step": 5550
    },
    {
      "entropy": 2.027844003587961,
      "epoch": 0.08319235707625998,
      "grad_norm": 0.34315750002861023,
      "learning_rate": 4.95262810135944e-05,
      "loss": 0.2539,
      "mean_token_accuracy": 0.9384838346391916,
      "num_tokens": 41169492.0,
      "step": 5560
    },
    {
      "entropy": 2.0327657498419285,
      "epoch": 0.08334198361776404,
      "grad_norm": 0.39672374725341797,
      "learning_rate": 4.95239514325906e-05,
      "loss": 0.2512,
      "mean_token_accuracy": 0.9384131621569395,
      "num_tokens": 41240275.0,
      "step": 5570
    },
    {
      "entropy": 1.9282449886202813,
      "epoch": 0.0834916101592681,
      "grad_norm": 0.398840069770813,
      "learning_rate": 4.95216161927739e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.942188385128975,
      "num_tokens": 41317339.0,
      "step": 5580
    },
    {
      "entropy": 1.9656348302960396,
      "epoch": 0.08364123670077217,
      "grad_norm": 0.3682236075401306,
      "learning_rate": 4.951927529468426e-05,
      "loss": 0.2354,
      "mean_token_accuracy": 0.9423908315598964,
      "num_tokens": 41391614.0,
      "step": 5590
    },
    {
      "entropy": 2.007145520299673,
      "epoch": 0.08379086324227623,
      "grad_norm": 0.35073068737983704,
      "learning_rate": 4.951692873886292e-05,
      "loss": 0.2521,
      "mean_token_accuracy": 0.9389020580798387,
      "num_tokens": 41464448.0,
      "step": 5600
    },
    {
      "entropy": 2.042485027760267,
      "epoch": 0.0839404897837803,
      "grad_norm": 0.4515027701854706,
      "learning_rate": 4.951457652585246e-05,
      "loss": 0.244,
      "mean_token_accuracy": 0.9409322731196881,
      "num_tokens": 41537142.0,
      "step": 5610
    },
    {
      "entropy": 2.037228162586689,
      "epoch": 0.08409011632528436,
      "grad_norm": 0.3204001188278198,
      "learning_rate": 4.951221865619676e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.9387986201792955,
      "num_tokens": 41612102.0,
      "step": 5620
    },
    {
      "entropy": 2.0044845595955847,
      "epoch": 0.08423974286678843,
      "grad_norm": 0.2963782250881195,
      "learning_rate": 4.950985513044098e-05,
      "loss": 0.2613,
      "mean_token_accuracy": 0.9353123210370541,
      "num_tokens": 41687261.0,
      "step": 5630
    },
    {
      "entropy": 2.0031675696372986,
      "epoch": 0.08438936940829249,
      "grad_norm": 0.357308030128479,
      "learning_rate": 4.9507485949131635e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9395338408648968,
      "num_tokens": 41757949.0,
      "step": 5640
    },
    {
      "entropy": 1.974797598272562,
      "epoch": 0.08453899594979655,
      "grad_norm": 0.30298322439193726,
      "learning_rate": 4.950511111281649e-05,
      "loss": 0.2299,
      "mean_token_accuracy": 0.9421533238142729,
      "num_tokens": 41834258.0,
      "step": 5650
    },
    {
      "entropy": 1.9879207871854305,
      "epoch": 0.08468862249130062,
      "grad_norm": 0.4356994330883026,
      "learning_rate": 4.950273062204468e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.9419903133064509,
      "num_tokens": 41908789.0,
      "step": 5660
    },
    {
      "entropy": 2.0031319215893744,
      "epoch": 0.08483824903280468,
      "grad_norm": 0.29541149735450745,
      "learning_rate": 4.95003444773666e-05,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9462953235954046,
      "num_tokens": 41983675.0,
      "step": 5670
    },
    {
      "entropy": 2.071229290217161,
      "epoch": 0.08498787557430874,
      "grad_norm": 0.29785752296447754,
      "learning_rate": 4.949795267933398e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9403899263590574,
      "num_tokens": 42057367.0,
      "step": 5680
    },
    {
      "entropy": 2.0707533314824103,
      "epoch": 0.08513750211581281,
      "grad_norm": 0.3975188434123993,
      "learning_rate": 4.949555522849983e-05,
      "loss": 0.2586,
      "mean_token_accuracy": 0.9354882039129734,
      "num_tokens": 42132310.0,
      "step": 5690
    },
    {
      "entropy": 2.0729260869324206,
      "epoch": 0.08528712865731687,
      "grad_norm": 0.3207266926765442,
      "learning_rate": 4.949315212541849e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9430180795490741,
      "num_tokens": 42206105.0,
      "step": 5700
    },
    {
      "entropy": 2.0568599477410316,
      "epoch": 0.08543675519882095,
      "grad_norm": 0.3376027047634125,
      "learning_rate": 4.949074337064561e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9385559841990471,
      "num_tokens": 42278607.0,
      "step": 5710
    },
    {
      "entropy": 2.0697998724877835,
      "epoch": 0.08558638174032501,
      "grad_norm": 0.3244185447692871,
      "learning_rate": 4.948832896473813e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.9413168095052242,
      "num_tokens": 42353702.0,
      "step": 5720
    },
    {
      "entropy": 2.0488916143774984,
      "epoch": 0.08573600828182908,
      "grad_norm": 0.3531823456287384,
      "learning_rate": 4.9485908908254295e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.9425504431128502,
      "num_tokens": 42430596.0,
      "step": 5730
    },
    {
      "entropy": 2.0975300915539266,
      "epoch": 0.08588563482333314,
      "grad_norm": 0.39805126190185547,
      "learning_rate": 4.948348320175368e-05,
      "loss": 0.2475,
      "mean_token_accuracy": 0.9424961566925049,
      "num_tokens": 42503367.0,
      "step": 5740
    },
    {
      "entropy": 2.107554052770138,
      "epoch": 0.0860352613648372,
      "grad_norm": 0.41154301166534424,
      "learning_rate": 4.948105184579714e-05,
      "loss": 0.251,
      "mean_token_accuracy": 0.9398801486939192,
      "num_tokens": 42574116.0,
      "step": 5750
    },
    {
      "entropy": 2.051720730215311,
      "epoch": 0.08618488790634127,
      "grad_norm": 0.3699057698249817,
      "learning_rate": 4.947861484094686e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9432928927242756,
      "num_tokens": 42648399.0,
      "step": 5760
    },
    {
      "entropy": 2.0431001022458077,
      "epoch": 0.08633451444784533,
      "grad_norm": 0.3378255367279053,
      "learning_rate": 4.947617218776631e-05,
      "loss": 0.2443,
      "mean_token_accuracy": 0.9411807004362345,
      "num_tokens": 42722007.0,
      "step": 5770
    },
    {
      "entropy": 2.063131057471037,
      "epoch": 0.0864841409893494,
      "grad_norm": 0.3617124557495117,
      "learning_rate": 4.947372388682027e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.9455738935619593,
      "num_tokens": 42797193.0,
      "step": 5780
    },
    {
      "entropy": 2.038436956703663,
      "epoch": 0.08663376753085346,
      "grad_norm": 0.3896268606185913,
      "learning_rate": 4.947126993867485e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9403161965310574,
      "num_tokens": 42871297.0,
      "step": 5790
    },
    {
      "entropy": 2.0094648763537406,
      "epoch": 0.08678339407235752,
      "grad_norm": 0.314296692609787,
      "learning_rate": 4.946881034389741e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9473994500935078,
      "num_tokens": 42946020.0,
      "step": 5800
    },
    {
      "entropy": 2.0251397870481016,
      "epoch": 0.08693302061386159,
      "grad_norm": 0.3231901228427887,
      "learning_rate": 4.94663451030567e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.9396975059062243,
      "num_tokens": 43017642.0,
      "step": 5810
    },
    {
      "entropy": 2.0104017682373523,
      "epoch": 0.08708264715536565,
      "grad_norm": 0.3585280478000641,
      "learning_rate": 4.946387421672269e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9434605691581964,
      "num_tokens": 43093091.0,
      "step": 5820
    },
    {
      "entropy": 2.0121307387948035,
      "epoch": 0.08723227369686971,
      "grad_norm": 0.3292883634567261,
      "learning_rate": 4.946139768546671e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9444511808454991,
      "num_tokens": 43166204.0,
      "step": 5830
    },
    {
      "entropy": 1.9911436684429646,
      "epoch": 0.08738190023837378,
      "grad_norm": 0.34359923005104065,
      "learning_rate": 4.945891550986137e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9435492575168609,
      "num_tokens": 43238420.0,
      "step": 5840
    },
    {
      "entropy": 2.0067301645874975,
      "epoch": 0.08753152677987786,
      "grad_norm": 0.4983045160770416,
      "learning_rate": 4.94564276904806e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9414449829608202,
      "num_tokens": 43313216.0,
      "step": 5850
    },
    {
      "entropy": 1.9915693342685699,
      "epoch": 0.08768115332138192,
      "grad_norm": 0.43152257800102234,
      "learning_rate": 4.9453934227899626e-05,
      "loss": 0.2236,
      "mean_token_accuracy": 0.943453136086464,
      "num_tokens": 43387416.0,
      "step": 5860
    },
    {
      "entropy": 1.9867277398705483,
      "epoch": 0.08783077986288598,
      "grad_norm": 0.3129936754703522,
      "learning_rate": 4.9451435122694973e-05,
      "loss": 0.2616,
      "mean_token_accuracy": 0.9379395052790642,
      "num_tokens": 43461449.0,
      "step": 5870
    },
    {
      "entropy": 2.008326721936464,
      "epoch": 0.08798040640439005,
      "grad_norm": 0.42647117376327515,
      "learning_rate": 4.94489303754445e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9424808446317912,
      "num_tokens": 43535580.0,
      "step": 5880
    },
    {
      "entropy": 2.0069543428719046,
      "epoch": 0.08813003294589411,
      "grad_norm": 0.3329504728317261,
      "learning_rate": 4.944641998672732e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9424146007746458,
      "num_tokens": 43613174.0,
      "step": 5890
    },
    {
      "entropy": 2.014348681271076,
      "epoch": 0.08827965948739817,
      "grad_norm": 0.3741479218006134,
      "learning_rate": 4.9443903957123904e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9430493362247944,
      "num_tokens": 43688837.0,
      "step": 5900
    },
    {
      "entropy": 1.9829587318003177,
      "epoch": 0.08842928602890224,
      "grad_norm": 0.3121398091316223,
      "learning_rate": 4.944138228721599e-05,
      "loss": 0.2415,
      "mean_token_accuracy": 0.9438747577369213,
      "num_tokens": 43762657.0,
      "step": 5910
    },
    {
      "entropy": 1.9873070389032363,
      "epoch": 0.0885789125704063,
      "grad_norm": 0.30037862062454224,
      "learning_rate": 4.943885497758664e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9449743703007698,
      "num_tokens": 43837696.0,
      "step": 5920
    },
    {
      "entropy": 2.0126344323158265,
      "epoch": 0.08872853911191037,
      "grad_norm": 0.32163476943969727,
      "learning_rate": 4.943632202882021e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9391754742711782,
      "num_tokens": 43912251.0,
      "step": 5930
    },
    {
      "entropy": 2.0232132084667684,
      "epoch": 0.08887816565341443,
      "grad_norm": 0.2829095721244812,
      "learning_rate": 4.943378344150235e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9450729068368673,
      "num_tokens": 43991186.0,
      "step": 5940
    },
    {
      "entropy": 2.000468822568655,
      "epoch": 0.08902779219491849,
      "grad_norm": 0.42535385489463806,
      "learning_rate": 4.943123921622004e-05,
      "loss": 0.2643,
      "mean_token_accuracy": 0.9356313236057758,
      "num_tokens": 44063806.0,
      "step": 5950
    },
    {
      "entropy": 1.9749486148357391,
      "epoch": 0.08917741873642256,
      "grad_norm": 0.38431838154792786,
      "learning_rate": 4.942868935356156e-05,
      "loss": 0.2427,
      "mean_token_accuracy": 0.945241916552186,
      "num_tokens": 44136549.0,
      "step": 5960
    },
    {
      "entropy": 2.006798943132162,
      "epoch": 0.08932704527792662,
      "grad_norm": 0.31501495838165283,
      "learning_rate": 4.942613385411646e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9401999864727258,
      "num_tokens": 44208111.0,
      "step": 5970
    },
    {
      "entropy": 2.007431797683239,
      "epoch": 0.08947667181943068,
      "grad_norm": 0.3073079288005829,
      "learning_rate": 4.942357271847565e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.943475567176938,
      "num_tokens": 44283154.0,
      "step": 5980
    },
    {
      "entropy": 2.0305110827088355,
      "epoch": 0.08962629836093476,
      "grad_norm": 0.3918016254901886,
      "learning_rate": 4.9421005947231274e-05,
      "loss": 0.2546,
      "mean_token_accuracy": 0.9383151553571224,
      "num_tokens": 44358057.0,
      "step": 5990
    },
    {
      "entropy": 1.9553543254733086,
      "epoch": 0.08977592490243883,
      "grad_norm": 0.3829839527606964,
      "learning_rate": 4.941843354097682e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9453389849513769,
      "num_tokens": 44432302.0,
      "step": 6000
    },
    {
      "epoch": 0.08977592490243883,
      "eval_entropy": 1.9293144409894942,
      "eval_loss": 0.26288577914237976,
      "eval_mean_token_accuracy": 0.9420813955187798,
      "eval_num_tokens": 44432302.0,
      "eval_runtime": 560.6707,
      "eval_samples_per_second": 35.672,
      "eval_steps_per_second": 8.918,
      "step": 6000
    },
    {
      "entropy": 1.9387284755706786,
      "epoch": 0.08992555144394289,
      "grad_norm": 0.32174190878868103,
      "learning_rate": 4.9415855500307104e-05,
      "loss": 0.2346,
      "mean_token_accuracy": 0.9433124139904976,
      "num_tokens": 44506534.0,
      "step": 6010
    },
    {
      "entropy": 1.9512828953564167,
      "epoch": 0.09007517798544695,
      "grad_norm": 0.2995511293411255,
      "learning_rate": 4.9413271825818184e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9435870587825775,
      "num_tokens": 44583415.0,
      "step": 6020
    },
    {
      "entropy": 1.9721534483134746,
      "epoch": 0.09022480452695102,
      "grad_norm": 0.33534887433052063,
      "learning_rate": 4.941068251810747e-05,
      "loss": 0.2415,
      "mean_token_accuracy": 0.9367950152605772,
      "num_tokens": 44656778.0,
      "step": 6030
    },
    {
      "entropy": 1.9898998245596886,
      "epoch": 0.09037443106845508,
      "grad_norm": 0.33429473638534546,
      "learning_rate": 4.940808757777364e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9434578020125628,
      "num_tokens": 44731491.0,
      "step": 6040
    },
    {
      "entropy": 1.9694059766829013,
      "epoch": 0.09052405760995914,
      "grad_norm": 0.375367134809494,
      "learning_rate": 4.9405487005416706e-05,
      "loss": 0.2519,
      "mean_token_accuracy": 0.9395961355417967,
      "num_tokens": 44802695.0,
      "step": 6050
    },
    {
      "entropy": 1.982146967202425,
      "epoch": 0.09067368415146321,
      "grad_norm": 0.30514591932296753,
      "learning_rate": 4.940288080163796e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.9415478222072124,
      "num_tokens": 44876832.0,
      "step": 6060
    },
    {
      "entropy": 1.985879722237587,
      "epoch": 0.09082331069296727,
      "grad_norm": 0.38759467005729675,
      "learning_rate": 4.940026896704e-05,
      "loss": 0.2573,
      "mean_token_accuracy": 0.9363418877124786,
      "num_tokens": 44949594.0,
      "step": 6070
    },
    {
      "entropy": 1.985973409563303,
      "epoch": 0.09097293723447133,
      "grad_norm": 0.34681272506713867,
      "learning_rate": 4.939765150222673e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.9444064289331436,
      "num_tokens": 45023824.0,
      "step": 6080
    },
    {
      "entropy": 1.9433952882885932,
      "epoch": 0.0911225637759754,
      "grad_norm": 0.2809903025627136,
      "learning_rate": 4.9395028407803356e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9462559569627047,
      "num_tokens": 45097998.0,
      "step": 6090
    },
    {
      "entropy": 1.9380916833877564,
      "epoch": 0.09127219031747946,
      "grad_norm": 0.33656659722328186,
      "learning_rate": 4.9392399684376386e-05,
      "loss": 0.2045,
      "mean_token_accuracy": 0.94820068590343,
      "num_tokens": 45175452.0,
      "step": 6100
    },
    {
      "entropy": 1.9890604570508004,
      "epoch": 0.09142181685898353,
      "grad_norm": 0.4189107418060303,
      "learning_rate": 4.9389765332553616e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9437408000230789,
      "num_tokens": 45247527.0,
      "step": 6110
    },
    {
      "entropy": 1.990535955131054,
      "epoch": 0.09157144340048759,
      "grad_norm": 0.4302513897418976,
      "learning_rate": 4.938712535294417e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9424699287861585,
      "num_tokens": 45319616.0,
      "step": 6120
    },
    {
      "entropy": 1.9907277204096316,
      "epoch": 0.09172106994199165,
      "grad_norm": 0.38890865445137024,
      "learning_rate": 4.9384479746158456e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.943159457668662,
      "num_tokens": 45395938.0,
      "step": 6130
    },
    {
      "entropy": 2.0076694644987585,
      "epoch": 0.09187069648349573,
      "grad_norm": 0.3142847716808319,
      "learning_rate": 4.938182851280818e-05,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9398007296025753,
      "num_tokens": 45470048.0,
      "step": 6140
    },
    {
      "entropy": 2.006206801533699,
      "epoch": 0.0920203230249998,
      "grad_norm": 0.29275137186050415,
      "learning_rate": 4.937917165350635e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9441610604524613,
      "num_tokens": 45547053.0,
      "step": 6150
    },
    {
      "entropy": 2.0174421705305576,
      "epoch": 0.09216994956650386,
      "grad_norm": 0.373707115650177,
      "learning_rate": 4.937650916886729e-05,
      "loss": 0.2382,
      "mean_token_accuracy": 0.942445170506835,
      "num_tokens": 45623274.0,
      "step": 6160
    },
    {
      "entropy": 2.057591637969017,
      "epoch": 0.09231957610800792,
      "grad_norm": 0.3343000113964081,
      "learning_rate": 4.9373841059506616e-05,
      "loss": 0.2637,
      "mean_token_accuracy": 0.9346455574035645,
      "num_tokens": 45696348.0,
      "step": 6170
    },
    {
      "entropy": 2.0000326104462145,
      "epoch": 0.09246920264951199,
      "grad_norm": 0.30674639344215393,
      "learning_rate": 4.9371167326041225e-05,
      "loss": 0.2029,
      "mean_token_accuracy": 0.950478482618928,
      "num_tokens": 45768484.0,
      "step": 6180
    },
    {
      "entropy": 1.9368852168321609,
      "epoch": 0.09261882919101605,
      "grad_norm": 0.30373847484588623,
      "learning_rate": 4.936848796908934e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9462162349373102,
      "num_tokens": 45844786.0,
      "step": 6190
    },
    {
      "entropy": 1.9643405690789222,
      "epoch": 0.09276845573252011,
      "grad_norm": 0.4404136538505554,
      "learning_rate": 4.9365802989270494e-05,
      "loss": 0.236,
      "mean_token_accuracy": 0.9428456198424101,
      "num_tokens": 45919576.0,
      "step": 6200
    },
    {
      "entropy": 1.9862743727862835,
      "epoch": 0.09291808227402418,
      "grad_norm": 0.3124238848686218,
      "learning_rate": 4.9363112387205473e-05,
      "loss": 0.2439,
      "mean_token_accuracy": 0.9411855451762676,
      "num_tokens": 45996903.0,
      "step": 6210
    },
    {
      "entropy": 2.018177844583988,
      "epoch": 0.09306770881552824,
      "grad_norm": 0.2903754711151123,
      "learning_rate": 4.936041616351641e-05,
      "loss": 0.2544,
      "mean_token_accuracy": 0.9379108726978302,
      "num_tokens": 46069089.0,
      "step": 6220
    },
    {
      "entropy": 2.010333774983883,
      "epoch": 0.0932173353570323,
      "grad_norm": 0.4018780589103699,
      "learning_rate": 4.935771431882671e-05,
      "loss": 0.2134,
      "mean_token_accuracy": 0.9473061062395572,
      "num_tokens": 46144505.0,
      "step": 6230
    },
    {
      "entropy": 2.02402925491333,
      "epoch": 0.09336696189853637,
      "grad_norm": 0.3372918367385864,
      "learning_rate": 4.93550068537611e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9411699272692203,
      "num_tokens": 46215836.0,
      "step": 6240
    },
    {
      "entropy": 1.9736057244241239,
      "epoch": 0.09351658844004043,
      "grad_norm": 0.38200995326042175,
      "learning_rate": 4.935229376894558e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9462323877960443,
      "num_tokens": 46291057.0,
      "step": 6250
    },
    {
      "entropy": 1.9556421659886838,
      "epoch": 0.0936662149815445,
      "grad_norm": 0.35541853308677673,
      "learning_rate": 4.934957506500748e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9449546594172716,
      "num_tokens": 46366432.0,
      "step": 6260
    },
    {
      "entropy": 1.9607943184673786,
      "epoch": 0.09381584152304856,
      "grad_norm": 0.410266637802124,
      "learning_rate": 4.934685074257539e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9432656101882457,
      "num_tokens": 46440554.0,
      "step": 6270
    },
    {
      "entropy": 1.9585503987967967,
      "epoch": 0.09396546806455264,
      "grad_norm": 0.363657146692276,
      "learning_rate": 4.9344120802279245e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.947188339009881,
      "num_tokens": 46516160.0,
      "step": 6280
    },
    {
      "entropy": 1.9898196868598461,
      "epoch": 0.0941150946060567,
      "grad_norm": 0.2982536554336548,
      "learning_rate": 4.9341385244750245e-05,
      "loss": 0.2744,
      "mean_token_accuracy": 0.9337642528116703,
      "num_tokens": 46587953.0,
      "step": 6290
    },
    {
      "entropy": 1.9880972765386105,
      "epoch": 0.09426472114756076,
      "grad_norm": 0.29255983233451843,
      "learning_rate": 4.933864407062089e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.9409552097320557,
      "num_tokens": 46662000.0,
      "step": 6300
    },
    {
      "entropy": 2.0375262178480624,
      "epoch": 0.09441434768906483,
      "grad_norm": 0.4662066102027893,
      "learning_rate": 4.933589728052501e-05,
      "loss": 0.2498,
      "mean_token_accuracy": 0.9390021458268165,
      "num_tokens": 46735412.0,
      "step": 6310
    },
    {
      "entropy": 2.0049594908952715,
      "epoch": 0.09456397423056889,
      "grad_norm": 0.2765514552593231,
      "learning_rate": 4.9333144875097694e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9438918050378561,
      "num_tokens": 46808113.0,
      "step": 6320
    },
    {
      "entropy": 2.0338924795389177,
      "epoch": 0.09471360077207296,
      "grad_norm": 0.3255186080932617,
      "learning_rate": 4.9330386854975344e-05,
      "loss": 0.247,
      "mean_token_accuracy": 0.9402040060609579,
      "num_tokens": 46881369.0,
      "step": 6330
    },
    {
      "entropy": 1.9966701485216618,
      "epoch": 0.09486322731357702,
      "grad_norm": 0.30386269092559814,
      "learning_rate": 4.932762322079569e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9451959274709225,
      "num_tokens": 46956376.0,
      "step": 6340
    },
    {
      "entropy": 1.9982624724507332,
      "epoch": 0.09501285385508108,
      "grad_norm": 0.36091679334640503,
      "learning_rate": 4.9324853973197696e-05,
      "loss": 0.2454,
      "mean_token_accuracy": 0.9422937951982021,
      "num_tokens": 47028078.0,
      "step": 6350
    },
    {
      "entropy": 1.9810258775949479,
      "epoch": 0.09516248039658515,
      "grad_norm": 0.3851760923862457,
      "learning_rate": 4.9322079112821694e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9429064743220806,
      "num_tokens": 47104536.0,
      "step": 6360
    },
    {
      "entropy": 1.9813734114170074,
      "epoch": 0.09531210693808921,
      "grad_norm": 0.388179212808609,
      "learning_rate": 4.9319298640309266e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.9423199363052845,
      "num_tokens": 47179926.0,
      "step": 6370
    },
    {
      "entropy": 2.0011019200086593,
      "epoch": 0.09546173347959327,
      "grad_norm": 0.3539544343948364,
      "learning_rate": 4.9316512556303315e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9431809086352587,
      "num_tokens": 47251731.0,
      "step": 6380
    },
    {
      "entropy": 2.0126077987253668,
      "epoch": 0.09561136002109734,
      "grad_norm": 0.4333111047744751,
      "learning_rate": 4.931372086144803e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9386585537344218,
      "num_tokens": 47325384.0,
      "step": 6390
    },
    {
      "entropy": 1.9916889756917953,
      "epoch": 0.0957609865626014,
      "grad_norm": 0.34524354338645935,
      "learning_rate": 4.931092355638888e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9421248972415924,
      "num_tokens": 47396632.0,
      "step": 6400
    },
    {
      "entropy": 1.978167374432087,
      "epoch": 0.09591061310410547,
      "grad_norm": 0.34311816096305847,
      "learning_rate": 4.930812064177267e-05,
      "loss": 0.2134,
      "mean_token_accuracy": 0.9490993659943342,
      "num_tokens": 47474205.0,
      "step": 6410
    },
    {
      "entropy": 1.9606876865029335,
      "epoch": 0.09606023964560953,
      "grad_norm": 0.34936586022377014,
      "learning_rate": 4.93053121182475e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9431053284555674,
      "num_tokens": 47550583.0,
      "step": 6420
    },
    {
      "entropy": 1.9932142451405526,
      "epoch": 0.09620986618711361,
      "grad_norm": 0.35197606682777405,
      "learning_rate": 4.930249798646272e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9452523529529572,
      "num_tokens": 47623963.0,
      "step": 6430
    },
    {
      "entropy": 1.995342443138361,
      "epoch": 0.09635949272861767,
      "grad_norm": 0.3068597614765167,
      "learning_rate": 4.929967824706902e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9454005528241396,
      "num_tokens": 47699964.0,
      "step": 6440
    },
    {
      "entropy": 1.9946649871766566,
      "epoch": 0.09650911927012173,
      "grad_norm": 0.3088610768318176,
      "learning_rate": 4.9296852900718384e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9475109238177538,
      "num_tokens": 47772795.0,
      "step": 6450
    },
    {
      "entropy": 1.984515665471554,
      "epoch": 0.0966587458116258,
      "grad_norm": 0.30914172530174255,
      "learning_rate": 4.929402194806407e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9443839006125927,
      "num_tokens": 47848033.0,
      "step": 6460
    },
    {
      "entropy": 2.0005695804953576,
      "epoch": 0.09680837235312986,
      "grad_norm": 0.3280348479747772,
      "learning_rate": 4.9291185389760644e-05,
      "loss": 0.2569,
      "mean_token_accuracy": 0.9374136213213206,
      "num_tokens": 47919398.0,
      "step": 6470
    },
    {
      "entropy": 2.022176963090897,
      "epoch": 0.09695799889463393,
      "grad_norm": 0.38380885124206543,
      "learning_rate": 4.928834322646398e-05,
      "loss": 0.2601,
      "mean_token_accuracy": 0.9359345212578773,
      "num_tokens": 47990822.0,
      "step": 6480
    },
    {
      "entropy": 1.9818934887647628,
      "epoch": 0.09710762543613799,
      "grad_norm": 0.32045137882232666,
      "learning_rate": 4.928549545883122e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9456462956964969,
      "num_tokens": 48064169.0,
      "step": 6490
    },
    {
      "entropy": 1.949671220779419,
      "epoch": 0.09725725197764205,
      "grad_norm": 0.3448818027973175,
      "learning_rate": 4.928264208752083e-05,
      "loss": 0.2475,
      "mean_token_accuracy": 0.9426261276006699,
      "num_tokens": 48138143.0,
      "step": 6500
    },
    {
      "entropy": 2.039323177933693,
      "epoch": 0.09740687851914612,
      "grad_norm": 0.37154433131217957,
      "learning_rate": 4.9279783113192566e-05,
      "loss": 0.2406,
      "mean_token_accuracy": 0.9428331695497036,
      "num_tokens": 48209901.0,
      "step": 6510
    },
    {
      "entropy": 2.0142661474645136,
      "epoch": 0.09755650506065018,
      "grad_norm": 0.3720242381095886,
      "learning_rate": 4.9276918536507455e-05,
      "loss": 0.2508,
      "mean_token_accuracy": 0.9381663110107183,
      "num_tokens": 48284210.0,
      "step": 6520
    },
    {
      "entropy": 2.028626708686352,
      "epoch": 0.09770613160215424,
      "grad_norm": 0.30136674642562866,
      "learning_rate": 4.9274048358127855e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9460651658475399,
      "num_tokens": 48361854.0,
      "step": 6530
    },
    {
      "entropy": 2.04504976272583,
      "epoch": 0.09785575814365831,
      "grad_norm": 0.34556055068969727,
      "learning_rate": 4.927117257871739e-05,
      "loss": 0.2555,
      "mean_token_accuracy": 0.939410300552845,
      "num_tokens": 48435619.0,
      "step": 6540
    },
    {
      "entropy": 2.0802429869771,
      "epoch": 0.09800538468516237,
      "grad_norm": 0.477430522441864,
      "learning_rate": 4.926829119894101e-05,
      "loss": 0.2455,
      "mean_token_accuracy": 0.9387106411159039,
      "num_tokens": 48510356.0,
      "step": 6550
    },
    {
      "entropy": 2.036254157871008,
      "epoch": 0.09815501122666644,
      "grad_norm": 0.313973605632782,
      "learning_rate": 4.926540421946491e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.943378721550107,
      "num_tokens": 48585360.0,
      "step": 6560
    },
    {
      "entropy": 2.0418919779360296,
      "epoch": 0.09830463776817051,
      "grad_norm": 0.3672398030757904,
      "learning_rate": 4.926251164095664e-05,
      "loss": 0.2328,
      "mean_token_accuracy": 0.942517563328147,
      "num_tokens": 48660057.0,
      "step": 6570
    },
    {
      "entropy": 2.020030361413956,
      "epoch": 0.09845426430967458,
      "grad_norm": 0.32402777671813965,
      "learning_rate": 4.9259613464085e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.943076952546835,
      "num_tokens": 48735642.0,
      "step": 6580
    },
    {
      "entropy": 2.0081481248140336,
      "epoch": 0.09860389085117864,
      "grad_norm": 0.3028557300567627,
      "learning_rate": 4.9256709689520113e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9416884869337082,
      "num_tokens": 48810395.0,
      "step": 6590
    },
    {
      "entropy": 2.0543888948857782,
      "epoch": 0.0987535173926827,
      "grad_norm": 0.3846410810947418,
      "learning_rate": 4.9253800317933376e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9429361782968044,
      "num_tokens": 48884559.0,
      "step": 6600
    },
    {
      "entropy": 2.040348903089762,
      "epoch": 0.09890314393418677,
      "grad_norm": 0.3542161285877228,
      "learning_rate": 4.9250885349997475e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9424649070948362,
      "num_tokens": 48958458.0,
      "step": 6610
    },
    {
      "entropy": 2.0048602439463137,
      "epoch": 0.09905277047569083,
      "grad_norm": 0.4752148687839508,
      "learning_rate": 4.924796478638643e-05,
      "loss": 0.2055,
      "mean_token_accuracy": 0.9504819098860026,
      "num_tokens": 49033149.0,
      "step": 6620
    },
    {
      "entropy": 1.9980697922408581,
      "epoch": 0.0992023970171949,
      "grad_norm": 0.3800557255744934,
      "learning_rate": 4.92450386277755e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9426403239369392,
      "num_tokens": 49107582.0,
      "step": 6630
    },
    {
      "entropy": 2.014701187610626,
      "epoch": 0.09935202355869896,
      "grad_norm": 0.45802465081214905,
      "learning_rate": 4.924210687484128e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9355220459401608,
      "num_tokens": 49180217.0,
      "step": 6640
    },
    {
      "entropy": 2.0101888939738273,
      "epoch": 0.09950165010020302,
      "grad_norm": 0.3397325873374939,
      "learning_rate": 4.9239169528261635e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9464712228626013,
      "num_tokens": 49254377.0,
      "step": 6650
    },
    {
      "entropy": 2.0421974755823613,
      "epoch": 0.09965127664170709,
      "grad_norm": 0.33602818846702576,
      "learning_rate": 4.923622658871575e-05,
      "loss": 0.2542,
      "mean_token_accuracy": 0.939394386857748,
      "num_tokens": 49327405.0,
      "step": 6660
    },
    {
      "entropy": 2.0122411526739596,
      "epoch": 0.09980090318321115,
      "grad_norm": 0.37172049283981323,
      "learning_rate": 4.923327805688406e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.9439492631703615,
      "num_tokens": 49401848.0,
      "step": 6670
    },
    {
      "entropy": 2.0287854827940466,
      "epoch": 0.09995052972471521,
      "grad_norm": 0.35554781556129456,
      "learning_rate": 4.923032393344834e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.945672607794404,
      "num_tokens": 49474409.0,
      "step": 6680
    },
    {
      "entropy": 2.03608126193285,
      "epoch": 0.10010015626621928,
      "grad_norm": 0.3851868808269501,
      "learning_rate": 4.9227364219091615e-05,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9394014824181796,
      "num_tokens": 49548922.0,
      "step": 6690
    },
    {
      "entropy": 2.0229356065392494,
      "epoch": 0.10024978280772334,
      "grad_norm": 0.3415566682815552,
      "learning_rate": 4.922439891449824e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9406227465718985,
      "num_tokens": 49621389.0,
      "step": 6700
    },
    {
      "entropy": 2.00074597671628,
      "epoch": 0.1003994093492274,
      "grad_norm": 0.30600619316101074,
      "learning_rate": 4.9221428020353835e-05,
      "loss": 0.2471,
      "mean_token_accuracy": 0.9415341597050428,
      "num_tokens": 49693931.0,
      "step": 6710
    },
    {
      "entropy": 2.0129027239978314,
      "epoch": 0.10054903589073148,
      "grad_norm": 0.31415054202079773,
      "learning_rate": 4.921845153734533e-05,
      "loss": 0.2496,
      "mean_token_accuracy": 0.9378685604780912,
      "num_tokens": 49767464.0,
      "step": 6720
    },
    {
      "entropy": 2.015584246814251,
      "epoch": 0.10069866243223555,
      "grad_norm": 0.2956080138683319,
      "learning_rate": 4.921546946616095e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9437808439135551,
      "num_tokens": 49843885.0,
      "step": 6730
    },
    {
      "entropy": 2.0356770806014537,
      "epoch": 0.10084828897373961,
      "grad_norm": 0.37869328260421753,
      "learning_rate": 4.921248180749019e-05,
      "loss": 0.2473,
      "mean_token_accuracy": 0.9398470185697079,
      "num_tokens": 49918791.0,
      "step": 6740
    },
    {
      "entropy": 2.0189469158649445,
      "epoch": 0.10099791551524367,
      "grad_norm": 0.29984787106513977,
      "learning_rate": 4.920948856202384e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.94599115177989,
      "num_tokens": 49992104.0,
      "step": 6750
    },
    {
      "entropy": 2.006312221288681,
      "epoch": 0.10114754205674774,
      "grad_norm": 0.3130849003791809,
      "learning_rate": 4.9206489730454014e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9456872925162315,
      "num_tokens": 50069858.0,
      "step": 6760
    },
    {
      "entropy": 2.044386038184166,
      "epoch": 0.1012971685982518,
      "grad_norm": 0.38148391246795654,
      "learning_rate": 4.9203485313474076e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9435559201985597,
      "num_tokens": 50144641.0,
      "step": 6770
    },
    {
      "entropy": 2.033870132267475,
      "epoch": 0.10144679513975587,
      "grad_norm": 0.3582657277584076,
      "learning_rate": 4.920047531177872e-05,
      "loss": 0.2132,
      "mean_token_accuracy": 0.9445033177733422,
      "num_tokens": 50217231.0,
      "step": 6780
    },
    {
      "entropy": 1.9934844501316547,
      "epoch": 0.10159642168125993,
      "grad_norm": 0.3550463914871216,
      "learning_rate": 4.91974597260639e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9455395940691232,
      "num_tokens": 50292031.0,
      "step": 6790
    },
    {
      "entropy": 1.997012511640787,
      "epoch": 0.101746048222764,
      "grad_norm": 0.4058363735675812,
      "learning_rate": 4.919443855702688e-05,
      "loss": 0.2467,
      "mean_token_accuracy": 0.940946352109313,
      "num_tokens": 50366730.0,
      "step": 6800
    },
    {
      "entropy": 2.0200984247028826,
      "epoch": 0.10189567476426806,
      "grad_norm": 0.2931559085845947,
      "learning_rate": 4.919141180536619e-05,
      "loss": 0.2406,
      "mean_token_accuracy": 0.9404024593532085,
      "num_tokens": 50438719.0,
      "step": 6810
    },
    {
      "entropy": 2.037948836386204,
      "epoch": 0.10204530130577212,
      "grad_norm": 0.35713475942611694,
      "learning_rate": 4.91883794717817e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9393110923469067,
      "num_tokens": 50511589.0,
      "step": 6820
    },
    {
      "entropy": 2.049623614549637,
      "epoch": 0.10219492784727618,
      "grad_norm": 0.3255009353160858,
      "learning_rate": 4.918534155697451e-05,
      "loss": 0.2459,
      "mean_token_accuracy": 0.9414567947387695,
      "num_tokens": 50583939.0,
      "step": 6830
    },
    {
      "entropy": 2.045039141178131,
      "epoch": 0.10234455438878025,
      "grad_norm": 0.4031001329421997,
      "learning_rate": 4.918229806164705e-05,
      "loss": 0.2483,
      "mean_token_accuracy": 0.9433498054742813,
      "num_tokens": 50661354.0,
      "step": 6840
    },
    {
      "entropy": 2.021258094161749,
      "epoch": 0.10249418093028431,
      "grad_norm": 0.37417805194854736,
      "learning_rate": 4.917924898650303e-05,
      "loss": 0.2599,
      "mean_token_accuracy": 0.937105530500412,
      "num_tokens": 50733094.0,
      "step": 6850
    },
    {
      "entropy": 2.0340081468224525,
      "epoch": 0.10264380747178839,
      "grad_norm": 0.30731359124183655,
      "learning_rate": 4.9176194332247456e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9450399167835712,
      "num_tokens": 50808446.0,
      "step": 6860
    },
    {
      "entropy": 2.046274674683809,
      "epoch": 0.10279343401329245,
      "grad_norm": 0.3464275598526001,
      "learning_rate": 4.917313409958661e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9422964431345463,
      "num_tokens": 50882222.0,
      "step": 6870
    },
    {
      "entropy": 2.0226226963102816,
      "epoch": 0.10294306055479652,
      "grad_norm": 0.27297133207321167,
      "learning_rate": 4.9170068289228085e-05,
      "loss": 0.2452,
      "mean_token_accuracy": 0.9412497516721487,
      "num_tokens": 50957201.0,
      "step": 6880
    },
    {
      "entropy": 2.015940083563328,
      "epoch": 0.10309268709630058,
      "grad_norm": 0.4369887411594391,
      "learning_rate": 4.916699690188074e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9450420372188091,
      "num_tokens": 51032750.0,
      "step": 6890
    },
    {
      "entropy": 2.0171053767204286,
      "epoch": 0.10324231363780464,
      "grad_norm": 0.3672684133052826,
      "learning_rate": 4.916391993825473e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.940800667181611,
      "num_tokens": 51104672.0,
      "step": 6900
    },
    {
      "entropy": 2.0696181930601596,
      "epoch": 0.10339194017930871,
      "grad_norm": 0.3166293799877167,
      "learning_rate": 4.916083739906151e-05,
      "loss": 0.2024,
      "mean_token_accuracy": 0.947877299413085,
      "num_tokens": 51181356.0,
      "step": 6910
    },
    {
      "entropy": 2.035859489440918,
      "epoch": 0.10354156672081277,
      "grad_norm": 0.3044778108596802,
      "learning_rate": 4.915774928501382e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9431015502661466,
      "num_tokens": 51254789.0,
      "step": 6920
    },
    {
      "entropy": 2.018730130046606,
      "epoch": 0.10369119326231684,
      "grad_norm": 0.32537880539894104,
      "learning_rate": 4.9154655596825685e-05,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9445736270397902,
      "num_tokens": 51330211.0,
      "step": 6930
    },
    {
      "entropy": 2.0382556296885013,
      "epoch": 0.1038408198038209,
      "grad_norm": 0.37756624817848206,
      "learning_rate": 4.915155633521243e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9416738353669644,
      "num_tokens": 51403345.0,
      "step": 6940
    },
    {
      "entropy": 2.0344260580837727,
      "epoch": 0.10399044634532496,
      "grad_norm": 0.312160849571228,
      "learning_rate": 4.914845150089064e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9483914408832789,
      "num_tokens": 51474987.0,
      "step": 6950
    },
    {
      "entropy": 2.000800547748804,
      "epoch": 0.10414007288682903,
      "grad_norm": 0.3466229736804962,
      "learning_rate": 4.9145341094578226e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9453424379229546,
      "num_tokens": 51549070.0,
      "step": 6960
    },
    {
      "entropy": 1.9574648439884186,
      "epoch": 0.10428969942833309,
      "grad_norm": 0.36725762486457825,
      "learning_rate": 4.914222511699435e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9428684048354625,
      "num_tokens": 51623393.0,
      "step": 6970
    },
    {
      "entropy": 1.9608295947313308,
      "epoch": 0.10443932596983715,
      "grad_norm": 0.37494465708732605,
      "learning_rate": 4.913910356885951e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9370633356273175,
      "num_tokens": 51696338.0,
      "step": 6980
    },
    {
      "entropy": 1.9949514374136925,
      "epoch": 0.10458895251134122,
      "grad_norm": 0.36238107085227966,
      "learning_rate": 4.913597645089544e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9397410385310649,
      "num_tokens": 51769611.0,
      "step": 6990
    },
    {
      "entropy": 1.959775475412607,
      "epoch": 0.10473857905284528,
      "grad_norm": 0.3562401235103607,
      "learning_rate": 4.913284376382518e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9438587129116058,
      "num_tokens": 51843379.0,
      "step": 7000
    },
    {
      "epoch": 0.10473857905284528,
      "eval_entropy": 1.956783408164978,
      "eval_loss": 0.26053309440612793,
      "eval_mean_token_accuracy": 0.9424084006667137,
      "eval_num_tokens": 51843379.0,
      "eval_runtime": 544.0839,
      "eval_samples_per_second": 36.759,
      "eval_steps_per_second": 9.19,
      "step": 7000
    },
    {
      "entropy": 1.9702783420681953,
      "epoch": 0.10488820559434936,
      "grad_norm": 0.34491923451423645,
      "learning_rate": 4.912970550837309e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9444088857620955,
      "num_tokens": 51918197.0,
      "step": 7010
    },
    {
      "entropy": 1.9687664464116097,
      "epoch": 0.10503783213585342,
      "grad_norm": 0.3593152165412903,
      "learning_rate": 4.912656168526477e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.944999472796917,
      "num_tokens": 51993262.0,
      "step": 7020
    },
    {
      "entropy": 1.9813758581876755,
      "epoch": 0.10518745867735749,
      "grad_norm": 0.45141223073005676,
      "learning_rate": 4.912341229522713e-05,
      "loss": 0.2447,
      "mean_token_accuracy": 0.9380789436399937,
      "num_tokens": 52065539.0,
      "step": 7030
    },
    {
      "entropy": 1.9591349750757217,
      "epoch": 0.10533708521886155,
      "grad_norm": 0.36457210779190063,
      "learning_rate": 4.912025733898837e-05,
      "loss": 0.231,
      "mean_token_accuracy": 0.9424976810812951,
      "num_tokens": 52141131.0,
      "step": 7040
    },
    {
      "entropy": 1.934197022020817,
      "epoch": 0.10548671176036561,
      "grad_norm": 0.3280206620693207,
      "learning_rate": 4.9117096817277974e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9416973732411862,
      "num_tokens": 52212532.0,
      "step": 7050
    },
    {
      "entropy": 1.9187272541224956,
      "epoch": 0.10563633830186968,
      "grad_norm": 0.3634321093559265,
      "learning_rate": 4.91139307308267e-05,
      "loss": 0.2456,
      "mean_token_accuracy": 0.9398303527384997,
      "num_tokens": 52286821.0,
      "step": 7060
    },
    {
      "entropy": 1.9340363413095474,
      "epoch": 0.10578596484337374,
      "grad_norm": 0.35386034846305847,
      "learning_rate": 4.911075908036662e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9458859506994486,
      "num_tokens": 52358956.0,
      "step": 7070
    },
    {
      "entropy": 1.9237247727811337,
      "epoch": 0.1059355913848778,
      "grad_norm": 0.3665013015270233,
      "learning_rate": 4.910758186663106e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9463939022272825,
      "num_tokens": 52433096.0,
      "step": 7080
    },
    {
      "entropy": 1.92852555885911,
      "epoch": 0.10608521792638187,
      "grad_norm": 0.32665374875068665,
      "learning_rate": 4.910439909035465e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.9428718138486147,
      "num_tokens": 52507270.0,
      "step": 7090
    },
    {
      "entropy": 1.9349713616073132,
      "epoch": 0.10623484446788593,
      "grad_norm": 0.29269692301750183,
      "learning_rate": 4.9101210752273316e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9449943996965885,
      "num_tokens": 52581430.0,
      "step": 7100
    },
    {
      "entropy": 1.9410242430865765,
      "epoch": 0.10638447100939,
      "grad_norm": 0.3780823349952698,
      "learning_rate": 4.909801685312425e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9380411934107542,
      "num_tokens": 52654274.0,
      "step": 7110
    },
    {
      "entropy": 1.9378709226846695,
      "epoch": 0.10653409755089406,
      "grad_norm": 0.36280402541160583,
      "learning_rate": 4.909481739364594e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9425738591700792,
      "num_tokens": 52728217.0,
      "step": 7120
    },
    {
      "entropy": 1.9600737757980824,
      "epoch": 0.10668372409239812,
      "grad_norm": 0.3633546531200409,
      "learning_rate": 4.909161237457816e-05,
      "loss": 0.2385,
      "mean_token_accuracy": 0.9408579252660274,
      "num_tokens": 52802991.0,
      "step": 7130
    },
    {
      "entropy": 1.9683450505137443,
      "epoch": 0.10683335063390219,
      "grad_norm": 0.35184985399246216,
      "learning_rate": 4.908840179666196e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.941391246393323,
      "num_tokens": 52878767.0,
      "step": 7140
    },
    {
      "entropy": 1.971848327666521,
      "epoch": 0.10698297717540627,
      "grad_norm": 0.2789813280105591,
      "learning_rate": 4.9085185660639696e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9402335368096828,
      "num_tokens": 52952657.0,
      "step": 7150
    },
    {
      "entropy": 1.9906848222017288,
      "epoch": 0.10713260371691033,
      "grad_norm": 0.2769067883491516,
      "learning_rate": 4.908196396725498e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9459978934377432,
      "num_tokens": 53028466.0,
      "step": 7160
    },
    {
      "entropy": 1.9547665022313594,
      "epoch": 0.1072822302584144,
      "grad_norm": 0.3809768855571747,
      "learning_rate": 4.907873671725274e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9410716891288757,
      "num_tokens": 53103411.0,
      "step": 7170
    },
    {
      "entropy": 1.9816002301871776,
      "epoch": 0.10743185679991846,
      "grad_norm": 0.3421342074871063,
      "learning_rate": 4.9075503911379156e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9437227305024862,
      "num_tokens": 53175845.0,
      "step": 7180
    },
    {
      "entropy": 1.924488516151905,
      "epoch": 0.10758148334142252,
      "grad_norm": 0.37161052227020264,
      "learning_rate": 4.9072265550381726e-05,
      "loss": 0.2395,
      "mean_token_accuracy": 0.9412518728524446,
      "num_tokens": 53252165.0,
      "step": 7190
    },
    {
      "entropy": 1.949691928178072,
      "epoch": 0.10773110988292658,
      "grad_norm": 0.36173954606056213,
      "learning_rate": 4.9069021635009214e-05,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9467463001608849,
      "num_tokens": 53327897.0,
      "step": 7200
    },
    {
      "entropy": 1.9609467647969723,
      "epoch": 0.10788073642443065,
      "grad_norm": 0.34035366773605347,
      "learning_rate": 4.906577216601168e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9427688263356686,
      "num_tokens": 53399947.0,
      "step": 7210
    },
    {
      "entropy": 1.9431313663721084,
      "epoch": 0.10803036296593471,
      "grad_norm": 0.37429338693618774,
      "learning_rate": 4.9062517144140416e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9409451283514499,
      "num_tokens": 53475080.0,
      "step": 7220
    },
    {
      "entropy": 1.9719017416238784,
      "epoch": 0.10817998950743878,
      "grad_norm": 0.3231341242790222,
      "learning_rate": 4.90592565701481e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9440370395779609,
      "num_tokens": 53550090.0,
      "step": 7230
    },
    {
      "entropy": 1.9454951584339142,
      "epoch": 0.10832961604894284,
      "grad_norm": 0.3078174889087677,
      "learning_rate": 4.905599044478859e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9440415371209383,
      "num_tokens": 53626682.0,
      "step": 7240
    },
    {
      "entropy": 1.9504641361534596,
      "epoch": 0.1084792425904469,
      "grad_norm": 0.427604615688324,
      "learning_rate": 4.90527187688171e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.9419248614460229,
      "num_tokens": 53699747.0,
      "step": 7250
    },
    {
      "entropy": 1.9466063216328622,
      "epoch": 0.10862886913195097,
      "grad_norm": 0.3223711848258972,
      "learning_rate": 4.904944154299009e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9433444648981094,
      "num_tokens": 53775107.0,
      "step": 7260
    },
    {
      "entropy": 1.9763004317879678,
      "epoch": 0.10877849567345503,
      "grad_norm": 0.3342854082584381,
      "learning_rate": 4.9046158768065295e-05,
      "loss": 0.2507,
      "mean_token_accuracy": 0.9385780848562717,
      "num_tokens": 53847138.0,
      "step": 7270
    },
    {
      "entropy": 1.9581170223653317,
      "epoch": 0.1089281222149591,
      "grad_norm": 0.3320242762565613,
      "learning_rate": 4.904287044480179e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9411128904670477,
      "num_tokens": 53922652.0,
      "step": 7280
    },
    {
      "entropy": 1.9770950928330422,
      "epoch": 0.10907774875646317,
      "grad_norm": 0.3976527452468872,
      "learning_rate": 4.903957657395986e-05,
      "loss": 0.2568,
      "mean_token_accuracy": 0.9363223131746053,
      "num_tokens": 53999118.0,
      "step": 7290
    },
    {
      "entropy": 1.9558130383491517,
      "epoch": 0.10922737529796724,
      "grad_norm": 0.3329937756061554,
      "learning_rate": 4.903627715630111e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9427610773593187,
      "num_tokens": 54072691.0,
      "step": 7300
    },
    {
      "entropy": 1.9647119984030723,
      "epoch": 0.1093770018394713,
      "grad_norm": 0.3223440945148468,
      "learning_rate": 4.9032972192588444e-05,
      "loss": 0.2134,
      "mean_token_accuracy": 0.9471665702760219,
      "num_tokens": 54147639.0,
      "step": 7310
    },
    {
      "entropy": 1.954242644459009,
      "epoch": 0.10952662838097536,
      "grad_norm": 0.4005436599254608,
      "learning_rate": 4.9029661683586016e-05,
      "loss": 0.2393,
      "mean_token_accuracy": 0.9430974654853344,
      "num_tokens": 54218787.0,
      "step": 7320
    },
    {
      "entropy": 1.9606421187520027,
      "epoch": 0.10967625492247943,
      "grad_norm": 0.382323682308197,
      "learning_rate": 4.902634563005927e-05,
      "loss": 0.2366,
      "mean_token_accuracy": 0.9392609771341085,
      "num_tokens": 54292462.0,
      "step": 7330
    },
    {
      "entropy": 2.00235248580575,
      "epoch": 0.10982588146398349,
      "grad_norm": 0.35012274980545044,
      "learning_rate": 4.9023024032774944e-05,
      "loss": 0.2528,
      "mean_token_accuracy": 0.9403872843831778,
      "num_tokens": 54366287.0,
      "step": 7340
    },
    {
      "entropy": 1.9762489810585975,
      "epoch": 0.10997550800548755,
      "grad_norm": 0.2740956246852875,
      "learning_rate": 4.901969689250105e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9452012725174427,
      "num_tokens": 54440711.0,
      "step": 7350
    },
    {
      "entropy": 1.970696920901537,
      "epoch": 0.11012513454699162,
      "grad_norm": 0.32421842217445374,
      "learning_rate": 4.901636421000687e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9400802738964558,
      "num_tokens": 54514526.0,
      "step": 7360
    },
    {
      "entropy": 1.9665995575487614,
      "epoch": 0.11027476108849568,
      "grad_norm": 0.32229533791542053,
      "learning_rate": 4.9013025986063e-05,
      "loss": 0.2348,
      "mean_token_accuracy": 0.9421021923422813,
      "num_tokens": 54587049.0,
      "step": 7370
    },
    {
      "entropy": 1.9783765882253648,
      "epoch": 0.11042438762999975,
      "grad_norm": 0.3116665184497833,
      "learning_rate": 4.9009682221441284e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9432570185512305,
      "num_tokens": 54660191.0,
      "step": 7380
    },
    {
      "entropy": 2.0220399662852286,
      "epoch": 0.11057401417150381,
      "grad_norm": 0.3588930368423462,
      "learning_rate": 4.900633291691486e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9432217977941036,
      "num_tokens": 54732859.0,
      "step": 7390
    },
    {
      "entropy": 1.9662847317755223,
      "epoch": 0.11072364071300787,
      "grad_norm": 0.30455639958381653,
      "learning_rate": 4.9002978073258156e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9416861586272717,
      "num_tokens": 54807109.0,
      "step": 7400
    },
    {
      "entropy": 1.9805108420550823,
      "epoch": 0.11087326725451194,
      "grad_norm": 0.26782509684562683,
      "learning_rate": 4.8999617691246855e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9467954777181149,
      "num_tokens": 54882861.0,
      "step": 7410
    },
    {
      "entropy": 1.9894406840205192,
      "epoch": 0.111022893796016,
      "grad_norm": 0.3435003459453583,
      "learning_rate": 4.899625177165795e-05,
      "loss": 0.2222,
      "mean_token_accuracy": 0.9434870753437281,
      "num_tokens": 54954654.0,
      "step": 7420
    },
    {
      "entropy": 1.9658081576228141,
      "epoch": 0.11117252033752006,
      "grad_norm": 0.28491100668907166,
      "learning_rate": 4.89928803152697e-05,
      "loss": 0.203,
      "mean_token_accuracy": 0.9494588177651166,
      "num_tokens": 55031006.0,
      "step": 7430
    },
    {
      "entropy": 1.9743023954331875,
      "epoch": 0.11132214687902414,
      "grad_norm": 0.405352383852005,
      "learning_rate": 4.898950332286165e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9441508024930954,
      "num_tokens": 55103646.0,
      "step": 7440
    },
    {
      "entropy": 1.976608407497406,
      "epoch": 0.1114717734205282,
      "grad_norm": 0.28956928849220276,
      "learning_rate": 4.89861207952146e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.9423568956553936,
      "num_tokens": 55181490.0,
      "step": 7450
    },
    {
      "entropy": 2.0138902612030507,
      "epoch": 0.11162139996203227,
      "grad_norm": 0.32477277517318726,
      "learning_rate": 4.898273273311067e-05,
      "loss": 0.2541,
      "mean_token_accuracy": 0.935784287005663,
      "num_tokens": 55256383.0,
      "step": 7460
    },
    {
      "entropy": 2.026325400173664,
      "epoch": 0.11177102650353633,
      "grad_norm": 0.3543660342693329,
      "learning_rate": 4.897933913733323e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9479182921350002,
      "num_tokens": 55330447.0,
      "step": 7470
    },
    {
      "entropy": 2.013948369026184,
      "epoch": 0.1119206530450404,
      "grad_norm": 0.3204508423805237,
      "learning_rate": 4.897594000866696e-05,
      "loss": 0.2239,
      "mean_token_accuracy": 0.9439801510423422,
      "num_tokens": 55403850.0,
      "step": 7480
    },
    {
      "entropy": 1.9721547089517117,
      "epoch": 0.11207027958654446,
      "grad_norm": 0.2992527484893799,
      "learning_rate": 4.897253534789777e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.94583880007267,
      "num_tokens": 55480909.0,
      "step": 7490
    },
    {
      "entropy": 1.9615136340260506,
      "epoch": 0.11221990612804852,
      "grad_norm": 0.4126185178756714,
      "learning_rate": 4.896912515581289e-05,
      "loss": 0.2417,
      "mean_token_accuracy": 0.9408015914261341,
      "num_tokens": 55554786.0,
      "step": 7500
    },
    {
      "entropy": 1.9778620898723602,
      "epoch": 0.11236953266955259,
      "grad_norm": 0.3288714587688446,
      "learning_rate": 4.896570943320083e-05,
      "loss": 0.2533,
      "mean_token_accuracy": 0.9393782801926136,
      "num_tokens": 55628173.0,
      "step": 7510
    },
    {
      "entropy": 1.9640972554683684,
      "epoch": 0.11251915921105665,
      "grad_norm": 0.3483368158340454,
      "learning_rate": 4.896228818085135e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9465756364166736,
      "num_tokens": 55703220.0,
      "step": 7520
    },
    {
      "entropy": 1.9573802798986435,
      "epoch": 0.11266878575256072,
      "grad_norm": 0.39200565218925476,
      "learning_rate": 4.8958861399555514e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.945859632268548,
      "num_tokens": 55776846.0,
      "step": 7530
    },
    {
      "entropy": 1.9335975103080272,
      "epoch": 0.11281841229406478,
      "grad_norm": 0.35961753129959106,
      "learning_rate": 4.8955429090105645e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9464996080845595,
      "num_tokens": 55849750.0,
      "step": 7540
    },
    {
      "entropy": 1.9627510450780392,
      "epoch": 0.11296803883556884,
      "grad_norm": 0.3746900260448456,
      "learning_rate": 4.895199125329536e-05,
      "loss": 0.272,
      "mean_token_accuracy": 0.9338338285684585,
      "num_tokens": 55920487.0,
      "step": 7550
    },
    {
      "entropy": 2.0039441332221033,
      "epoch": 0.1131176653770729,
      "grad_norm": 0.39414066076278687,
      "learning_rate": 4.894854788991954e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9390516452491283,
      "num_tokens": 55992613.0,
      "step": 7560
    },
    {
      "entropy": 2.008208806812763,
      "epoch": 0.11326729191857697,
      "grad_norm": 0.37331703305244446,
      "learning_rate": 4.894509900077437e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9450124565511941,
      "num_tokens": 56068322.0,
      "step": 7570
    },
    {
      "entropy": 1.9876067832112312,
      "epoch": 0.11341691846008105,
      "grad_norm": 0.3615008294582367,
      "learning_rate": 4.894164458665728e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9443882420659065,
      "num_tokens": 56143571.0,
      "step": 7580
    },
    {
      "entropy": 1.949827364832163,
      "epoch": 0.11356654500158511,
      "grad_norm": 0.28961238265037537,
      "learning_rate": 4.8938184648366995e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.939967067539692,
      "num_tokens": 56219944.0,
      "step": 7590
    },
    {
      "entropy": 1.9500743009150028,
      "epoch": 0.11371617154308918,
      "grad_norm": 0.2952779531478882,
      "learning_rate": 4.893471918670351e-05,
      "loss": 0.2153,
      "mean_token_accuracy": 0.944991498067975,
      "num_tokens": 56295919.0,
      "step": 7600
    },
    {
      "entropy": 1.959369846433401,
      "epoch": 0.11386579808459324,
      "grad_norm": 0.36025580763816833,
      "learning_rate": 4.8931248202468116e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9389270719140768,
      "num_tokens": 56371871.0,
      "step": 7610
    },
    {
      "entropy": 1.911109447479248,
      "epoch": 0.1140154246260973,
      "grad_norm": 0.3745344877243042,
      "learning_rate": 4.892777169646335e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9435254398733377,
      "num_tokens": 56446458.0,
      "step": 7620
    },
    {
      "entropy": 1.9355987779796124,
      "epoch": 0.11416505116760137,
      "grad_norm": 0.31572961807250977,
      "learning_rate": 4.892428966949305e-05,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9436810456216336,
      "num_tokens": 56522482.0,
      "step": 7630
    },
    {
      "entropy": 1.9828083612024785,
      "epoch": 0.11431467770910543,
      "grad_norm": 0.36202681064605713,
      "learning_rate": 4.892080212236232e-05,
      "loss": 0.255,
      "mean_token_accuracy": 0.9377515502274036,
      "num_tokens": 56595306.0,
      "step": 7640
    },
    {
      "entropy": 1.9993874795734883,
      "epoch": 0.1144643042506095,
      "grad_norm": 0.31084680557250977,
      "learning_rate": 4.891730905587755e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9476212445646525,
      "num_tokens": 56668696.0,
      "step": 7650
    },
    {
      "entropy": 1.958421205729246,
      "epoch": 0.11461393079211356,
      "grad_norm": 0.3913757801055908,
      "learning_rate": 4.891381047084638e-05,
      "loss": 0.2575,
      "mean_token_accuracy": 0.9382905591279268,
      "num_tokens": 56740383.0,
      "step": 7660
    },
    {
      "entropy": 1.9506193578243256,
      "epoch": 0.11476355733361762,
      "grad_norm": 0.31375589966773987,
      "learning_rate": 4.891030636807777e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9412986893206835,
      "num_tokens": 56813692.0,
      "step": 7670
    },
    {
      "entropy": 1.951964906603098,
      "epoch": 0.11491318387512169,
      "grad_norm": 0.3148193657398224,
      "learning_rate": 4.890679674838191e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9430453546345234,
      "num_tokens": 56888009.0,
      "step": 7680
    },
    {
      "entropy": 1.9718525499105453,
      "epoch": 0.11506281041662575,
      "grad_norm": 0.40108856558799744,
      "learning_rate": 4.89032816125703e-05,
      "loss": 0.2589,
      "mean_token_accuracy": 0.9353983599692584,
      "num_tokens": 56961083.0,
      "step": 7690
    },
    {
      "entropy": 1.9764004416763783,
      "epoch": 0.11521243695812981,
      "grad_norm": 0.3184696435928345,
      "learning_rate": 4.88997609614557e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9441411580890418,
      "num_tokens": 57032566.0,
      "step": 7700
    },
    {
      "entropy": 1.988452311605215,
      "epoch": 0.11536206349963388,
      "grad_norm": 0.3779258728027344,
      "learning_rate": 4.889623479585214e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9406640086323023,
      "num_tokens": 57105401.0,
      "step": 7710
    },
    {
      "entropy": 1.9512231409549714,
      "epoch": 0.11551169004113794,
      "grad_norm": 0.3697732985019684,
      "learning_rate": 4.889270311657494e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9416694547981024,
      "num_tokens": 57178212.0,
      "step": 7720
    },
    {
      "entropy": 1.9566611379384995,
      "epoch": 0.11566131658264202,
      "grad_norm": 0.300411581993103,
      "learning_rate": 4.888916592444068e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9436910539865494,
      "num_tokens": 57253184.0,
      "step": 7730
    },
    {
      "entropy": 1.9884315066039562,
      "epoch": 0.11581094312414608,
      "grad_norm": 0.2767464220523834,
      "learning_rate": 4.8885623220267226e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9415495052933693,
      "num_tokens": 57324820.0,
      "step": 7740
    },
    {
      "entropy": 1.9842980623245239,
      "epoch": 0.11596056966565015,
      "grad_norm": 0.3031434714794159,
      "learning_rate": 4.8882075004873715e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9417913638055324,
      "num_tokens": 57401696.0,
      "step": 7750
    },
    {
      "entropy": 1.9897237218916417,
      "epoch": 0.11611019620715421,
      "grad_norm": 0.2866101861000061,
      "learning_rate": 4.8878521279080565e-05,
      "loss": 0.2104,
      "mean_token_accuracy": 0.9432899966835976,
      "num_tokens": 57477092.0,
      "step": 7760
    },
    {
      "entropy": 2.0056554906070234,
      "epoch": 0.11625982274865827,
      "grad_norm": 0.3836924433708191,
      "learning_rate": 4.887496204370944e-05,
      "loss": 0.2439,
      "mean_token_accuracy": 0.9392777111381292,
      "num_tokens": 57551545.0,
      "step": 7770
    },
    {
      "entropy": 1.984122658520937,
      "epoch": 0.11640944929016234,
      "grad_norm": 0.3618398904800415,
      "learning_rate": 4.8871397299583315e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9456167384982109,
      "num_tokens": 57627461.0,
      "step": 7780
    },
    {
      "entropy": 1.98211180716753,
      "epoch": 0.1165590758316664,
      "grad_norm": 0.31075379252433777,
      "learning_rate": 4.8867827047526416e-05,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9441700298339128,
      "num_tokens": 57703230.0,
      "step": 7790
    },
    {
      "entropy": 1.975354964286089,
      "epoch": 0.11670870237317046,
      "grad_norm": 0.35607680678367615,
      "learning_rate": 4.8864251288364254e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9414236027747392,
      "num_tokens": 57777026.0,
      "step": 7800
    },
    {
      "entropy": 1.957238256931305,
      "epoch": 0.11685832891467453,
      "grad_norm": 0.2830517292022705,
      "learning_rate": 4.88606700229236e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9407104577869176,
      "num_tokens": 57854652.0,
      "step": 7810
    },
    {
      "entropy": 1.9861981220543385,
      "epoch": 0.11700795545617859,
      "grad_norm": 0.32007896900177,
      "learning_rate": 4.8857083252032516e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9444633319973945,
      "num_tokens": 57932134.0,
      "step": 7820
    },
    {
      "entropy": 2.0038851484656335,
      "epoch": 0.11715758199768266,
      "grad_norm": 0.3092794418334961,
      "learning_rate": 4.885349097652032e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.9408882360905408,
      "num_tokens": 58006010.0,
      "step": 7830
    },
    {
      "entropy": 2.012570309638977,
      "epoch": 0.11730720853918672,
      "grad_norm": 0.3337852656841278,
      "learning_rate": 4.884989319721761e-05,
      "loss": 0.2461,
      "mean_token_accuracy": 0.9410578105598688,
      "num_tokens": 58078986.0,
      "step": 7840
    },
    {
      "entropy": 1.9848682075738906,
      "epoch": 0.11745683508069078,
      "grad_norm": 0.3809310495853424,
      "learning_rate": 4.884628991495626e-05,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9422338467091322,
      "num_tokens": 58154694.0,
      "step": 7850
    },
    {
      "entropy": 1.9879748292267323,
      "epoch": 0.11760646162219485,
      "grad_norm": 0.38753095269203186,
      "learning_rate": 4.884268113056942e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9426141887903213,
      "num_tokens": 58228951.0,
      "step": 7860
    },
    {
      "entropy": 2.008814745396376,
      "epoch": 0.11775608816369892,
      "grad_norm": 0.29451096057891846,
      "learning_rate": 4.883906684489148e-05,
      "loss": 0.2454,
      "mean_token_accuracy": 0.93609982393682,
      "num_tokens": 58305245.0,
      "step": 7870
    },
    {
      "entropy": 2.018505879491568,
      "epoch": 0.11790571470520299,
      "grad_norm": 0.38933372497558594,
      "learning_rate": 4.883544705875815e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9428345333784819,
      "num_tokens": 58379500.0,
      "step": 7880
    },
    {
      "entropy": 2.0120041079819204,
      "epoch": 0.11805534124670705,
      "grad_norm": 0.3472827970981598,
      "learning_rate": 4.8831821773006394e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9430873706936836,
      "num_tokens": 58452311.0,
      "step": 7890
    },
    {
      "entropy": 2.0061072528362276,
      "epoch": 0.11820496778821112,
      "grad_norm": 0.32910364866256714,
      "learning_rate": 4.882819098847442e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9434091363102197,
      "num_tokens": 58524760.0,
      "step": 7900
    },
    {
      "entropy": 1.942064929753542,
      "epoch": 0.11835459432971518,
      "grad_norm": 0.28931930661201477,
      "learning_rate": 4.8824554706001746e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9441337864845991,
      "num_tokens": 58600926.0,
      "step": 7910
    },
    {
      "entropy": 1.9873508840799332,
      "epoch": 0.11850422087121924,
      "grad_norm": 0.4340669810771942,
      "learning_rate": 4.882091292642913e-05,
      "loss": 0.2411,
      "mean_token_accuracy": 0.9440523099154234,
      "num_tokens": 58676526.0,
      "step": 7920
    },
    {
      "entropy": 2.007344464212656,
      "epoch": 0.1186538474127233,
      "grad_norm": 0.339648574590683,
      "learning_rate": 4.8817265650598636e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9427611395716667,
      "num_tokens": 58753103.0,
      "step": 7930
    },
    {
      "entropy": 2.018076355010271,
      "epoch": 0.11880347395422737,
      "grad_norm": 0.38849717378616333,
      "learning_rate": 4.8813612879353565e-05,
      "loss": 0.2493,
      "mean_token_accuracy": 0.9375931937247515,
      "num_tokens": 58825096.0,
      "step": 7940
    },
    {
      "entropy": 1.9962676964700221,
      "epoch": 0.11895310049573143,
      "grad_norm": 0.36552372574806213,
      "learning_rate": 4.880995461353852e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9422419529408217,
      "num_tokens": 58900156.0,
      "step": 7950
    },
    {
      "entropy": 2.0157137289643288,
      "epoch": 0.1191027270372355,
      "grad_norm": 0.435794472694397,
      "learning_rate": 4.880629085399934e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9419899836182595,
      "num_tokens": 58973674.0,
      "step": 7960
    },
    {
      "entropy": 2.025240730494261,
      "epoch": 0.11925235357873956,
      "grad_norm": 0.3175814747810364,
      "learning_rate": 4.8802621601583145e-05,
      "loss": 0.2372,
      "mean_token_accuracy": 0.942645663022995,
      "num_tokens": 59047924.0,
      "step": 7970
    },
    {
      "entropy": 2.038414245098829,
      "epoch": 0.11940198012024363,
      "grad_norm": 0.25668326020240784,
      "learning_rate": 4.879894685713835e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9440921753644943,
      "num_tokens": 59121408.0,
      "step": 7980
    },
    {
      "entropy": 2.007302012294531,
      "epoch": 0.11955160666174769,
      "grad_norm": 0.32251331210136414,
      "learning_rate": 4.879526662151461e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.941983325779438,
      "num_tokens": 59195756.0,
      "step": 7990
    },
    {
      "entropy": 1.9465818949043752,
      "epoch": 0.11970123320325175,
      "grad_norm": 0.31368765234947205,
      "learning_rate": 4.8791580895562874e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9449691016227006,
      "num_tokens": 59269815.0,
      "step": 8000
    },
    {
      "epoch": 0.11970123320325175,
      "eval_entropy": 1.9632114042282105,
      "eval_loss": 0.2563895881175995,
      "eval_mean_token_accuracy": 0.9431099826931953,
      "eval_num_tokens": 59269815.0,
      "eval_runtime": 545.3533,
      "eval_samples_per_second": 36.673,
      "eval_steps_per_second": 9.168,
      "step": 8000
    },
    {
      "entropy": 1.9650651969015598,
      "epoch": 0.11985085974475582,
      "grad_norm": 0.2659401595592499,
      "learning_rate": 4.878788968013533e-05,
      "loss": 0.2181,
      "mean_token_accuracy": 0.944899171590805,
      "num_tokens": 59344461.0,
      "step": 8010
    },
    {
      "entropy": 1.991673719882965,
      "epoch": 0.1200004862862599,
      "grad_norm": 0.35998767614364624,
      "learning_rate": 4.878419297608546e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9390778996050357,
      "num_tokens": 59420050.0,
      "step": 8020
    },
    {
      "entropy": 1.937024176120758,
      "epoch": 0.12015011282776396,
      "grad_norm": 0.35680779814720154,
      "learning_rate": 4.878049078426802e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9437963265925646,
      "num_tokens": 59493562.0,
      "step": 8030
    },
    {
      "entropy": 1.9642414554953576,
      "epoch": 0.12029973936926802,
      "grad_norm": 0.35885006189346313,
      "learning_rate": 4.8776783105539e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9442583430558443,
      "num_tokens": 59566032.0,
      "step": 8040
    },
    {
      "entropy": 1.97511575371027,
      "epoch": 0.12044936591077209,
      "grad_norm": 0.30620771646499634,
      "learning_rate": 4.877306994075569e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9460680555552244,
      "num_tokens": 59638638.0,
      "step": 8050
    },
    {
      "entropy": 2.012221439927816,
      "epoch": 0.12059899245227615,
      "grad_norm": 0.3005710244178772,
      "learning_rate": 4.8769351290776646e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9400858175009489,
      "num_tokens": 59712230.0,
      "step": 8060
    },
    {
      "entropy": 2.000872851908207,
      "epoch": 0.12074861899378021,
      "grad_norm": 0.3741800785064697,
      "learning_rate": 4.8765627156461676e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9425716027617455,
      "num_tokens": 59783956.0,
      "step": 8070
    },
    {
      "entropy": 1.9883229665458202,
      "epoch": 0.12089824553528428,
      "grad_norm": 0.2875189185142517,
      "learning_rate": 4.876189753867188e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9458576399832964,
      "num_tokens": 59857966.0,
      "step": 8080
    },
    {
      "entropy": 1.9670287810266018,
      "epoch": 0.12104787207678834,
      "grad_norm": 0.3263150751590729,
      "learning_rate": 4.875816243826961e-05,
      "loss": 0.2044,
      "mean_token_accuracy": 0.9494062159210443,
      "num_tokens": 59932912.0,
      "step": 8090
    },
    {
      "entropy": 1.9842902772128581,
      "epoch": 0.1211974986182924,
      "grad_norm": 0.388075590133667,
      "learning_rate": 4.875442185611848e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9413359355181455,
      "num_tokens": 60004261.0,
      "step": 8100
    },
    {
      "entropy": 2.007620833069086,
      "epoch": 0.12134712515979647,
      "grad_norm": 0.3924224376678467,
      "learning_rate": 4.8750675793083365e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9395794782787561,
      "num_tokens": 60076089.0,
      "step": 8110
    },
    {
      "entropy": 1.9913847222924232,
      "epoch": 0.12149675170130053,
      "grad_norm": 0.309541255235672,
      "learning_rate": 4.874692425003045e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9450324464589357,
      "num_tokens": 60151306.0,
      "step": 8120
    },
    {
      "entropy": 2.0059898398816585,
      "epoch": 0.1216463782428046,
      "grad_norm": 0.3752894401550293,
      "learning_rate": 4.8743167227827144e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.9418621595948935,
      "num_tokens": 60223069.0,
      "step": 8130
    },
    {
      "entropy": 1.9932882629334927,
      "epoch": 0.12179600478430866,
      "grad_norm": 0.34809330105781555,
      "learning_rate": 4.8739404727342144e-05,
      "loss": 0.2446,
      "mean_token_accuracy": 0.9416219037026167,
      "num_tokens": 60295102.0,
      "step": 8140
    },
    {
      "entropy": 2.0084681786596774,
      "epoch": 0.12194563132581272,
      "grad_norm": 0.2914487421512604,
      "learning_rate": 4.8735636749445394e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.9412002809345722,
      "num_tokens": 60368578.0,
      "step": 8150
    },
    {
      "entropy": 2.0299769453704357,
      "epoch": 0.1220952578673168,
      "grad_norm": 0.34598615765571594,
      "learning_rate": 4.8731863295008123e-05,
      "loss": 0.2472,
      "mean_token_accuracy": 0.9423439908772707,
      "num_tokens": 60441122.0,
      "step": 8160
    },
    {
      "entropy": 1.963605634868145,
      "epoch": 0.12224488440882086,
      "grad_norm": 0.25985008478164673,
      "learning_rate": 4.872808436490283e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9456941228359937,
      "num_tokens": 60516860.0,
      "step": 8170
    },
    {
      "entropy": 1.9466576144099235,
      "epoch": 0.12239451095032493,
      "grad_norm": 0.3242815434932709,
      "learning_rate": 4.8724299960003264e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9434987168759108,
      "num_tokens": 60590189.0,
      "step": 8180
    },
    {
      "entropy": 1.978720387071371,
      "epoch": 0.12254413749182899,
      "grad_norm": 0.3736899793148041,
      "learning_rate": 4.8720510081184445e-05,
      "loss": 0.2591,
      "mean_token_accuracy": 0.9377596624195575,
      "num_tokens": 60663502.0,
      "step": 8190
    },
    {
      "entropy": 1.9611873246729374,
      "epoch": 0.12269376403333306,
      "grad_norm": 0.3095279932022095,
      "learning_rate": 4.871671472932267e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.9423503495752812,
      "num_tokens": 60741492.0,
      "step": 8200
    },
    {
      "entropy": 1.9340177446603775,
      "epoch": 0.12284339057483712,
      "grad_norm": 0.37733903527259827,
      "learning_rate": 4.871291390529547e-05,
      "loss": 0.2022,
      "mean_token_accuracy": 0.950228575989604,
      "num_tokens": 60815537.0,
      "step": 8210
    },
    {
      "entropy": 1.9657929264009,
      "epoch": 0.12299301711634118,
      "grad_norm": 0.31044384837150574,
      "learning_rate": 4.8709107609981693e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9424935258924961,
      "num_tokens": 60891905.0,
      "step": 8220
    },
    {
      "entropy": 1.9878283530473708,
      "epoch": 0.12314264365784525,
      "grad_norm": 0.3591158390045166,
      "learning_rate": 4.8705295844261404e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9429921135306358,
      "num_tokens": 60966336.0,
      "step": 8230
    },
    {
      "entropy": 2.0191489554941655,
      "epoch": 0.12329227019934931,
      "grad_norm": 0.3593141734600067,
      "learning_rate": 4.8701478609015945e-05,
      "loss": 0.2486,
      "mean_token_accuracy": 0.9380486898124218,
      "num_tokens": 61039406.0,
      "step": 8240
    },
    {
      "entropy": 1.991200564801693,
      "epoch": 0.12344189674085337,
      "grad_norm": 0.34738439321517944,
      "learning_rate": 4.869765590512794e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9457454591989517,
      "num_tokens": 61115576.0,
      "step": 8250
    },
    {
      "entropy": 1.9950225219130515,
      "epoch": 0.12359152328235744,
      "grad_norm": 0.2894030511379242,
      "learning_rate": 4.869382773348126e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9479049224406481,
      "num_tokens": 61190313.0,
      "step": 8260
    },
    {
      "entropy": 1.9342955321073532,
      "epoch": 0.1237411498238615,
      "grad_norm": 0.34559860825538635,
      "learning_rate": 4.868999409496106e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9464694540947676,
      "num_tokens": 61266698.0,
      "step": 8270
    },
    {
      "entropy": 1.974451456964016,
      "epoch": 0.12389077636536557,
      "grad_norm": 0.30783629417419434,
      "learning_rate": 4.8686154990453734e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9425033900886775,
      "num_tokens": 61342997.0,
      "step": 8280
    },
    {
      "entropy": 1.9569523245096208,
      "epoch": 0.12404040290686963,
      "grad_norm": 0.34296104311943054,
      "learning_rate": 4.8682310420846945e-05,
      "loss": 0.2406,
      "mean_token_accuracy": 0.9422730423510075,
      "num_tokens": 61417826.0,
      "step": 8290
    },
    {
      "entropy": 1.9692927330732346,
      "epoch": 0.1241900294483737,
      "grad_norm": 0.33801189064979553,
      "learning_rate": 4.867846038702964e-05,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9448234625160694,
      "num_tokens": 61489462.0,
      "step": 8300
    },
    {
      "entropy": 1.9861831665039062,
      "epoch": 0.12433965598987777,
      "grad_norm": 0.28949058055877686,
      "learning_rate": 4.8674604889892e-05,
      "loss": 0.2434,
      "mean_token_accuracy": 0.9391167432069778,
      "num_tokens": 61563180.0,
      "step": 8310
    },
    {
      "entropy": 1.974212247878313,
      "epoch": 0.12448928253138183,
      "grad_norm": 0.33900365233421326,
      "learning_rate": 4.8670743930325504e-05,
      "loss": 0.2251,
      "mean_token_accuracy": 0.9438469767570495,
      "num_tokens": 61636792.0,
      "step": 8320
    },
    {
      "entropy": 1.9894649021327495,
      "epoch": 0.1246389090728859,
      "grad_norm": 0.3019603192806244,
      "learning_rate": 4.8666877509222856e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9452763896435499,
      "num_tokens": 61712724.0,
      "step": 8330
    },
    {
      "entropy": 1.98419756218791,
      "epoch": 0.12478853561438996,
      "grad_norm": 0.34396257996559143,
      "learning_rate": 4.866300562747805e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9397640496492385,
      "num_tokens": 61784426.0,
      "step": 8340
    },
    {
      "entropy": 1.9590938612818718,
      "epoch": 0.12493816215589403,
      "grad_norm": 0.2474859058856964,
      "learning_rate": 4.8659128285986336e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9466206256300211,
      "num_tokens": 61858840.0,
      "step": 8350
    },
    {
      "entropy": 1.980057169497013,
      "epoch": 0.1250877886973981,
      "grad_norm": 0.25387272238731384,
      "learning_rate": 4.8655245485644224e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9466661866754293,
      "num_tokens": 61930747.0,
      "step": 8360
    },
    {
      "entropy": 1.971390776336193,
      "epoch": 0.12523741523890214,
      "grad_norm": 0.3569907546043396,
      "learning_rate": 4.865135722734947e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9422731041908264,
      "num_tokens": 62005126.0,
      "step": 8370
    },
    {
      "entropy": 1.9663595296442509,
      "epoch": 0.12538704178040622,
      "grad_norm": 0.3019673526287079,
      "learning_rate": 4.864746351200114e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9442000284790992,
      "num_tokens": 62078363.0,
      "step": 8380
    },
    {
      "entropy": 1.9822776213288307,
      "epoch": 0.1255366683219103,
      "grad_norm": 0.3434697091579437,
      "learning_rate": 4.86435643404995e-05,
      "loss": 0.2456,
      "mean_token_accuracy": 0.9392819691449403,
      "num_tokens": 62150028.0,
      "step": 8390
    },
    {
      "entropy": 1.9773451015353203,
      "epoch": 0.12568629486341434,
      "grad_norm": 0.3503270447254181,
      "learning_rate": 4.863965971374612e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9415654823184013,
      "num_tokens": 62224857.0,
      "step": 8400
    },
    {
      "entropy": 1.9774516694247724,
      "epoch": 0.12583592140491842,
      "grad_norm": 0.31939077377319336,
      "learning_rate": 4.863574963264383e-05,
      "loss": 0.2348,
      "mean_token_accuracy": 0.9402480907738209,
      "num_tokens": 62299817.0,
      "step": 8410
    },
    {
      "entropy": 1.9766998663544655,
      "epoch": 0.12598554794642247,
      "grad_norm": 0.32501330971717834,
      "learning_rate": 4.86318340980967e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.944223378226161,
      "num_tokens": 62374989.0,
      "step": 8420
    },
    {
      "entropy": 1.9741141736507415,
      "epoch": 0.12613517448792655,
      "grad_norm": 0.39732980728149414,
      "learning_rate": 4.862791311101006e-05,
      "loss": 0.2612,
      "mean_token_accuracy": 0.9363386739045382,
      "num_tokens": 62448885.0,
      "step": 8430
    },
    {
      "entropy": 1.9976093538105488,
      "epoch": 0.1262848010294306,
      "grad_norm": 0.36971303820610046,
      "learning_rate": 4.862398667229053e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.9399280592799186,
      "num_tokens": 62522704.0,
      "step": 8440
    },
    {
      "entropy": 2.0156882248818873,
      "epoch": 0.12643442757093468,
      "grad_norm": 0.3868907392024994,
      "learning_rate": 4.862005478284596e-05,
      "loss": 0.2457,
      "mean_token_accuracy": 0.9397084049880504,
      "num_tokens": 62596028.0,
      "step": 8450
    },
    {
      "entropy": 1.9764956191182137,
      "epoch": 0.12658405411243873,
      "grad_norm": 0.3057219684123993,
      "learning_rate": 4.8616117443585486e-05,
      "loss": 0.2078,
      "mean_token_accuracy": 0.9495420183986425,
      "num_tokens": 62670102.0,
      "step": 8460
    },
    {
      "entropy": 1.9228065252304076,
      "epoch": 0.1267336806539428,
      "grad_norm": 0.35979214310646057,
      "learning_rate": 4.861217465541948e-05,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9445719916373492,
      "num_tokens": 62743599.0,
      "step": 8470
    },
    {
      "entropy": 1.9121742151677608,
      "epoch": 0.12688330719544685,
      "grad_norm": 0.40340152382850647,
      "learning_rate": 4.860822641925958e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9463967125862837,
      "num_tokens": 62818162.0,
      "step": 8480
    },
    {
      "entropy": 1.9151701487600803,
      "epoch": 0.12703293373695093,
      "grad_norm": 0.330057829618454,
      "learning_rate": 4.860427273601871e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9443833488970995,
      "num_tokens": 62890626.0,
      "step": 8490
    },
    {
      "entropy": 1.9346738435328006,
      "epoch": 0.12718256027845498,
      "grad_norm": 0.3881211280822754,
      "learning_rate": 4.8600313606611005e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9424592532217503,
      "num_tokens": 62963133.0,
      "step": 8500
    },
    {
      "entropy": 1.9652894951403141,
      "epoch": 0.12733218681995906,
      "grad_norm": 0.3115762770175934,
      "learning_rate": 4.859634903195191e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9442177213728428,
      "num_tokens": 63035573.0,
      "step": 8510
    },
    {
      "entropy": 1.9568308904767036,
      "epoch": 0.12748181336146314,
      "grad_norm": 0.2675648629665375,
      "learning_rate": 4.8592379012958095e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9401015724986792,
      "num_tokens": 63109590.0,
      "step": 8520
    },
    {
      "entropy": 1.9584609642624855,
      "epoch": 0.1276314399029672,
      "grad_norm": 0.32072627544403076,
      "learning_rate": 4.8588403550547494e-05,
      "loss": 0.2451,
      "mean_token_accuracy": 0.9389824490994215,
      "num_tokens": 63181119.0,
      "step": 8530
    },
    {
      "entropy": 1.9569327257573605,
      "epoch": 0.12778106644447126,
      "grad_norm": 0.32235363125801086,
      "learning_rate": 4.858442264563931e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9429894834756851,
      "num_tokens": 63255238.0,
      "step": 8540
    },
    {
      "entropy": 1.9441192679107189,
      "epoch": 0.12793069298597531,
      "grad_norm": 0.30793941020965576,
      "learning_rate": 4.8580436299154e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9416972581297159,
      "num_tokens": 63329349.0,
      "step": 8550
    },
    {
      "entropy": 1.9822468265891076,
      "epoch": 0.1280803195274794,
      "grad_norm": 0.25205153226852417,
      "learning_rate": 4.857644451201327e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9472266159951687,
      "num_tokens": 63401535.0,
      "step": 8560
    },
    {
      "entropy": 1.988240273296833,
      "epoch": 0.12822994606898344,
      "grad_norm": 0.7536655068397522,
      "learning_rate": 4.857244728514011e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9464554693549871,
      "num_tokens": 63475745.0,
      "step": 8570
    },
    {
      "entropy": 1.947568230330944,
      "epoch": 0.12837957261048752,
      "grad_norm": 0.3762826919555664,
      "learning_rate": 4.856844461945872e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9428681518882513,
      "num_tokens": 63549531.0,
      "step": 8580
    },
    {
      "entropy": 1.9757441848516464,
      "epoch": 0.12852919915199157,
      "grad_norm": 0.46594375371932983,
      "learning_rate": 4.8564436515894616e-05,
      "loss": 0.2538,
      "mean_token_accuracy": 0.9405520241707563,
      "num_tokens": 63624345.0,
      "step": 8590
    },
    {
      "entropy": 1.9864204876124858,
      "epoch": 0.12867882569349565,
      "grad_norm": 0.42728391289711,
      "learning_rate": 4.8560422975374536e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9395755406469106,
      "num_tokens": 63698537.0,
      "step": 8600
    },
    {
      "entropy": 1.9666790924966335,
      "epoch": 0.1288284522349997,
      "grad_norm": 0.35553333163261414,
      "learning_rate": 4.855640399882648e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9401620116084815,
      "num_tokens": 63771118.0,
      "step": 8610
    },
    {
      "entropy": 1.9578314311802387,
      "epoch": 0.12897807877650377,
      "grad_norm": 0.39668187499046326,
      "learning_rate": 4.8552379587179694e-05,
      "loss": 0.2479,
      "mean_token_accuracy": 0.9411626156419516,
      "num_tokens": 63846471.0,
      "step": 8620
    },
    {
      "entropy": 1.9691228158771992,
      "epoch": 0.12912770531800782,
      "grad_norm": 0.400918573141098,
      "learning_rate": 4.854834974136472e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9434386190026999,
      "num_tokens": 63919635.0,
      "step": 8630
    },
    {
      "entropy": 1.9797006890177726,
      "epoch": 0.1292773318595119,
      "grad_norm": 0.32591941952705383,
      "learning_rate": 4.854431446231331e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9438564013689756,
      "num_tokens": 63993647.0,
      "step": 8640
    },
    {
      "entropy": 1.985445711016655,
      "epoch": 0.12942695840101595,
      "grad_norm": 0.2766413390636444,
      "learning_rate": 4.854027375095849e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9435212541371584,
      "num_tokens": 64065422.0,
      "step": 8650
    },
    {
      "entropy": 1.9521594598889351,
      "epoch": 0.12957658494252003,
      "grad_norm": 0.36526623368263245,
      "learning_rate": 4.8536227608234564e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.945048525929451,
      "num_tokens": 64138632.0,
      "step": 8660
    },
    {
      "entropy": 1.9529690995812417,
      "epoch": 0.1297262114840241,
      "grad_norm": 0.2847371995449066,
      "learning_rate": 4.8532176035077054e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9371548358350992,
      "num_tokens": 64212992.0,
      "step": 8670
    },
    {
      "entropy": 1.9735078819096088,
      "epoch": 0.12987583802552816,
      "grad_norm": 0.7598893046379089,
      "learning_rate": 4.852811903242276e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9439747896045446,
      "num_tokens": 64288444.0,
      "step": 8680
    },
    {
      "entropy": 1.9900413297116757,
      "epoch": 0.13002546456703223,
      "grad_norm": 0.30402979254722595,
      "learning_rate": 4.852405660120975e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9427242401987315,
      "num_tokens": 64359164.0,
      "step": 8690
    },
    {
      "entropy": 1.9761988900601863,
      "epoch": 0.13017509110853628,
      "grad_norm": 0.3588244318962097,
      "learning_rate": 4.85199887423773e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9416104767471551,
      "num_tokens": 64432524.0,
      "step": 8700
    },
    {
      "entropy": 1.9465205788612365,
      "epoch": 0.13032471765004036,
      "grad_norm": 0.346652626991272,
      "learning_rate": 4.851591545686598e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9439924176782369,
      "num_tokens": 64506680.0,
      "step": 8710
    },
    {
      "entropy": 1.9636968933045864,
      "epoch": 0.1304743441915444,
      "grad_norm": 0.35844936966896057,
      "learning_rate": 4.851183674561762e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9411437481641769,
      "num_tokens": 64578859.0,
      "step": 8720
    },
    {
      "entropy": 1.9625427171587944,
      "epoch": 0.1306239707330485,
      "grad_norm": 0.2936355471611023,
      "learning_rate": 4.850775260957529e-05,
      "loss": 0.219,
      "mean_token_accuracy": 0.9431131657212972,
      "num_tokens": 64651074.0,
      "step": 8730
    },
    {
      "entropy": 1.9437426067888737,
      "epoch": 0.13077359727455254,
      "grad_norm": 0.26664796471595764,
      "learning_rate": 4.850366304968331e-05,
      "loss": 0.2315,
      "mean_token_accuracy": 0.9437053076922893,
      "num_tokens": 64726339.0,
      "step": 8740
    },
    {
      "entropy": 1.9261946514248849,
      "epoch": 0.13092322381605662,
      "grad_norm": 0.25835856795310974,
      "learning_rate": 4.849956806688725e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9456098906695842,
      "num_tokens": 64800390.0,
      "step": 8750
    },
    {
      "entropy": 1.9333353698253632,
      "epoch": 0.13107285035756067,
      "grad_norm": 0.3477896749973297,
      "learning_rate": 4.849546766213395e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9439190872013569,
      "num_tokens": 64875756.0,
      "step": 8760
    },
    {
      "entropy": 1.9803166374564172,
      "epoch": 0.13122247689906474,
      "grad_norm": 0.3693682849407196,
      "learning_rate": 4.8491361836371494e-05,
      "loss": 0.2472,
      "mean_token_accuracy": 0.9403591029345989,
      "num_tokens": 64949220.0,
      "step": 8770
    },
    {
      "entropy": 1.9510506846010685,
      "epoch": 0.1313721034405688,
      "grad_norm": 0.368045449256897,
      "learning_rate": 4.848725059054923e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9450150433927774,
      "num_tokens": 65023063.0,
      "step": 8780
    },
    {
      "entropy": 1.9711753949522972,
      "epoch": 0.13152172998207287,
      "grad_norm": 0.3556571304798126,
      "learning_rate": 4.8483133925617744e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.9398034173995257,
      "num_tokens": 65097830.0,
      "step": 8790
    },
    {
      "entropy": 1.9685128338634967,
      "epoch": 0.13167135652357692,
      "grad_norm": 0.3265618681907654,
      "learning_rate": 4.8479011842528885e-05,
      "loss": 0.2476,
      "mean_token_accuracy": 0.9389249615371227,
      "num_tokens": 65168882.0,
      "step": 8800
    },
    {
      "entropy": 1.9480599597096444,
      "epoch": 0.131820983065081,
      "grad_norm": 0.3606414198875427,
      "learning_rate": 4.8474884342235756e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9424274686723948,
      "num_tokens": 65245162.0,
      "step": 8810
    },
    {
      "entropy": 1.9709650918841362,
      "epoch": 0.13197060960658508,
      "grad_norm": 0.3648437559604645,
      "learning_rate": 4.8470751425692694e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9458903253078461,
      "num_tokens": 65317847.0,
      "step": 8820
    },
    {
      "entropy": 1.9740999616682529,
      "epoch": 0.13212023614808913,
      "grad_norm": 0.381575345993042,
      "learning_rate": 4.846661309385532e-05,
      "loss": 0.2513,
      "mean_token_accuracy": 0.9397554129362107,
      "num_tokens": 65389230.0,
      "step": 8830
    },
    {
      "entropy": 1.957600075751543,
      "epoch": 0.1322698626895932,
      "grad_norm": 0.24571581184864044,
      "learning_rate": 4.846246934768047e-05,
      "loss": 0.2481,
      "mean_token_accuracy": 0.9400928698480129,
      "num_tokens": 65461623.0,
      "step": 8840
    },
    {
      "entropy": 1.9673214673995971,
      "epoch": 0.13241948923109725,
      "grad_norm": 0.28099003434181213,
      "learning_rate": 4.845832018812627e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9440156791359187,
      "num_tokens": 65536396.0,
      "step": 8850
    },
    {
      "entropy": 1.969651285558939,
      "epoch": 0.13256911577260133,
      "grad_norm": 0.37584996223449707,
      "learning_rate": 4.845416561615207e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9477973401546478,
      "num_tokens": 65611860.0,
      "step": 8860
    },
    {
      "entropy": 1.9568484649062157,
      "epoch": 0.13271874231410538,
      "grad_norm": 0.33930739760398865,
      "learning_rate": 4.84500056327185e-05,
      "loss": 0.2333,
      "mean_token_accuracy": 0.9430764563381672,
      "num_tokens": 65683517.0,
      "step": 8870
    },
    {
      "entropy": 1.944619332998991,
      "epoch": 0.13286836885560946,
      "grad_norm": 0.4050168991088867,
      "learning_rate": 4.8445840238787396e-05,
      "loss": 0.2603,
      "mean_token_accuracy": 0.9399784963577986,
      "num_tokens": 65755427.0,
      "step": 8880
    },
    {
      "entropy": 1.959000463783741,
      "epoch": 0.1330179953971135,
      "grad_norm": 0.3912133276462555,
      "learning_rate": 4.844166943532188e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.943807939067483,
      "num_tokens": 65832991.0,
      "step": 8890
    },
    {
      "entropy": 1.9239009767770767,
      "epoch": 0.1331676219386176,
      "grad_norm": 0.330244243144989,
      "learning_rate": 4.843749322328632e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9459849793463946,
      "num_tokens": 65907618.0,
      "step": 8900
    },
    {
      "entropy": 1.942949976027012,
      "epoch": 0.13331724848012164,
      "grad_norm": 0.4121508002281189,
      "learning_rate": 4.8433311603646317e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9438702866435051,
      "num_tokens": 65981665.0,
      "step": 8910
    },
    {
      "entropy": 1.9682985998690128,
      "epoch": 0.1334668750216257,
      "grad_norm": 0.2904795706272125,
      "learning_rate": 4.842912457736876e-05,
      "loss": 0.231,
      "mean_token_accuracy": 0.941966163739562,
      "num_tokens": 66055836.0,
      "step": 8920
    },
    {
      "entropy": 1.9720845118165016,
      "epoch": 0.13361650156312976,
      "grad_norm": 0.293618768453598,
      "learning_rate": 4.842493214542175e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9437258597463369,
      "num_tokens": 66131212.0,
      "step": 8930
    },
    {
      "entropy": 1.9385132156312466,
      "epoch": 0.13376612810463384,
      "grad_norm": 0.2917327582836151,
      "learning_rate": 4.8420734308774655e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9486747324466706,
      "num_tokens": 66205858.0,
      "step": 8940
    },
    {
      "entropy": 1.9383563965559005,
      "epoch": 0.13391575464613792,
      "grad_norm": 0.3826505243778229,
      "learning_rate": 4.841653106839808e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9395735554397107,
      "num_tokens": 66280012.0,
      "step": 8950
    },
    {
      "entropy": 1.9099912263453007,
      "epoch": 0.13406538118764197,
      "grad_norm": 0.34202826023101807,
      "learning_rate": 4.84123224252639e-05,
      "loss": 0.1976,
      "mean_token_accuracy": 0.9502735130488873,
      "num_tokens": 66355311.0,
      "step": 8960
    },
    {
      "entropy": 1.885715789347887,
      "epoch": 0.13421500772914605,
      "grad_norm": 0.42412397265434265,
      "learning_rate": 4.840810838034523e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9424956567585469,
      "num_tokens": 66430508.0,
      "step": 8970
    },
    {
      "entropy": 1.8879559002816677,
      "epoch": 0.1343646342706501,
      "grad_norm": 0.40387916564941406,
      "learning_rate": 4.840388893461641e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9418501652777195,
      "num_tokens": 66503574.0,
      "step": 8980
    },
    {
      "entropy": 1.917030405253172,
      "epoch": 0.13451426081215417,
      "grad_norm": 0.3102967441082001,
      "learning_rate": 4.8399664089053085e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9448459714651107,
      "num_tokens": 66580316.0,
      "step": 8990
    },
    {
      "entropy": 1.9436975404620171,
      "epoch": 0.13466388735365822,
      "grad_norm": 0.3890587091445923,
      "learning_rate": 4.8395433844632074e-05,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9456178724765778,
      "num_tokens": 66655430.0,
      "step": 9000
    },
    {
      "epoch": 0.13466388735365822,
      "eval_entropy": 1.9488787143468858,
      "eval_loss": 0.25563159584999084,
      "eval_mean_token_accuracy": 0.9433935329318046,
      "eval_num_tokens": 66655430.0,
      "eval_runtime": 545.2853,
      "eval_samples_per_second": 36.678,
      "eval_steps_per_second": 9.17,
      "step": 9000
    },
    {
      "entropy": 1.9483337968587875,
      "epoch": 0.1348135138951623,
      "grad_norm": 0.3011627495288849,
      "learning_rate": 4.839119820233153e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9459673687815666,
      "num_tokens": 66729719.0,
      "step": 9010
    },
    {
      "entropy": 1.9654812686145307,
      "epoch": 0.13496314043666635,
      "grad_norm": 0.31736207008361816,
      "learning_rate": 4.838695716313078e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9453720789402723,
      "num_tokens": 66801812.0,
      "step": 9020
    },
    {
      "entropy": 1.9585977770388125,
      "epoch": 0.13511276697817043,
      "grad_norm": 0.27295032143592834,
      "learning_rate": 4.838271072801043e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9412912257015705,
      "num_tokens": 66875970.0,
      "step": 9030
    },
    {
      "entropy": 1.970735015720129,
      "epoch": 0.13526239351967448,
      "grad_norm": 0.3993614912033081,
      "learning_rate": 4.8378458897952336e-05,
      "loss": 0.2397,
      "mean_token_accuracy": 0.9421390369534492,
      "num_tokens": 66949110.0,
      "step": 9040
    },
    {
      "entropy": 1.9808103710412979,
      "epoch": 0.13541202006117856,
      "grad_norm": 0.3647657036781311,
      "learning_rate": 4.837420167393961e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9441422343254089,
      "num_tokens": 67023299.0,
      "step": 9050
    },
    {
      "entropy": 1.9768221914768218,
      "epoch": 0.1355616466026826,
      "grad_norm": 0.37747153639793396,
      "learning_rate": 4.836993905695657e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9398123681545257,
      "num_tokens": 67096505.0,
      "step": 9060
    },
    {
      "entropy": 1.9879881963133812,
      "epoch": 0.13571127314418668,
      "grad_norm": 0.3683839440345764,
      "learning_rate": 4.8365671047988825e-05,
      "loss": 0.2417,
      "mean_token_accuracy": 0.9429855939000845,
      "num_tokens": 67167639.0,
      "step": 9070
    },
    {
      "entropy": 1.9715983614325523,
      "epoch": 0.13586089968569073,
      "grad_norm": 0.36192578077316284,
      "learning_rate": 4.836139764802321e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9431883241981268,
      "num_tokens": 67241224.0,
      "step": 9080
    },
    {
      "entropy": 1.9246913336217404,
      "epoch": 0.1360105262271948,
      "grad_norm": 0.35266897082328796,
      "learning_rate": 4.835711885804782e-05,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9453090026974678,
      "num_tokens": 67316377.0,
      "step": 9090
    },
    {
      "entropy": 1.9322432443499564,
      "epoch": 0.1361601527686989,
      "grad_norm": 0.34745660424232483,
      "learning_rate": 4.835283467905198e-05,
      "loss": 0.2487,
      "mean_token_accuracy": 0.9406419951468706,
      "num_tokens": 67388526.0,
      "step": 9100
    },
    {
      "entropy": 1.9272822305560111,
      "epoch": 0.13630977931020294,
      "grad_norm": 0.2859017252922058,
      "learning_rate": 4.834854511202627e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9472344610840082,
      "num_tokens": 67465106.0,
      "step": 9110
    },
    {
      "entropy": 1.9539577946066857,
      "epoch": 0.13645940585170702,
      "grad_norm": 0.3548308312892914,
      "learning_rate": 4.8344250157962504e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9439939834177494,
      "num_tokens": 67539732.0,
      "step": 9120
    },
    {
      "entropy": 1.977692713588476,
      "epoch": 0.13660903239321107,
      "grad_norm": 0.3132769465446472,
      "learning_rate": 4.833994981785377e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9426475591957569,
      "num_tokens": 67611979.0,
      "step": 9130
    },
    {
      "entropy": 1.949458361417055,
      "epoch": 0.13675865893471514,
      "grad_norm": 0.3377133905887604,
      "learning_rate": 4.833564409269437e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9496698271483183,
      "num_tokens": 67686569.0,
      "step": 9140
    },
    {
      "entropy": 1.9355627067387104,
      "epoch": 0.1369082854762192,
      "grad_norm": 0.3213011920452118,
      "learning_rate": 4.833133298347987e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9421371828764677,
      "num_tokens": 67761774.0,
      "step": 9150
    },
    {
      "entropy": 1.9401809476315974,
      "epoch": 0.13705791201772327,
      "grad_norm": 0.32401928305625916,
      "learning_rate": 4.8327016491207075e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9449021417647601,
      "num_tokens": 67839957.0,
      "step": 9160
    },
    {
      "entropy": 1.9816595241427422,
      "epoch": 0.13720753855922732,
      "grad_norm": 0.3979884088039398,
      "learning_rate": 4.832269461687403e-05,
      "loss": 0.2358,
      "mean_token_accuracy": 0.9417010087519884,
      "num_tokens": 67916378.0,
      "step": 9170
    },
    {
      "entropy": 1.9495150744915009,
      "epoch": 0.1373571651007314,
      "grad_norm": 0.3539530634880066,
      "learning_rate": 4.831836736148004e-05,
      "loss": 0.205,
      "mean_token_accuracy": 0.9505281843245029,
      "num_tokens": 67990229.0,
      "step": 9180
    },
    {
      "entropy": 1.9691606126725674,
      "epoch": 0.13750679164223545,
      "grad_norm": 0.44842231273651123,
      "learning_rate": 4.831403472602563e-05,
      "loss": 0.2326,
      "mean_token_accuracy": 0.9405070416629314,
      "num_tokens": 68061907.0,
      "step": 9190
    },
    {
      "entropy": 1.9491675458848476,
      "epoch": 0.13765641818373953,
      "grad_norm": 0.32721734046936035,
      "learning_rate": 4.8309696711512594e-05,
      "loss": 0.2195,
      "mean_token_accuracy": 0.9451438773423433,
      "num_tokens": 68135653.0,
      "step": 9200
    },
    {
      "entropy": 1.9939965434372424,
      "epoch": 0.13780604472524358,
      "grad_norm": 0.353633850812912,
      "learning_rate": 4.830535331894395e-05,
      "loss": 0.2566,
      "mean_token_accuracy": 0.937218252569437,
      "num_tokens": 68206742.0,
      "step": 9210
    },
    {
      "entropy": 1.9459476798772812,
      "epoch": 0.13795567126674765,
      "grad_norm": 0.31654617190361023,
      "learning_rate": 4.830100454932397e-05,
      "loss": 0.2097,
      "mean_token_accuracy": 0.9479497592896223,
      "num_tokens": 68281111.0,
      "step": 9220
    },
    {
      "entropy": 1.9277146644890308,
      "epoch": 0.1381052978082517,
      "grad_norm": 0.3213835656642914,
      "learning_rate": 4.829665040365817e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9485133994370699,
      "num_tokens": 68357239.0,
      "step": 9230
    },
    {
      "entropy": 1.9786666579544545,
      "epoch": 0.13825492434975578,
      "grad_norm": 0.4365866184234619,
      "learning_rate": 4.82922908829533e-05,
      "loss": 0.2684,
      "mean_token_accuracy": 0.9337970435619354,
      "num_tokens": 68430584.0,
      "step": 9240
    },
    {
      "entropy": 1.9404628470540046,
      "epoch": 0.13840455089125986,
      "grad_norm": 0.3002017140388489,
      "learning_rate": 4.8287925988217365e-05,
      "loss": 0.198,
      "mean_token_accuracy": 0.9502225946635008,
      "num_tokens": 68507158.0,
      "step": 9250
    },
    {
      "entropy": 1.951265572756529,
      "epoch": 0.1385541774327639,
      "grad_norm": 0.3586018979549408,
      "learning_rate": 4.82835557204596e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9417462438344956,
      "num_tokens": 68581231.0,
      "step": 9260
    },
    {
      "entropy": 1.9776667036116122,
      "epoch": 0.138703803974268,
      "grad_norm": 0.3314315378665924,
      "learning_rate": 4.827918008069049e-05,
      "loss": 0.2596,
      "mean_token_accuracy": 0.9385028939694167,
      "num_tokens": 68654225.0,
      "step": 9270
    },
    {
      "entropy": 1.9743095569312572,
      "epoch": 0.13885343051577204,
      "grad_norm": 0.29083719849586487,
      "learning_rate": 4.827479906992177e-05,
      "loss": 0.2467,
      "mean_token_accuracy": 0.9407394751906395,
      "num_tokens": 68728882.0,
      "step": 9280
    },
    {
      "entropy": 1.9809728540480136,
      "epoch": 0.1390030570572761,
      "grad_norm": 0.25985953211784363,
      "learning_rate": 4.827041268916639e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9479349613189697,
      "num_tokens": 68807811.0,
      "step": 9290
    },
    {
      "entropy": 1.9703376397490502,
      "epoch": 0.13915268359878016,
      "grad_norm": 0.36498844623565674,
      "learning_rate": 4.826602093943858e-05,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9456536643207073,
      "num_tokens": 68880398.0,
      "step": 9300
    },
    {
      "entropy": 1.9479986652731895,
      "epoch": 0.13930231014028424,
      "grad_norm": 0.27892500162124634,
      "learning_rate": 4.826162382175378e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9468015041202307,
      "num_tokens": 68951729.0,
      "step": 9310
    },
    {
      "entropy": 1.9210216835141183,
      "epoch": 0.1394519366817883,
      "grad_norm": 0.3597976565361023,
      "learning_rate": 4.8257221337128674e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.9406485389918089,
      "num_tokens": 69024086.0,
      "step": 9320
    },
    {
      "entropy": 1.9612861469388008,
      "epoch": 0.13960156322329237,
      "grad_norm": 0.31205272674560547,
      "learning_rate": 4.82528134865812e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9433183468878269,
      "num_tokens": 69095614.0,
      "step": 9330
    },
    {
      "entropy": 1.9877513609826565,
      "epoch": 0.13975118976479642,
      "grad_norm": 0.37088486552238464,
      "learning_rate": 4.824840027113055e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9408521112054586,
      "num_tokens": 69167626.0,
      "step": 9340
    },
    {
      "entropy": 1.98278241828084,
      "epoch": 0.1399008163063005,
      "grad_norm": 0.36493024230003357,
      "learning_rate": 4.8243981691797106e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9410956777632237,
      "num_tokens": 69239211.0,
      "step": 9350
    },
    {
      "entropy": 1.9933070227503777,
      "epoch": 0.14005044284780455,
      "grad_norm": 0.31293556094169617,
      "learning_rate": 4.823955774960255e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9408943265676498,
      "num_tokens": 69312805.0,
      "step": 9360
    },
    {
      "entropy": 1.958333135396242,
      "epoch": 0.14020006938930862,
      "grad_norm": 0.31713417172431946,
      "learning_rate": 4.8235128445569754e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9468443047255277,
      "num_tokens": 69388158.0,
      "step": 9370
    },
    {
      "entropy": 1.9638317182660103,
      "epoch": 0.14034969593081267,
      "grad_norm": 0.3242224454879761,
      "learning_rate": 4.823069378072287e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9442442692816257,
      "num_tokens": 69463918.0,
      "step": 9380
    },
    {
      "entropy": 1.986253874003887,
      "epoch": 0.14049932247231675,
      "grad_norm": 0.3108769357204437,
      "learning_rate": 4.822625375608726e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.942235691845417,
      "num_tokens": 69535509.0,
      "step": 9390
    },
    {
      "entropy": 1.9511330835521221,
      "epoch": 0.14064894901382083,
      "grad_norm": 0.3007988929748535,
      "learning_rate": 4.8221808372689545e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9486309554427862,
      "num_tokens": 69611068.0,
      "step": 9400
    },
    {
      "entropy": 1.951339303702116,
      "epoch": 0.14079857555532488,
      "grad_norm": 0.3171694874763489,
      "learning_rate": 4.821735763155756e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9465548057109118,
      "num_tokens": 69685186.0,
      "step": 9410
    },
    {
      "entropy": 1.953355387598276,
      "epoch": 0.14094820209682896,
      "grad_norm": 0.29984259605407715,
      "learning_rate": 4.821290153372042e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9436756044626236,
      "num_tokens": 69760338.0,
      "step": 9420
    },
    {
      "entropy": 1.9204163372516632,
      "epoch": 0.141097828638333,
      "grad_norm": 0.29189491271972656,
      "learning_rate": 4.820844008020844e-05,
      "loss": 0.2531,
      "mean_token_accuracy": 0.9399379406124353,
      "num_tokens": 69834564.0,
      "step": 9430
    },
    {
      "entropy": 1.9474875345826148,
      "epoch": 0.14124745517983708,
      "grad_norm": 0.3574906885623932,
      "learning_rate": 4.820397327205318e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9390805486589671,
      "num_tokens": 69908034.0,
      "step": 9440
    },
    {
      "entropy": 1.9551872149109841,
      "epoch": 0.14139708172134113,
      "grad_norm": 0.35252702236175537,
      "learning_rate": 4.819950111028746e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9408782787621022,
      "num_tokens": 69980070.0,
      "step": 9450
    },
    {
      "entropy": 1.9115084245800973,
      "epoch": 0.1415467082628452,
      "grad_norm": 0.3842913508415222,
      "learning_rate": 4.819502359594533e-05,
      "loss": 0.2301,
      "mean_token_accuracy": 0.941709129139781,
      "num_tokens": 70054460.0,
      "step": 9460
    },
    {
      "entropy": 1.8893527925014495,
      "epoch": 0.14169633480434926,
      "grad_norm": 0.3047802746295929,
      "learning_rate": 4.8190540730062036e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9447365637868643,
      "num_tokens": 70128759.0,
      "step": 9470
    },
    {
      "entropy": 1.9327017694711686,
      "epoch": 0.14184596134585334,
      "grad_norm": 0.33507052063941956,
      "learning_rate": 4.818605251367414e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9416280146688223,
      "num_tokens": 70202817.0,
      "step": 9480
    },
    {
      "entropy": 1.9426468513906001,
      "epoch": 0.1419955878873574,
      "grad_norm": 0.35069891810417175,
      "learning_rate": 4.818155894781936e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9464975386857987,
      "num_tokens": 70276096.0,
      "step": 9490
    },
    {
      "entropy": 1.9518881015479566,
      "epoch": 0.14214521442886147,
      "grad_norm": 0.3290864825248718,
      "learning_rate": 4.817706003353671e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9452050339430571,
      "num_tokens": 70351479.0,
      "step": 9500
    },
    {
      "entropy": 1.970929603278637,
      "epoch": 0.14229484097036552,
      "grad_norm": 0.34592166543006897,
      "learning_rate": 4.817255577186642e-05,
      "loss": 0.2429,
      "mean_token_accuracy": 0.9381695881485939,
      "num_tokens": 70425876.0,
      "step": 9510
    },
    {
      "entropy": 1.9552265472710133,
      "epoch": 0.1424444675118696,
      "grad_norm": 0.32449761033058167,
      "learning_rate": 4.816804616384995e-05,
      "loss": 0.2324,
      "mean_token_accuracy": 0.9449277278035879,
      "num_tokens": 70499060.0,
      "step": 9520
    },
    {
      "entropy": 2.0136697702109814,
      "epoch": 0.14259409405337367,
      "grad_norm": 0.3378818929195404,
      "learning_rate": 4.816353121052999e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9438359316438436,
      "num_tokens": 70574024.0,
      "step": 9530
    },
    {
      "entropy": 1.95682624951005,
      "epoch": 0.14274372059487772,
      "grad_norm": 0.2584880590438843,
      "learning_rate": 4.81590109129505e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.944714231044054,
      "num_tokens": 70648369.0,
      "step": 9540
    },
    {
      "entropy": 1.9372911527752876,
      "epoch": 0.1428933471363818,
      "grad_norm": 0.31865113973617554,
      "learning_rate": 4.815448527215663e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9439224701374769,
      "num_tokens": 70722875.0,
      "step": 9550
    },
    {
      "entropy": 1.9243810951709748,
      "epoch": 0.14304297367788585,
      "grad_norm": 0.413815438747406,
      "learning_rate": 4.814995428919481e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9394006479531527,
      "num_tokens": 70795601.0,
      "step": 9560
    },
    {
      "entropy": 1.9187439888715745,
      "epoch": 0.14319260021938993,
      "grad_norm": 0.286629319190979,
      "learning_rate": 4.814541796511267e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.9442979108542204,
      "num_tokens": 70871964.0,
      "step": 9570
    },
    {
      "entropy": 1.9147816732525826,
      "epoch": 0.14334222676089398,
      "grad_norm": 0.3704766631126404,
      "learning_rate": 4.814087630095909e-05,
      "loss": 0.2429,
      "mean_token_accuracy": 0.9406113352626562,
      "num_tokens": 70945352.0,
      "step": 9580
    },
    {
      "entropy": 1.942322688549757,
      "epoch": 0.14349185330239805,
      "grad_norm": 0.4035572409629822,
      "learning_rate": 4.813632929778419e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9434312421828508,
      "num_tokens": 71022681.0,
      "step": 9590
    },
    {
      "entropy": 1.9356301598250867,
      "epoch": 0.1436414798439021,
      "grad_norm": 0.32776445150375366,
      "learning_rate": 4.8131776956639315e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9457470294088125,
      "num_tokens": 71098359.0,
      "step": 9600
    },
    {
      "entropy": 1.9147076666355134,
      "epoch": 0.14379110638540618,
      "grad_norm": 0.36388885974884033,
      "learning_rate": 4.812721927857705e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9398420371115208,
      "num_tokens": 71170989.0,
      "step": 9610
    },
    {
      "entropy": 1.882090973854065,
      "epoch": 0.14394073292691023,
      "grad_norm": 0.3057663142681122,
      "learning_rate": 4.812265626465121e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9435542799532414,
      "num_tokens": 71244688.0,
      "step": 9620
    },
    {
      "entropy": 1.9075574971735478,
      "epoch": 0.1440903594684143,
      "grad_norm": 0.30157583951950073,
      "learning_rate": 4.811808791591683e-05,
      "loss": 0.2262,
      "mean_token_accuracy": 0.9443697270005942,
      "num_tokens": 71319184.0,
      "step": 9630
    },
    {
      "entropy": 1.89164879322052,
      "epoch": 0.14423998600991836,
      "grad_norm": 0.37271201610565186,
      "learning_rate": 4.8113514233430216e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9475700814276934,
      "num_tokens": 71393750.0,
      "step": 9640
    },
    {
      "entropy": 1.9133434474468232,
      "epoch": 0.14438961255142244,
      "grad_norm": 0.32016822695732117,
      "learning_rate": 4.810893521824888e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9441170435398817,
      "num_tokens": 71468750.0,
      "step": 9650
    },
    {
      "entropy": 1.9390408918261528,
      "epoch": 0.14453923909292649,
      "grad_norm": 0.31049180030822754,
      "learning_rate": 4.810435087143155e-05,
      "loss": 0.2051,
      "mean_token_accuracy": 0.9521675903350115,
      "num_tokens": 71543368.0,
      "step": 9660
    },
    {
      "entropy": 1.9402896292507648,
      "epoch": 0.14468886563443056,
      "grad_norm": 0.3493243157863617,
      "learning_rate": 4.809976119403824e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9445060800760985,
      "num_tokens": 71617937.0,
      "step": 9670
    },
    {
      "entropy": 1.928921538591385,
      "epoch": 0.14483849217593464,
      "grad_norm": 0.35896921157836914,
      "learning_rate": 4.8095166187130144e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9461550883948803,
      "num_tokens": 71693335.0,
      "step": 9680
    },
    {
      "entropy": 1.90728582367301,
      "epoch": 0.1449881187174387,
      "grad_norm": 0.2883624732494354,
      "learning_rate": 4.8090565851769726e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9413074858486652,
      "num_tokens": 71767892.0,
      "step": 9690
    },
    {
      "entropy": 1.9456420816481113,
      "epoch": 0.14513774525894277,
      "grad_norm": 0.36232709884643555,
      "learning_rate": 4.8085960189020644e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9420846290886402,
      "num_tokens": 71841878.0,
      "step": 9700
    },
    {
      "entropy": 1.9357500843703748,
      "epoch": 0.14528737180044682,
      "grad_norm": 0.3139575719833374,
      "learning_rate": 4.808134919994782e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9423670586198568,
      "num_tokens": 71914722.0,
      "step": 9710
    },
    {
      "entropy": 1.9303450793027879,
      "epoch": 0.1454369983419509,
      "grad_norm": 0.3181741237640381,
      "learning_rate": 4.8076732885617406e-05,
      "loss": 0.2059,
      "mean_token_accuracy": 0.9462727215141058,
      "num_tokens": 71988909.0,
      "step": 9720
    },
    {
      "entropy": 1.9360411137342453,
      "epoch": 0.14558662488345495,
      "grad_norm": 0.33108997344970703,
      "learning_rate": 4.807211124709676e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.943679092824459,
      "num_tokens": 72061144.0,
      "step": 9730
    },
    {
      "entropy": 1.9307071655988692,
      "epoch": 0.14573625142495902,
      "grad_norm": 0.360953152179718,
      "learning_rate": 4.80674842854545e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9412772305309772,
      "num_tokens": 72134390.0,
      "step": 9740
    },
    {
      "entropy": 1.9613996863365173,
      "epoch": 0.14588587796646307,
      "grad_norm": 0.2520855963230133,
      "learning_rate": 4.806285200176045e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.944409403204918,
      "num_tokens": 72209576.0,
      "step": 9750
    },
    {
      "entropy": 1.9498137004673481,
      "epoch": 0.14603550450796715,
      "grad_norm": 0.3211948871612549,
      "learning_rate": 4.805821439708569e-05,
      "loss": 0.2122,
      "mean_token_accuracy": 0.947629764676094,
      "num_tokens": 72285525.0,
      "step": 9760
    },
    {
      "entropy": 1.9193257749080659,
      "epoch": 0.1461851310494712,
      "grad_norm": 0.3163833022117615,
      "learning_rate": 4.80535714725025e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9416702345013619,
      "num_tokens": 72358358.0,
      "step": 9770
    },
    {
      "entropy": 1.9338900044560432,
      "epoch": 0.14633475759097528,
      "grad_norm": 0.34347960352897644,
      "learning_rate": 4.804892322908442e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.945580180734396,
      "num_tokens": 72435095.0,
      "step": 9780
    },
    {
      "entropy": 1.9366588167846204,
      "epoch": 0.14648438413247933,
      "grad_norm": 0.31336262822151184,
      "learning_rate": 4.80442696679062e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9434244647622109,
      "num_tokens": 72511427.0,
      "step": 9790
    },
    {
      "entropy": 1.9242891520261765,
      "epoch": 0.1466340106739834,
      "grad_norm": 0.3171961307525635,
      "learning_rate": 4.803961079004383e-05,
      "loss": 0.2102,
      "mean_token_accuracy": 0.9478205628693104,
      "num_tokens": 72588228.0,
      "step": 9800
    },
    {
      "entropy": 1.9427091382443904,
      "epoch": 0.14678363721548746,
      "grad_norm": 0.29199811816215515,
      "learning_rate": 4.803494659657452e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9438135959208012,
      "num_tokens": 72662552.0,
      "step": 9810
    },
    {
      "entropy": 1.9504983954131603,
      "epoch": 0.14693326375699153,
      "grad_norm": 0.281848281621933,
      "learning_rate": 4.803027708857673e-05,
      "loss": 0.2159,
      "mean_token_accuracy": 0.9442804489284754,
      "num_tokens": 72737781.0,
      "step": 9820
    },
    {
      "entropy": 1.9234144739806651,
      "epoch": 0.1470828902984956,
      "grad_norm": 0.3392612934112549,
      "learning_rate": 4.802560226713012e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9445745177567005,
      "num_tokens": 72813053.0,
      "step": 9830
    },
    {
      "entropy": 1.9352587692439556,
      "epoch": 0.14723251683999966,
      "grad_norm": 0.3580477237701416,
      "learning_rate": 4.802092213331558e-05,
      "loss": 0.2641,
      "mean_token_accuracy": 0.9382480446249246,
      "num_tokens": 72886466.0,
      "step": 9840
    },
    {
      "entropy": 1.9515813373029232,
      "epoch": 0.14738214338150374,
      "grad_norm": 0.32320213317871094,
      "learning_rate": 4.801623668821526e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9468538921326399,
      "num_tokens": 72963216.0,
      "step": 9850
    },
    {
      "entropy": 1.9166609413921833,
      "epoch": 0.1475317699230078,
      "grad_norm": 0.3752511143684387,
      "learning_rate": 4.801154593291252e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9402341291308403,
      "num_tokens": 73036539.0,
      "step": 9860
    },
    {
      "entropy": 1.91079076603055,
      "epoch": 0.14768139646451187,
      "grad_norm": 0.34397321939468384,
      "learning_rate": 4.800684986849193e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9436853427439928,
      "num_tokens": 73109440.0,
      "step": 9870
    },
    {
      "entropy": 1.9356561787426472,
      "epoch": 0.14783102300601592,
      "grad_norm": 0.3052953779697418,
      "learning_rate": 4.800214849603931e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9470526415854692,
      "num_tokens": 73184990.0,
      "step": 9880
    },
    {
      "entropy": 1.9489891722798347,
      "epoch": 0.14798064954752,
      "grad_norm": 0.3975132703781128,
      "learning_rate": 4.799744181664171e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9412062238901854,
      "num_tokens": 73258482.0,
      "step": 9890
    },
    {
      "entropy": 1.9365462079644202,
      "epoch": 0.14813027608902404,
      "grad_norm": 0.4482927620410919,
      "learning_rate": 4.7992729831387386e-05,
      "loss": 0.22,
      "mean_token_accuracy": 0.9461224015802145,
      "num_tokens": 73330435.0,
      "step": 9900
    },
    {
      "entropy": 1.885791876167059,
      "epoch": 0.14827990263052812,
      "grad_norm": 0.3477848172187805,
      "learning_rate": 4.798801254136584e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9464099742472172,
      "num_tokens": 73405957.0,
      "step": 9910
    },
    {
      "entropy": 1.9172332286834717,
      "epoch": 0.14842952917203217,
      "grad_norm": 0.36400294303894043,
      "learning_rate": 4.798328994766779e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9457256682217121,
      "num_tokens": 73479729.0,
      "step": 9920
    },
    {
      "entropy": 1.9637786597013474,
      "epoch": 0.14857915571353625,
      "grad_norm": 0.333106130361557,
      "learning_rate": 4.7978562051385184e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.9398229122161865,
      "num_tokens": 73555138.0,
      "step": 9930
    },
    {
      "entropy": 1.951609492301941,
      "epoch": 0.1487287822550403,
      "grad_norm": 0.332767516374588,
      "learning_rate": 4.7973828853611196e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9455506943166256,
      "num_tokens": 73628761.0,
      "step": 9940
    },
    {
      "entropy": 1.932226374000311,
      "epoch": 0.14887840879654438,
      "grad_norm": 0.30349162220954895,
      "learning_rate": 4.796909035544022e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9437657959759236,
      "num_tokens": 73703967.0,
      "step": 9950
    },
    {
      "entropy": 1.992712678015232,
      "epoch": 0.14902803533804845,
      "grad_norm": 0.36695024371147156,
      "learning_rate": 4.7964346557967894e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9439554255455732,
      "num_tokens": 73778880.0,
      "step": 9960
    },
    {
      "entropy": 1.9757874220609666,
      "epoch": 0.1491776618795525,
      "grad_norm": 0.37456706166267395,
      "learning_rate": 4.795959746229106e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9443719271570444,
      "num_tokens": 73852833.0,
      "step": 9970
    },
    {
      "entropy": 1.965779746323824,
      "epoch": 0.14932728842105658,
      "grad_norm": 0.3689342439174652,
      "learning_rate": 4.795484306950779e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9431554324924946,
      "num_tokens": 73926444.0,
      "step": 9980
    },
    {
      "entropy": 1.9717483691871167,
      "epoch": 0.14947691496256063,
      "grad_norm": 0.30721649527549744,
      "learning_rate": 4.795008338071739e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.945174703001976,
      "num_tokens": 74000217.0,
      "step": 9990
    },
    {
      "entropy": 2.0146261431276797,
      "epoch": 0.1496265415040647,
      "grad_norm": 0.31610745191574097,
      "learning_rate": 4.794531839702039e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9427457839250565,
      "num_tokens": 74073489.0,
      "step": 10000
    },
    {
      "epoch": 0.1496265415040647,
      "eval_entropy": 2.0023350311994554,
      "eval_loss": 0.2541130483150482,
      "eval_mean_token_accuracy": 0.9434884492874146,
      "eval_num_tokens": 74073489.0,
      "eval_runtime": 547.2649,
      "eval_samples_per_second": 36.545,
      "eval_steps_per_second": 9.136,
      "step": 10000
    },
    {
      "entropy": 2.0096257492899894,
      "epoch": 0.14977616804556876,
      "grad_norm": 0.41175577044487,
      "learning_rate": 4.794054811951852e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9445735428482294,
      "num_tokens": 74145452.0,
      "step": 10010
    },
    {
      "entropy": 1.96277888789773,
      "epoch": 0.14992579458707284,
      "grad_norm": 0.38437095284461975,
      "learning_rate": 4.793577254931477e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9410177625715732,
      "num_tokens": 74218173.0,
      "step": 10020
    },
    {
      "entropy": 1.9713745146989823,
      "epoch": 0.15007542112857689,
      "grad_norm": 0.4264318346977234,
      "learning_rate": 4.793099168751334e-05,
      "loss": 0.2552,
      "mean_token_accuracy": 0.9402187038213015,
      "num_tokens": 74289714.0,
      "step": 10030
    },
    {
      "entropy": 2.0118980705738068,
      "epoch": 0.15022504767008096,
      "grad_norm": 0.3127113878726959,
      "learning_rate": 4.7926205535219644e-05,
      "loss": 0.249,
      "mean_token_accuracy": 0.9399946581572294,
      "num_tokens": 74363240.0,
      "step": 10040
    },
    {
      "entropy": 2.0039468236267566,
      "epoch": 0.150374674211585,
      "grad_norm": 0.3341583013534546,
      "learning_rate": 4.792141409354031e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9471007533371448,
      "num_tokens": 74440048.0,
      "step": 10050
    },
    {
      "entropy": 1.9907129302620887,
      "epoch": 0.1505243007530891,
      "grad_norm": 0.32159239053726196,
      "learning_rate": 4.7916617363583236e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9476275566965342,
      "num_tokens": 74515465.0,
      "step": 10060
    },
    {
      "entropy": 1.9853538058698177,
      "epoch": 0.15067392729459314,
      "grad_norm": 0.31087031960487366,
      "learning_rate": 4.791181534645748e-05,
      "loss": 0.2692,
      "mean_token_accuracy": 0.9375228494405746,
      "num_tokens": 74590380.0,
      "step": 10070
    },
    {
      "entropy": 2.044235463440418,
      "epoch": 0.15082355383609722,
      "grad_norm": 0.4257383644580841,
      "learning_rate": 4.790700804327337e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9427407011389732,
      "num_tokens": 74664814.0,
      "step": 10080
    },
    {
      "entropy": 2.0140422962605955,
      "epoch": 0.15097318037760127,
      "grad_norm": 0.3344215154647827,
      "learning_rate": 4.790219545514244e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9469365004450083,
      "num_tokens": 74739249.0,
      "step": 10090
    },
    {
      "entropy": 1.9841054931282998,
      "epoch": 0.15112280691910535,
      "grad_norm": 0.4196813404560089,
      "learning_rate": 4.789737758317744e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9458187758922577,
      "num_tokens": 74814142.0,
      "step": 10100
    },
    {
      "entropy": 2.0130374021828175,
      "epoch": 0.15127243346060942,
      "grad_norm": 0.25859564542770386,
      "learning_rate": 4.7892554428492355e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9443059764802456,
      "num_tokens": 74889368.0,
      "step": 10110
    },
    {
      "entropy": 1.9955432578921317,
      "epoch": 0.15142206000211347,
      "grad_norm": 0.2633899748325348,
      "learning_rate": 4.788772599220238e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9447741836309433,
      "num_tokens": 74968209.0,
      "step": 10120
    },
    {
      "entropy": 2.027172787487507,
      "epoch": 0.15157168654361755,
      "grad_norm": 0.3068169057369232,
      "learning_rate": 4.7882892275423944e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9453667365014553,
      "num_tokens": 75041250.0,
      "step": 10130
    },
    {
      "entropy": 2.00287516489625,
      "epoch": 0.1517213130851216,
      "grad_norm": 0.28688597679138184,
      "learning_rate": 4.787805327927466e-05,
      "loss": 0.202,
      "mean_token_accuracy": 0.9475355301052332,
      "num_tokens": 75118005.0,
      "step": 10140
    },
    {
      "entropy": 1.9912812247872353,
      "epoch": 0.15187093962662568,
      "grad_norm": 0.3301286995410919,
      "learning_rate": 4.787320900487342e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9416725657880306,
      "num_tokens": 75192995.0,
      "step": 10150
    },
    {
      "entropy": 1.984282349795103,
      "epoch": 0.15202056616812973,
      "grad_norm": 0.36690813302993774,
      "learning_rate": 4.7868359453340306e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9486179735511542,
      "num_tokens": 75266359.0,
      "step": 10160
    },
    {
      "entropy": 1.9733359031379223,
      "epoch": 0.1521701927096338,
      "grad_norm": 0.3084751069545746,
      "learning_rate": 4.78635046257966e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.944027767330408,
      "num_tokens": 75339651.0,
      "step": 10170
    },
    {
      "entropy": 1.9452820114791394,
      "epoch": 0.15231981925113786,
      "grad_norm": 0.4038937985897064,
      "learning_rate": 4.7858644523364836e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9453101187944413,
      "num_tokens": 75413711.0,
      "step": 10180
    },
    {
      "entropy": 1.9682258747518062,
      "epoch": 0.15246944579264193,
      "grad_norm": 0.3249014914035797,
      "learning_rate": 4.785377914716876e-05,
      "loss": 0.2481,
      "mean_token_accuracy": 0.9381590314209461,
      "num_tokens": 75489487.0,
      "step": 10190
    },
    {
      "entropy": 1.9567937403917313,
      "epoch": 0.15261907233414598,
      "grad_norm": 0.40018409490585327,
      "learning_rate": 4.7848908498333326e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.9459907744079828,
      "num_tokens": 75562406.0,
      "step": 10200
    },
    {
      "entropy": 1.9470438182353973,
      "epoch": 0.15276869887565006,
      "grad_norm": 0.44914180040359497,
      "learning_rate": 4.784403257798472e-05,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9474494758993387,
      "num_tokens": 75637282.0,
      "step": 10210
    },
    {
      "entropy": 1.9817422211170197,
      "epoch": 0.1529183254171541,
      "grad_norm": 0.39441195130348206,
      "learning_rate": 4.7839151387250345e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9444709490984678,
      "num_tokens": 75713969.0,
      "step": 10220
    },
    {
      "entropy": 1.9814308978617192,
      "epoch": 0.1530679519586582,
      "grad_norm": 0.4253191351890564,
      "learning_rate": 4.78342649272588e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9437417417764664,
      "num_tokens": 75786449.0,
      "step": 10230
    },
    {
      "entropy": 1.9705582112073898,
      "epoch": 0.15321757850016224,
      "grad_norm": 0.38298362493515015,
      "learning_rate": 4.7829373199139946e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9441553324460983,
      "num_tokens": 75860517.0,
      "step": 10240
    },
    {
      "entropy": 1.961432795971632,
      "epoch": 0.15336720504166632,
      "grad_norm": 0.3164595663547516,
      "learning_rate": 4.782447620402482e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9451293453574181,
      "num_tokens": 75936023.0,
      "step": 10250
    },
    {
      "entropy": 1.9676277592778206,
      "epoch": 0.1535168315831704,
      "grad_norm": 0.34096574783325195,
      "learning_rate": 4.7819573943045695e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9452019326388836,
      "num_tokens": 76010359.0,
      "step": 10260
    },
    {
      "entropy": 1.9231061466038226,
      "epoch": 0.15366645812467444,
      "grad_norm": 0.3452071249485016,
      "learning_rate": 4.781466641733608e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.947759136930108,
      "num_tokens": 76084169.0,
      "step": 10270
    },
    {
      "entropy": 1.9195040717720986,
      "epoch": 0.15381608466617852,
      "grad_norm": 0.35254359245300293,
      "learning_rate": 4.7809753628030663e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.950443197414279,
      "num_tokens": 76159506.0,
      "step": 10280
    },
    {
      "entropy": 1.9417597696185112,
      "epoch": 0.15396571120768257,
      "grad_norm": 0.32364755868911743,
      "learning_rate": 4.780483557626537e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9476958364248276,
      "num_tokens": 76234246.0,
      "step": 10290
    },
    {
      "entropy": 1.9783832170069218,
      "epoch": 0.15411533774918665,
      "grad_norm": 0.3573232591152191,
      "learning_rate": 4.7799912263177354e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9468444090336561,
      "num_tokens": 76308556.0,
      "step": 10300
    },
    {
      "entropy": 1.9693367302417755,
      "epoch": 0.1542649642906907,
      "grad_norm": 0.3826085031032562,
      "learning_rate": 4.779498368990497e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9394370969384909,
      "num_tokens": 76381224.0,
      "step": 10310
    },
    {
      "entropy": 1.9444942571222783,
      "epoch": 0.15441459083219478,
      "grad_norm": 0.3278908133506775,
      "learning_rate": 4.779004985758777e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9443727150559426,
      "num_tokens": 76455875.0,
      "step": 10320
    },
    {
      "entropy": 1.9720619566738606,
      "epoch": 0.15456421737369883,
      "grad_norm": 0.3902450501918793,
      "learning_rate": 4.7785110767366576e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9407512798905373,
      "num_tokens": 76527498.0,
      "step": 10330
    },
    {
      "entropy": 1.9660998113453387,
      "epoch": 0.1547138439152029,
      "grad_norm": 0.3909050226211548,
      "learning_rate": 4.778016642038337e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9423815131187439,
      "num_tokens": 76600289.0,
      "step": 10340
    },
    {
      "entropy": 1.9942153520882129,
      "epoch": 0.15486347045670695,
      "grad_norm": 0.3505004048347473,
      "learning_rate": 4.777521681778138e-05,
      "loss": 0.2375,
      "mean_token_accuracy": 0.9432413939386606,
      "num_tokens": 76674964.0,
      "step": 10350
    },
    {
      "entropy": 1.9955847837030887,
      "epoch": 0.15501309699821103,
      "grad_norm": 0.3075355887413025,
      "learning_rate": 4.7770261960705046e-05,
      "loss": 0.2173,
      "mean_token_accuracy": 0.9458872582763433,
      "num_tokens": 76749212.0,
      "step": 10360
    },
    {
      "entropy": 1.9781630665063858,
      "epoch": 0.15516272353971508,
      "grad_norm": 0.5010891556739807,
      "learning_rate": 4.776530185030001e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9421981893479824,
      "num_tokens": 76826068.0,
      "step": 10370
    },
    {
      "entropy": 1.9448955468833447,
      "epoch": 0.15531235008121916,
      "grad_norm": 0.31132784485816956,
      "learning_rate": 4.776033648771315e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9477094251662492,
      "num_tokens": 76900784.0,
      "step": 10380
    },
    {
      "entropy": 1.9730580471456052,
      "epoch": 0.1554619766227232,
      "grad_norm": 0.32480528950691223,
      "learning_rate": 4.7755365874092544e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9397799544036388,
      "num_tokens": 76975783.0,
      "step": 10390
    },
    {
      "entropy": 1.9658220052719115,
      "epoch": 0.15561160316422729,
      "grad_norm": 0.3758400082588196,
      "learning_rate": 4.7750390010587473e-05,
      "loss": 0.2431,
      "mean_token_accuracy": 0.9406151827424765,
      "num_tokens": 77048362.0,
      "step": 10400
    },
    {
      "entropy": 1.9778623804450035,
      "epoch": 0.15576122970573136,
      "grad_norm": 0.3254123628139496,
      "learning_rate": 4.774540889834847e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9480778209865093,
      "num_tokens": 77121836.0,
      "step": 10410
    },
    {
      "entropy": 1.935928375273943,
      "epoch": 0.1559108562472354,
      "grad_norm": 0.36201465129852295,
      "learning_rate": 4.7740422538527225e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9439206875860691,
      "num_tokens": 77200511.0,
      "step": 10420
    },
    {
      "entropy": 1.9682085797190667,
      "epoch": 0.1560604827887395,
      "grad_norm": 0.3414900302886963,
      "learning_rate": 4.7735430932276696e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9455634739249945,
      "num_tokens": 77273849.0,
      "step": 10430
    },
    {
      "entropy": 1.9416078969836235,
      "epoch": 0.15621010933024354,
      "grad_norm": 0.3483978509902954,
      "learning_rate": 4.773043408075102e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9449137635529041,
      "num_tokens": 77346544.0,
      "step": 10440
    },
    {
      "entropy": 1.9785201512277126,
      "epoch": 0.15635973587174762,
      "grad_norm": 0.3955497145652771,
      "learning_rate": 4.772543198510556e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9469719376415014,
      "num_tokens": 77420226.0,
      "step": 10450
    },
    {
      "entropy": 2.0132290981709957,
      "epoch": 0.15650936241325167,
      "grad_norm": 0.36287185549736023,
      "learning_rate": 4.7720424646496905e-05,
      "loss": 0.2493,
      "mean_token_accuracy": 0.9358021222054959,
      "num_tokens": 77495676.0,
      "step": 10460
    },
    {
      "entropy": 2.0146973818540572,
      "epoch": 0.15665898895475575,
      "grad_norm": 0.36396801471710205,
      "learning_rate": 4.771541206608282e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.9452680945396423,
      "num_tokens": 77568766.0,
      "step": 10470
    },
    {
      "entropy": 2.0186705641448497,
      "epoch": 0.1568086154962598,
      "grad_norm": 0.3387264013290405,
      "learning_rate": 4.771039424502232e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9423900414258242,
      "num_tokens": 77642064.0,
      "step": 10480
    },
    {
      "entropy": 1.973207425326109,
      "epoch": 0.15695824203776387,
      "grad_norm": 0.3171241283416748,
      "learning_rate": 4.770537118447559e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9448492452502251,
      "num_tokens": 77716321.0,
      "step": 10490
    },
    {
      "entropy": 2.012285700440407,
      "epoch": 0.15710786857926792,
      "grad_norm": 0.2792969346046448,
      "learning_rate": 4.7700342885604066e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9414599124342203,
      "num_tokens": 77791143.0,
      "step": 10500
    },
    {
      "entropy": 1.9563880451023579,
      "epoch": 0.157257495120772,
      "grad_norm": 0.3146459460258484,
      "learning_rate": 4.7695309349570384e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.94557551369071,
      "num_tokens": 77865517.0,
      "step": 10510
    },
    {
      "entropy": 1.9683196663856506,
      "epoch": 0.15740712166227605,
      "grad_norm": 0.31142696738243103,
      "learning_rate": 4.769027057753839e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9453274752944708,
      "num_tokens": 77942415.0,
      "step": 10520
    },
    {
      "entropy": 2.0028581485152244,
      "epoch": 0.15755674820378013,
      "grad_norm": 0.3106737434864044,
      "learning_rate": 4.768522657067313e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.9446032267063856,
      "num_tokens": 78013607.0,
      "step": 10530
    },
    {
      "entropy": 1.991431586444378,
      "epoch": 0.1577063747452842,
      "grad_norm": 0.41271111369132996,
      "learning_rate": 4.7680177330140864e-05,
      "loss": 0.2343,
      "mean_token_accuracy": 0.9441709138453007,
      "num_tokens": 78088094.0,
      "step": 10540
    },
    {
      "entropy": 2.0153347618877886,
      "epoch": 0.15785600128678826,
      "grad_norm": 0.356224000453949,
      "learning_rate": 4.767512285710906e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9443151004612446,
      "num_tokens": 78162858.0,
      "step": 10550
    },
    {
      "entropy": 2.0088924184441566,
      "epoch": 0.15800562782829233,
      "grad_norm": 0.3527098596096039,
      "learning_rate": 4.7670063152746436e-05,
      "loss": 0.2497,
      "mean_token_accuracy": 0.9393941190093755,
      "num_tokens": 78236392.0,
      "step": 10560
    },
    {
      "entropy": 2.0209972254931925,
      "epoch": 0.15815525436979638,
      "grad_norm": 0.3115454614162445,
      "learning_rate": 4.7664998218222846e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.94420220926404,
      "num_tokens": 78311048.0,
      "step": 10570
    },
    {
      "entropy": 1.9964174672961235,
      "epoch": 0.15830488091130046,
      "grad_norm": 0.4216262996196747,
      "learning_rate": 4.7659928054709415e-05,
      "loss": 0.2484,
      "mean_token_accuracy": 0.9414950624108315,
      "num_tokens": 78385869.0,
      "step": 10580
    },
    {
      "entropy": 1.9904039442539214,
      "epoch": 0.1584545074528045,
      "grad_norm": 0.40144437551498413,
      "learning_rate": 4.765485266337845e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9442124094814062,
      "num_tokens": 78459652.0,
      "step": 10590
    },
    {
      "entropy": 1.9699851073324681,
      "epoch": 0.1586041339943086,
      "grad_norm": 0.3338618576526642,
      "learning_rate": 4.7649772045403476e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9430190443992614,
      "num_tokens": 78533879.0,
      "step": 10600
    },
    {
      "entropy": 1.9537183180451394,
      "epoch": 0.15875376053581264,
      "grad_norm": 0.33143433928489685,
      "learning_rate": 4.7644686201959214e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9432450734078884,
      "num_tokens": 78607206.0,
      "step": 10610
    },
    {
      "entropy": 1.941721798479557,
      "epoch": 0.15890338707731672,
      "grad_norm": 0.34236812591552734,
      "learning_rate": 4.76395951342216e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9421915169805288,
      "num_tokens": 78681555.0,
      "step": 10620
    },
    {
      "entropy": 1.989346131682396,
      "epoch": 0.15905301361882077,
      "grad_norm": 0.3201507031917572,
      "learning_rate": 4.7634498843367796e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9434388693422079,
      "num_tokens": 78753899.0,
      "step": 10630
    },
    {
      "entropy": 2.0069540180265903,
      "epoch": 0.15920264016032484,
      "grad_norm": 0.38621047139167786,
      "learning_rate": 4.762939733057614e-05,
      "loss": 0.2103,
      "mean_token_accuracy": 0.943544851243496,
      "num_tokens": 78827981.0,
      "step": 10640
    },
    {
      "entropy": 1.9651928126811982,
      "epoch": 0.1593522667018289,
      "grad_norm": 0.31219255924224854,
      "learning_rate": 4.762429059702619e-05,
      "loss": 0.231,
      "mean_token_accuracy": 0.9450302854180336,
      "num_tokens": 78901733.0,
      "step": 10650
    },
    {
      "entropy": 1.9831410489976407,
      "epoch": 0.15950189324333297,
      "grad_norm": 0.26025286316871643,
      "learning_rate": 4.7619178643898725e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9404673781245947,
      "num_tokens": 78975108.0,
      "step": 10660
    },
    {
      "entropy": 1.978158713877201,
      "epoch": 0.15965151978483702,
      "grad_norm": 0.4043017029762268,
      "learning_rate": 4.7614061472375715e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9456560432910919,
      "num_tokens": 79049229.0,
      "step": 10670
    },
    {
      "entropy": 1.9537462018430234,
      "epoch": 0.1598011463263411,
      "grad_norm": 0.27361536026000977,
      "learning_rate": 4.760893908364034e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9488750763237477,
      "num_tokens": 79124867.0,
      "step": 10680
    },
    {
      "entropy": 1.9781516090035438,
      "epoch": 0.15995077286784518,
      "grad_norm": 0.41842177510261536,
      "learning_rate": 4.760381147887698e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9431587524712086,
      "num_tokens": 79196736.0,
      "step": 10690
    },
    {
      "entropy": 1.9668852277100086,
      "epoch": 0.16010039940934923,
      "grad_norm": 0.33304363489151,
      "learning_rate": 4.759867865927124e-05,
      "loss": 0.2228,
      "mean_token_accuracy": 0.9462146919220686,
      "num_tokens": 79270489.0,
      "step": 10700
    },
    {
      "entropy": 1.970092374831438,
      "epoch": 0.1602500259508533,
      "grad_norm": 0.3183971643447876,
      "learning_rate": 4.759354062600993e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.946943337470293,
      "num_tokens": 79346060.0,
      "step": 10710
    },
    {
      "entropy": 1.9573637008666993,
      "epoch": 0.16039965249235735,
      "grad_norm": 0.5096936821937561,
      "learning_rate": 4.7588397380281024e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.9416359167546033,
      "num_tokens": 79416908.0,
      "step": 10720
    },
    {
      "entropy": 1.9613184303045272,
      "epoch": 0.16054927903386143,
      "grad_norm": 0.32115495204925537,
      "learning_rate": 4.7583248923273746e-05,
      "loss": 0.1964,
      "mean_token_accuracy": 0.9490230828523636,
      "num_tokens": 79492025.0,
      "step": 10730
    },
    {
      "entropy": 1.9644240103662014,
      "epoch": 0.16069890557536548,
      "grad_norm": 0.3410327434539795,
      "learning_rate": 4.7578095256178526e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.945114828273654,
      "num_tokens": 79566212.0,
      "step": 10740
    },
    {
      "entropy": 1.952405869215727,
      "epoch": 0.16084853211686956,
      "grad_norm": 0.40386343002319336,
      "learning_rate": 4.757293638018697e-05,
      "loss": 0.236,
      "mean_token_accuracy": 0.9415817927569151,
      "num_tokens": 79639653.0,
      "step": 10750
    },
    {
      "entropy": 1.9546678058803082,
      "epoch": 0.1609981586583736,
      "grad_norm": 0.37465283274650574,
      "learning_rate": 4.7567772296491895e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9458628330379725,
      "num_tokens": 79713793.0,
      "step": 10760
    },
    {
      "entropy": 1.9670131176710128,
      "epoch": 0.16114778519987769,
      "grad_norm": 0.34774649143218994,
      "learning_rate": 4.7562603006287335e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9412055384367705,
      "num_tokens": 79790490.0,
      "step": 10770
    },
    {
      "entropy": 1.9672094404697418,
      "epoch": 0.16129741174138174,
      "grad_norm": 0.33496513962745667,
      "learning_rate": 4.755742851076853e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9471564702689648,
      "num_tokens": 79864554.0,
      "step": 10780
    },
    {
      "entropy": 1.9885177597403527,
      "epoch": 0.1614470382828858,
      "grad_norm": 0.30325421690940857,
      "learning_rate": 4.7552248811131896e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.945768577978015,
      "num_tokens": 79940683.0,
      "step": 10790
    },
    {
      "entropy": 1.975594438612461,
      "epoch": 0.16159666482438986,
      "grad_norm": 0.355462908744812,
      "learning_rate": 4.75470639085751e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9456601209938527,
      "num_tokens": 80013159.0,
      "step": 10800
    },
    {
      "entropy": 1.9485169537365437,
      "epoch": 0.16174629136589394,
      "grad_norm": 0.34689298272132874,
      "learning_rate": 4.754187380429695e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9436494067311287,
      "num_tokens": 80083557.0,
      "step": 10810
    },
    {
      "entropy": 1.971580981463194,
      "epoch": 0.161895917907398,
      "grad_norm": 0.4138765335083008,
      "learning_rate": 4.753667849949752e-05,
      "loss": 0.2534,
      "mean_token_accuracy": 0.9404000699520111,
      "num_tokens": 80155087.0,
      "step": 10820
    },
    {
      "entropy": 2.004280689358711,
      "epoch": 0.16204554444890207,
      "grad_norm": 0.3035665452480316,
      "learning_rate": 4.753147799537804e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9413384258747101,
      "num_tokens": 80229524.0,
      "step": 10830
    },
    {
      "entropy": 1.9797514848411084,
      "epoch": 0.16219517099040615,
      "grad_norm": 0.33744123578071594,
      "learning_rate": 4.7526272293140975e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9379522424191237,
      "num_tokens": 80303836.0,
      "step": 10840
    },
    {
      "entropy": 1.958918285369873,
      "epoch": 0.1623447975319102,
      "grad_norm": 0.3103746473789215,
      "learning_rate": 4.7521061393989964e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9470954809337855,
      "num_tokens": 80378194.0,
      "step": 10850
    },
    {
      "entropy": 1.9701179198920726,
      "epoch": 0.16249442407341427,
      "grad_norm": 0.4412486255168915,
      "learning_rate": 4.7515845299129854e-05,
      "loss": 0.2445,
      "mean_token_accuracy": 0.9429891966283321,
      "num_tokens": 80452563.0,
      "step": 10860
    },
    {
      "entropy": 1.949977456778288,
      "epoch": 0.16264405061491832,
      "grad_norm": 0.37699609994888306,
      "learning_rate": 4.75106240097667e-05,
      "loss": 0.2142,
      "mean_token_accuracy": 0.9463824927806854,
      "num_tokens": 80526005.0,
      "step": 10870
    },
    {
      "entropy": 1.9736156642436982,
      "epoch": 0.1627936771564224,
      "grad_norm": 0.27534738183021545,
      "learning_rate": 4.7505397527107776e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9467197686433793,
      "num_tokens": 80603133.0,
      "step": 10880
    },
    {
      "entropy": 2.000217464566231,
      "epoch": 0.16294330369792645,
      "grad_norm": 0.33747947216033936,
      "learning_rate": 4.750016585236152e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.9405313212424516,
      "num_tokens": 80676285.0,
      "step": 10890
    },
    {
      "entropy": 1.9920235708355905,
      "epoch": 0.16309293023943053,
      "grad_norm": 0.33371323347091675,
      "learning_rate": 4.749492898673758e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9417155481874943,
      "num_tokens": 80750781.0,
      "step": 10900
    },
    {
      "entropy": 1.9735712334513664,
      "epoch": 0.16324255678093458,
      "grad_norm": 0.35176360607147217,
      "learning_rate": 4.7489686931446843e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9438793353736401,
      "num_tokens": 80824533.0,
      "step": 10910
    },
    {
      "entropy": 2.008169595897198,
      "epoch": 0.16339218332243866,
      "grad_norm": 0.2957383990287781,
      "learning_rate": 4.7484439687701336e-05,
      "loss": 0.2195,
      "mean_token_accuracy": 0.9459058258682489,
      "num_tokens": 80902321.0,
      "step": 10920
    },
    {
      "entropy": 1.9874602533876895,
      "epoch": 0.1635418098639427,
      "grad_norm": 0.31329163908958435,
      "learning_rate": 4.7479187256714336e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9472841840237379,
      "num_tokens": 80978387.0,
      "step": 10930
    },
    {
      "entropy": 1.968823428452015,
      "epoch": 0.16369143640544678,
      "grad_norm": 0.3770764172077179,
      "learning_rate": 4.747392963970028e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.945463977009058,
      "num_tokens": 81054543.0,
      "step": 10940
    },
    {
      "entropy": 1.941917323321104,
      "epoch": 0.16384106294695083,
      "grad_norm": 0.407034307718277,
      "learning_rate": 4.746866683787484e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.9455674681812525,
      "num_tokens": 81126519.0,
      "step": 10950
    },
    {
      "entropy": 1.968162039667368,
      "epoch": 0.1639906894884549,
      "grad_norm": 0.4003603756427765,
      "learning_rate": 4.746339885245486e-05,
      "loss": 0.2481,
      "mean_token_accuracy": 0.9375172656029462,
      "num_tokens": 81202184.0,
      "step": 10960
    },
    {
      "entropy": 1.983988807350397,
      "epoch": 0.164140316029959,
      "grad_norm": 0.34252604842185974,
      "learning_rate": 4.745812568465839e-05,
      "loss": 0.2398,
      "mean_token_accuracy": 0.9432325001806021,
      "num_tokens": 81274197.0,
      "step": 10970
    },
    {
      "entropy": 2.0024105235934258,
      "epoch": 0.16428994257146304,
      "grad_norm": 0.36099374294281006,
      "learning_rate": 4.745284733570468e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9424055811017752,
      "num_tokens": 81347282.0,
      "step": 10980
    },
    {
      "entropy": 1.9782360464334487,
      "epoch": 0.16443956911296712,
      "grad_norm": 0.24694211781024933,
      "learning_rate": 4.7447563806814185e-05,
      "loss": 0.2077,
      "mean_token_accuracy": 0.9477418273687362,
      "num_tokens": 81423730.0,
      "step": 10990
    },
    {
      "entropy": 1.9813645020127297,
      "epoch": 0.16458919565447117,
      "grad_norm": 0.4146910309791565,
      "learning_rate": 4.744227509920855e-05,
      "loss": 0.2422,
      "mean_token_accuracy": 0.9429894223809242,
      "num_tokens": 81496003.0,
      "step": 11000
    },
    {
      "epoch": 0.16458919565447117,
      "eval_entropy": 2.004666271686554,
      "eval_loss": 0.25224271416664124,
      "eval_mean_token_accuracy": 0.9437370450496674,
      "eval_num_tokens": 81496003.0,
      "eval_runtime": 543.6838,
      "eval_samples_per_second": 36.786,
      "eval_steps_per_second": 9.197,
      "step": 11000
    },
    {
      "entropy": 1.9927356079220773,
      "epoch": 0.16473882219597524,
      "grad_norm": 0.33182159066200256,
      "learning_rate": 4.743698121411061e-05,
      "loss": 0.209,
      "mean_token_accuracy": 0.9461609624326229,
      "num_tokens": 81569623.0,
      "step": 11010
    },
    {
      "entropy": 2.0417598888278006,
      "epoch": 0.1648884487374793,
      "grad_norm": 0.32985419034957886,
      "learning_rate": 4.7431682152744414e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9359782136976719,
      "num_tokens": 81644163.0,
      "step": 11020
    },
    {
      "entropy": 2.046129982918501,
      "epoch": 0.16503807527898337,
      "grad_norm": 0.4606972634792328,
      "learning_rate": 4.74263779163352e-05,
      "loss": 0.2545,
      "mean_token_accuracy": 0.9384373672306537,
      "num_tokens": 81716254.0,
      "step": 11030
    },
    {
      "entropy": 2.00350741147995,
      "epoch": 0.16518770182048742,
      "grad_norm": 0.3878551721572876,
      "learning_rate": 4.7421068506109394e-05,
      "loss": 0.2489,
      "mean_token_accuracy": 0.9392802346497774,
      "num_tokens": 81788120.0,
      "step": 11040
    },
    {
      "entropy": 2.003596521168947,
      "epoch": 0.1653373283619915,
      "grad_norm": 0.3278098404407501,
      "learning_rate": 4.741575392329462e-05,
      "loss": 0.2449,
      "mean_token_accuracy": 0.9429786529392004,
      "num_tokens": 81862624.0,
      "step": 11050
    },
    {
      "entropy": 2.0163043096661566,
      "epoch": 0.16548695490349555,
      "grad_norm": 0.3412734270095825,
      "learning_rate": 4.741043416911974e-05,
      "loss": 0.2541,
      "mean_token_accuracy": 0.935042281076312,
      "num_tokens": 81936836.0,
      "step": 11060
    },
    {
      "entropy": 1.989291739463806,
      "epoch": 0.16563658144499963,
      "grad_norm": 0.34780606627464294,
      "learning_rate": 4.740510924481473e-05,
      "loss": 0.2589,
      "mean_token_accuracy": 0.9357319053262472,
      "num_tokens": 82009898.0,
      "step": 11070
    },
    {
      "entropy": 1.9842719361186028,
      "epoch": 0.16578620798650368,
      "grad_norm": 0.34832021594047546,
      "learning_rate": 4.739977915161083e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9419632039964199,
      "num_tokens": 82085868.0,
      "step": 11080
    },
    {
      "entropy": 2.022157220542431,
      "epoch": 0.16593583452800775,
      "grad_norm": 0.31946173310279846,
      "learning_rate": 4.7394443890740456e-05,
      "loss": 0.2451,
      "mean_token_accuracy": 0.9380356334149837,
      "num_tokens": 82160104.0,
      "step": 11090
    },
    {
      "entropy": 1.9865759693086147,
      "epoch": 0.1660854610695118,
      "grad_norm": 0.3690767288208008,
      "learning_rate": 4.7389103463437196e-05,
      "loss": 0.2449,
      "mean_token_accuracy": 0.940768975764513,
      "num_tokens": 82235915.0,
      "step": 11100
    },
    {
      "entropy": 1.9949030950665474,
      "epoch": 0.16623508761101588,
      "grad_norm": 0.2824763357639313,
      "learning_rate": 4.738375787093587e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9438487235456705,
      "num_tokens": 82309623.0,
      "step": 11110
    },
    {
      "entropy": 2.0076543390750885,
      "epoch": 0.16638471415251996,
      "grad_norm": 0.3778853714466095,
      "learning_rate": 4.737840711447248e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9425571031868458,
      "num_tokens": 82387678.0,
      "step": 11120
    },
    {
      "entropy": 2.030814906209707,
      "epoch": 0.166534340694024,
      "grad_norm": 0.3743649423122406,
      "learning_rate": 4.7373051195284186e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9437940642237663,
      "num_tokens": 82461894.0,
      "step": 11130
    },
    {
      "entropy": 2.0106933273375036,
      "epoch": 0.16668396723552809,
      "grad_norm": 0.44694194197654724,
      "learning_rate": 4.73676901146094e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9449918989092112,
      "num_tokens": 82537066.0,
      "step": 11140
    },
    {
      "entropy": 1.9608184173703194,
      "epoch": 0.16683359377703214,
      "grad_norm": 0.3679609000682831,
      "learning_rate": 4.73623238736877e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9452104546129704,
      "num_tokens": 82615122.0,
      "step": 11150
    },
    {
      "entropy": 1.9855125054717064,
      "epoch": 0.1669832203185362,
      "grad_norm": 0.40195468068122864,
      "learning_rate": 4.7356952473759835e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9458672877401113,
      "num_tokens": 82687518.0,
      "step": 11160
    },
    {
      "entropy": 1.966302090138197,
      "epoch": 0.16713284686004026,
      "grad_norm": 0.3887089788913727,
      "learning_rate": 4.735157591606777e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9384690564125776,
      "num_tokens": 82759632.0,
      "step": 11170
    },
    {
      "entropy": 1.9579239383339881,
      "epoch": 0.16728247340154434,
      "grad_norm": 0.2957284152507782,
      "learning_rate": 4.734619420185468e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9465036384761334,
      "num_tokens": 82835445.0,
      "step": 11180
    },
    {
      "entropy": 1.997496996819973,
      "epoch": 0.1674320999430484,
      "grad_norm": 0.3416714668273926,
      "learning_rate": 4.7340807332364905e-05,
      "loss": 0.2378,
      "mean_token_accuracy": 0.9424118213355541,
      "num_tokens": 82908756.0,
      "step": 11190
    },
    {
      "entropy": 2.019940835982561,
      "epoch": 0.16758172648455247,
      "grad_norm": 0.32623180747032166,
      "learning_rate": 4.733541530884398e-05,
      "loss": 0.2078,
      "mean_token_accuracy": 0.9473976008594036,
      "num_tokens": 82984683.0,
      "step": 11200
    },
    {
      "entropy": 2.042385031282902,
      "epoch": 0.16773135302605652,
      "grad_norm": 0.37172645330429077,
      "learning_rate": 4.7330018132538645e-05,
      "loss": 0.2422,
      "mean_token_accuracy": 0.9400179572403431,
      "num_tokens": 83057513.0,
      "step": 11210
    },
    {
      "entropy": 2.015123999118805,
      "epoch": 0.1678809795675606,
      "grad_norm": 0.3376302421092987,
      "learning_rate": 4.7324615804696816e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9445292435586452,
      "num_tokens": 83132521.0,
      "step": 11220
    },
    {
      "entropy": 2.0334195122122765,
      "epoch": 0.16803060610906465,
      "grad_norm": 0.3890132009983063,
      "learning_rate": 4.731920832656761e-05,
      "loss": 0.2399,
      "mean_token_accuracy": 0.940800054743886,
      "num_tokens": 83206297.0,
      "step": 11230
    },
    {
      "entropy": 1.9911458320915698,
      "epoch": 0.16818023265056872,
      "grad_norm": 0.27696743607521057,
      "learning_rate": 4.7313795699401326e-05,
      "loss": 0.212,
      "mean_token_accuracy": 0.9480526529252529,
      "num_tokens": 83279423.0,
      "step": 11240
    },
    {
      "entropy": 2.006420696526766,
      "epoch": 0.16832985919207277,
      "grad_norm": 0.26376888155937195,
      "learning_rate": 4.730837792444947e-05,
      "loss": 0.2424,
      "mean_token_accuracy": 0.9410385452210903,
      "num_tokens": 83357460.0,
      "step": 11250
    },
    {
      "entropy": 2.003609035909176,
      "epoch": 0.16847948573357685,
      "grad_norm": 0.3280031085014343,
      "learning_rate": 4.7302955002964725e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9427330803126097,
      "num_tokens": 83433069.0,
      "step": 11260
    },
    {
      "entropy": 2.0081531308591365,
      "epoch": 0.16862911227508093,
      "grad_norm": 0.30559009313583374,
      "learning_rate": 4.7297526936200956e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.943436112254858,
      "num_tokens": 83506417.0,
      "step": 11270
    },
    {
      "entropy": 1.9873912274837493,
      "epoch": 0.16877873881658498,
      "grad_norm": 0.3467153012752533,
      "learning_rate": 4.729209372541324e-05,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9395671043545007,
      "num_tokens": 83580790.0,
      "step": 11280
    },
    {
      "entropy": 1.9938005469739437,
      "epoch": 0.16892836535808906,
      "grad_norm": 0.33694988489151,
      "learning_rate": 4.728665537185783e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9443761296570301,
      "num_tokens": 83656969.0,
      "step": 11290
    },
    {
      "entropy": 1.9792845122516156,
      "epoch": 0.1690779918995931,
      "grad_norm": 0.37290507555007935,
      "learning_rate": 4.7281211876792177e-05,
      "loss": 0.2315,
      "mean_token_accuracy": 0.9427817400544882,
      "num_tokens": 83735036.0,
      "step": 11300
    },
    {
      "entropy": 1.9764795929193497,
      "epoch": 0.16922761844109718,
      "grad_norm": 0.31503549218177795,
      "learning_rate": 4.7275763241474895e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9433816157281398,
      "num_tokens": 83810555.0,
      "step": 11310
    },
    {
      "entropy": 2.0068880416452886,
      "epoch": 0.16937724498260123,
      "grad_norm": 0.3658764064311981,
      "learning_rate": 4.727030946716582e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.942457003891468,
      "num_tokens": 83886670.0,
      "step": 11320
    },
    {
      "entropy": 1.9784133471548557,
      "epoch": 0.1695268715241053,
      "grad_norm": 0.30738013982772827,
      "learning_rate": 4.726485055512596e-05,
      "loss": 0.2061,
      "mean_token_accuracy": 0.9475188191980124,
      "num_tokens": 83962578.0,
      "step": 11330
    },
    {
      "entropy": 1.9734949164092541,
      "epoch": 0.16967649806560936,
      "grad_norm": 0.322635680437088,
      "learning_rate": 4.725938650661752e-05,
      "loss": 0.252,
      "mean_token_accuracy": 0.9433252871036529,
      "num_tokens": 84038224.0,
      "step": 11340
    },
    {
      "entropy": 1.9685859464108943,
      "epoch": 0.16982612460711344,
      "grad_norm": 0.3552205562591553,
      "learning_rate": 4.7253917322903865e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9435147665441036,
      "num_tokens": 84114270.0,
      "step": 11350
    },
    {
      "entropy": 1.9976129665970803,
      "epoch": 0.1699757511486175,
      "grad_norm": 0.3485437035560608,
      "learning_rate": 4.724844300524958e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9430071525275707,
      "num_tokens": 84187117.0,
      "step": 11360
    },
    {
      "entropy": 2.025367783755064,
      "epoch": 0.17012537769012157,
      "grad_norm": 0.3402247726917267,
      "learning_rate": 4.724296355492043e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9379003636538983,
      "num_tokens": 84259224.0,
      "step": 11370
    },
    {
      "entropy": 2.0367686569690706,
      "epoch": 0.17027500423162562,
      "grad_norm": 0.2624521553516388,
      "learning_rate": 4.723747897318334e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9414156951010227,
      "num_tokens": 84333292.0,
      "step": 11380
    },
    {
      "entropy": 1.998896025121212,
      "epoch": 0.1704246307731297,
      "grad_norm": 0.37605130672454834,
      "learning_rate": 4.7231989261306474e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9418044328689575,
      "num_tokens": 84406272.0,
      "step": 11390
    },
    {
      "entropy": 1.990217313170433,
      "epoch": 0.17057425731463374,
      "grad_norm": 0.25922632217407227,
      "learning_rate": 4.722649442055913e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9490035511553288,
      "num_tokens": 84482908.0,
      "step": 11400
    },
    {
      "entropy": 2.0117429167032244,
      "epoch": 0.17072388385613782,
      "grad_norm": 0.3807252049446106,
      "learning_rate": 4.722099445221181e-05,
      "loss": 0.2504,
      "mean_token_accuracy": 0.9396326608955861,
      "num_tokens": 84552310.0,
      "step": 11410
    },
    {
      "entropy": 1.9989011235535146,
      "epoch": 0.1708735103976419,
      "grad_norm": 0.33218055963516235,
      "learning_rate": 4.7215489357536224e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9399367064237595,
      "num_tokens": 84626001.0,
      "step": 11420
    },
    {
      "entropy": 1.9518065460026264,
      "epoch": 0.17102313693914595,
      "grad_norm": 0.2501559555530548,
      "learning_rate": 4.720997913780523e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9479783162474632,
      "num_tokens": 84701016.0,
      "step": 11430
    },
    {
      "entropy": 1.9699207291007041,
      "epoch": 0.17117276348065003,
      "grad_norm": 0.43689748644828796,
      "learning_rate": 4.72044637942929e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9467272449284792,
      "num_tokens": 84775505.0,
      "step": 11440
    },
    {
      "entropy": 1.977792189270258,
      "epoch": 0.17132239002215408,
      "grad_norm": 0.3523460030555725,
      "learning_rate": 4.719894332827447e-05,
      "loss": 0.241,
      "mean_token_accuracy": 0.9422855533659458,
      "num_tokens": 84847919.0,
      "step": 11450
    },
    {
      "entropy": 1.9940011978149415,
      "epoch": 0.17147201656365815,
      "grad_norm": 0.2924376428127289,
      "learning_rate": 4.719341774102636e-05,
      "loss": 0.2498,
      "mean_token_accuracy": 0.9395606476813555,
      "num_tokens": 84922478.0,
      "step": 11460
    },
    {
      "entropy": 2.009524831175804,
      "epoch": 0.1716216431051622,
      "grad_norm": 0.31076130270957947,
      "learning_rate": 4.7187887033826214e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9428077362477779,
      "num_tokens": 84993883.0,
      "step": 11470
    },
    {
      "entropy": 1.9925081998109817,
      "epoch": 0.17177126964666628,
      "grad_norm": 0.31091538071632385,
      "learning_rate": 4.718235120795282e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9450667701661587,
      "num_tokens": 85066121.0,
      "step": 11480
    },
    {
      "entropy": 1.9914929665625096,
      "epoch": 0.17192089618817033,
      "grad_norm": 0.45386621356010437,
      "learning_rate": 4.717681026468615e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.939835711568594,
      "num_tokens": 85138439.0,
      "step": 11490
    },
    {
      "entropy": 1.989299291372299,
      "epoch": 0.1720705227296744,
      "grad_norm": 0.32469138503074646,
      "learning_rate": 4.717126420530736e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9470309380441904,
      "num_tokens": 85209552.0,
      "step": 11500
    },
    {
      "entropy": 2.009778030216694,
      "epoch": 0.17222014927117846,
      "grad_norm": 0.2888270616531372,
      "learning_rate": 4.716571303109881e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9456829790025949,
      "num_tokens": 85282340.0,
      "step": 11510
    },
    {
      "entropy": 1.995441696792841,
      "epoch": 0.17236977581268254,
      "grad_norm": 0.34296390414237976,
      "learning_rate": 4.716015674334405e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9426594771444797,
      "num_tokens": 85354536.0,
      "step": 11520
    },
    {
      "entropy": 1.987555281072855,
      "epoch": 0.17251940235418659,
      "grad_norm": 0.26210397481918335,
      "learning_rate": 4.715459534332775e-05,
      "loss": 0.2555,
      "mean_token_accuracy": 0.9388522062450647,
      "num_tokens": 85427841.0,
      "step": 11530
    },
    {
      "entropy": 1.9448085844516754,
      "epoch": 0.17266902889569066,
      "grad_norm": 0.27720192074775696,
      "learning_rate": 4.714902883233584e-05,
      "loss": 0.1969,
      "mean_token_accuracy": 0.9489927809685469,
      "num_tokens": 85502180.0,
      "step": 11540
    },
    {
      "entropy": 1.9452966034412384,
      "epoch": 0.17281865543719474,
      "grad_norm": 0.3437153100967407,
      "learning_rate": 4.7143457211655376e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9438325099647045,
      "num_tokens": 85575647.0,
      "step": 11550
    },
    {
      "entropy": 1.9413255617022513,
      "epoch": 0.1729682819786988,
      "grad_norm": 0.3672594726085663,
      "learning_rate": 4.713788048257463e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9421920731663704,
      "num_tokens": 85647727.0,
      "step": 11560
    },
    {
      "entropy": 1.937819854915142,
      "epoch": 0.17311790852020287,
      "grad_norm": 0.33253198862075806,
      "learning_rate": 4.713229864638303e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9448194082826376,
      "num_tokens": 85723911.0,
      "step": 11570
    },
    {
      "entropy": 1.967685354501009,
      "epoch": 0.17326753506170692,
      "grad_norm": 0.3533225357532501,
      "learning_rate": 4.712671170437121e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9368028469383717,
      "num_tokens": 85796999.0,
      "step": 11580
    },
    {
      "entropy": 1.9381090596318244,
      "epoch": 0.173417161603211,
      "grad_norm": 0.3203708231449127,
      "learning_rate": 4.712111965783095e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9443447958678007,
      "num_tokens": 85871746.0,
      "step": 11590
    },
    {
      "entropy": 1.93919398188591,
      "epoch": 0.17356678814471505,
      "grad_norm": 0.32674136757850647,
      "learning_rate": 4.711552250805525e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.944812948256731,
      "num_tokens": 85943385.0,
      "step": 11600
    },
    {
      "entropy": 1.9240055955946445,
      "epoch": 0.17371641468621912,
      "grad_norm": 0.3292785882949829,
      "learning_rate": 4.710992025633825e-05,
      "loss": 0.2122,
      "mean_token_accuracy": 0.947202593833208,
      "num_tokens": 86019547.0,
      "step": 11610
    },
    {
      "entropy": 1.9504281006753446,
      "epoch": 0.17386604122772317,
      "grad_norm": 0.3692404329776764,
      "learning_rate": 4.710431290397531e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9428993128240108,
      "num_tokens": 86095600.0,
      "step": 11620
    },
    {
      "entropy": 2.0045026652514935,
      "epoch": 0.17401566776922725,
      "grad_norm": 0.31418874859809875,
      "learning_rate": 4.709870045226293e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9433543309569359,
      "num_tokens": 86168060.0,
      "step": 11630
    },
    {
      "entropy": 1.9617283843457698,
      "epoch": 0.1741652943107313,
      "grad_norm": 0.3664930462837219,
      "learning_rate": 4.709308290249883e-05,
      "loss": 0.2098,
      "mean_token_accuracy": 0.9459586538374424,
      "num_tokens": 86239125.0,
      "step": 11640
    },
    {
      "entropy": 1.929988507926464,
      "epoch": 0.17431492085223538,
      "grad_norm": 0.3214315176010132,
      "learning_rate": 4.708746025598188e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9470377992838621,
      "num_tokens": 86311942.0,
      "step": 11650
    },
    {
      "entropy": 1.9321693368256092,
      "epoch": 0.17446454739373943,
      "grad_norm": 0.4081851840019226,
      "learning_rate": 4.708183251401213e-05,
      "loss": 0.2326,
      "mean_token_accuracy": 0.9418060846626759,
      "num_tokens": 86385061.0,
      "step": 11660
    },
    {
      "entropy": 1.9304345116019248,
      "epoch": 0.1746141739352435,
      "grad_norm": 0.47714775800704956,
      "learning_rate": 4.707619967789082e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9451507557183504,
      "num_tokens": 86459482.0,
      "step": 11670
    },
    {
      "entropy": 1.9033852018415929,
      "epoch": 0.17476380047674756,
      "grad_norm": 0.3254261612892151,
      "learning_rate": 4.707056174892035e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.947445360198617,
      "num_tokens": 86537158.0,
      "step": 11680
    },
    {
      "entropy": 1.9561389043927193,
      "epoch": 0.17491342701825163,
      "grad_norm": 0.3169887065887451,
      "learning_rate": 4.706491872840433e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9413825996220112,
      "num_tokens": 86609901.0,
      "step": 11690
    },
    {
      "entropy": 1.9344462431967258,
      "epoch": 0.1750630535597557,
      "grad_norm": 0.31042397022247314,
      "learning_rate": 4.705927061764751e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9433655422180891,
      "num_tokens": 86683481.0,
      "step": 11700
    },
    {
      "entropy": 1.9256453976035117,
      "epoch": 0.17521268010125976,
      "grad_norm": 0.41656678915023804,
      "learning_rate": 4.7053617417955844e-05,
      "loss": 0.2308,
      "mean_token_accuracy": 0.9437326103448868,
      "num_tokens": 86755408.0,
      "step": 11710
    },
    {
      "entropy": 1.9313046686351298,
      "epoch": 0.17536230664276384,
      "grad_norm": 0.32611384987831116,
      "learning_rate": 4.7047959130636437e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9393491044640541,
      "num_tokens": 86828799.0,
      "step": 11720
    },
    {
      "entropy": 1.919238407164812,
      "epoch": 0.1755119331842679,
      "grad_norm": 0.28555989265441895,
      "learning_rate": 4.704229575699761e-05,
      "loss": 0.2038,
      "mean_token_accuracy": 0.9483412735164165,
      "num_tokens": 86907131.0,
      "step": 11730
    },
    {
      "entropy": 1.959150043874979,
      "epoch": 0.17566155972577197,
      "grad_norm": 0.3012308180332184,
      "learning_rate": 4.703662729834881e-05,
      "loss": 0.2425,
      "mean_token_accuracy": 0.9386565640568734,
      "num_tokens": 86979766.0,
      "step": 11740
    },
    {
      "entropy": 1.9357104040682316,
      "epoch": 0.17581118626727602,
      "grad_norm": 0.35708725452423096,
      "learning_rate": 4.7030953756000706e-05,
      "loss": 0.2399,
      "mean_token_accuracy": 0.9450996421277523,
      "num_tokens": 87055950.0,
      "step": 11750
    },
    {
      "entropy": 1.9493321485817432,
      "epoch": 0.1759608128087801,
      "grad_norm": 0.38161128759384155,
      "learning_rate": 4.7025275131265106e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9417205080389977,
      "num_tokens": 87131067.0,
      "step": 11760
    },
    {
      "entropy": 1.9052909553050994,
      "epoch": 0.17611043935028414,
      "grad_norm": 0.3973505198955536,
      "learning_rate": 4.701959142545502e-05,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9479531172662974,
      "num_tokens": 87206555.0,
      "step": 11770
    },
    {
      "entropy": 1.9235304296016693,
      "epoch": 0.17626006589178822,
      "grad_norm": 0.5511879324913025,
      "learning_rate": 4.701390263988462e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9396278414875269,
      "num_tokens": 87280966.0,
      "step": 11780
    },
    {
      "entropy": 1.9063782408833503,
      "epoch": 0.17640969243329227,
      "grad_norm": 0.34514760971069336,
      "learning_rate": 4.7008208775869254e-05,
      "loss": 0.2201,
      "mean_token_accuracy": 0.9461074750870466,
      "num_tokens": 87355770.0,
      "step": 11790
    },
    {
      "entropy": 1.908871029317379,
      "epoch": 0.17655931897479635,
      "grad_norm": 0.39832258224487305,
      "learning_rate": 4.700250983472544e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9455657251179218,
      "num_tokens": 87429928.0,
      "step": 11800
    },
    {
      "entropy": 1.93466479703784,
      "epoch": 0.1767089455163004,
      "grad_norm": 0.4126843810081482,
      "learning_rate": 4.699680581777088e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9448823489248752,
      "num_tokens": 87504561.0,
      "step": 11810
    },
    {
      "entropy": 1.9009818837046624,
      "epoch": 0.17685857205780448,
      "grad_norm": 0.3930802345275879,
      "learning_rate": 4.699109672632444e-05,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9463172253221274,
      "num_tokens": 87579014.0,
      "step": 11820
    },
    {
      "entropy": 1.9089251399040221,
      "epoch": 0.17700819859930852,
      "grad_norm": 0.30471739172935486,
      "learning_rate": 4.698538256170616e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9431010823696852,
      "num_tokens": 87653519.0,
      "step": 11830
    },
    {
      "entropy": 1.9564978040754795,
      "epoch": 0.1771578251408126,
      "grad_norm": 0.3308776319026947,
      "learning_rate": 4.697966332523728e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9403819229453803,
      "num_tokens": 87726421.0,
      "step": 11840
    },
    {
      "entropy": 1.967709169536829,
      "epoch": 0.17730745168231668,
      "grad_norm": 0.35049569606781006,
      "learning_rate": 4.697393901824016e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9409485526382924,
      "num_tokens": 87801350.0,
      "step": 11850
    },
    {
      "entropy": 1.9941087171435357,
      "epoch": 0.17745707822382073,
      "grad_norm": 0.40882498025894165,
      "learning_rate": 4.696820964203837e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9398682795464992,
      "num_tokens": 87873180.0,
      "step": 11860
    },
    {
      "entropy": 1.967032478749752,
      "epoch": 0.1776067047653248,
      "grad_norm": 0.3035028874874115,
      "learning_rate": 4.696247519795665e-05,
      "loss": 0.2471,
      "mean_token_accuracy": 0.9369981177151203,
      "num_tokens": 87946210.0,
      "step": 11870
    },
    {
      "entropy": 1.9317103445529937,
      "epoch": 0.17775633130682886,
      "grad_norm": 0.38312920928001404,
      "learning_rate": 4.69567356873209e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.94202882014215,
      "num_tokens": 88023240.0,
      "step": 11880
    },
    {
      "entropy": 1.939635204523802,
      "epoch": 0.17790595784833294,
      "grad_norm": 0.41997671127319336,
      "learning_rate": 4.6950991111458206e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.9437234051525593,
      "num_tokens": 88096232.0,
      "step": 11890
    },
    {
      "entropy": 1.9522278435528277,
      "epoch": 0.17805558438983698,
      "grad_norm": 0.3657970130443573,
      "learning_rate": 4.6945241471696796e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9440057579427957,
      "num_tokens": 88170037.0,
      "step": 11900
    },
    {
      "entropy": 1.9423182480037213,
      "epoch": 0.17820521093134106,
      "grad_norm": 0.3174639642238617,
      "learning_rate": 4.6939486769366106e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9480380184948445,
      "num_tokens": 88244594.0,
      "step": 11910
    },
    {
      "entropy": 1.9633292973041534,
      "epoch": 0.1783548374728451,
      "grad_norm": 0.3805256187915802,
      "learning_rate": 4.6933727005796736e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9393668442964553,
      "num_tokens": 88319947.0,
      "step": 11920
    },
    {
      "entropy": 1.9683417588472367,
      "epoch": 0.1785044640143492,
      "grad_norm": 0.4310803711414337,
      "learning_rate": 4.692796218232042e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.942944435030222,
      "num_tokens": 88394082.0,
      "step": 11930
    },
    {
      "entropy": 1.9801807649433614,
      "epoch": 0.17865409055585324,
      "grad_norm": 0.3662787079811096,
      "learning_rate": 4.6922192300270104e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9408070523291826,
      "num_tokens": 88472198.0,
      "step": 11940
    },
    {
      "entropy": 1.950430803745985,
      "epoch": 0.17880371709735732,
      "grad_norm": 0.28596171736717224,
      "learning_rate": 4.691641736097987e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9443738542497158,
      "num_tokens": 88545012.0,
      "step": 11950
    },
    {
      "entropy": 1.944110681116581,
      "epoch": 0.17895334363886137,
      "grad_norm": 0.3334936499595642,
      "learning_rate": 4.6910637365785004e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9389479156583548,
      "num_tokens": 88617233.0,
      "step": 11960
    },
    {
      "entropy": 1.918237602710724,
      "epoch": 0.17910297018036545,
      "grad_norm": 0.34052762389183044,
      "learning_rate": 4.690485231602195e-05,
      "loss": 0.2023,
      "mean_token_accuracy": 0.9470183115452528,
      "num_tokens": 88694388.0,
      "step": 11970
    },
    {
      "entropy": 1.9329943858087062,
      "epoch": 0.17925259672186952,
      "grad_norm": 0.34407472610473633,
      "learning_rate": 4.689906221302829e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9479840029031038,
      "num_tokens": 88770862.0,
      "step": 11980
    },
    {
      "entropy": 1.9657779030501843,
      "epoch": 0.17940222326337357,
      "grad_norm": 0.3621869683265686,
      "learning_rate": 4.689326705814282e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9447189796715975,
      "num_tokens": 88842562.0,
      "step": 11990
    },
    {
      "entropy": 1.988500027358532,
      "epoch": 0.17955184980487765,
      "grad_norm": 0.5374886393547058,
      "learning_rate": 4.6887466852705474e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9480006195604801,
      "num_tokens": 88915917.0,
      "step": 12000
    },
    {
      "epoch": 0.17955184980487765,
      "eval_entropy": 1.9875859952926636,
      "eval_loss": 0.2513410151004791,
      "eval_mean_token_accuracy": 0.9441012032866478,
      "eval_num_tokens": 88915917.0,
      "eval_runtime": 545.636,
      "eval_samples_per_second": 36.654,
      "eval_steps_per_second": 9.164,
      "step": 12000
    },
    {
      "entropy": 2.012895941734314,
      "epoch": 0.1797014763463817,
      "grad_norm": 0.3534221947193146,
      "learning_rate": 4.688166159805736e-05,
      "loss": 0.2492,
      "mean_token_accuracy": 0.9361440792679787,
      "num_tokens": 88989913.0,
      "step": 12010
    },
    {
      "entropy": 1.9872764982283115,
      "epoch": 0.17985110288788578,
      "grad_norm": 0.35632598400115967,
      "learning_rate": 4.687585129554076e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9454539440572262,
      "num_tokens": 89062190.0,
      "step": 12020
    },
    {
      "entropy": 1.9924490667879582,
      "epoch": 0.18000072942938983,
      "grad_norm": 0.423711895942688,
      "learning_rate": 4.6870035946499124e-05,
      "loss": 0.2437,
      "mean_token_accuracy": 0.9400301907211542,
      "num_tokens": 89136097.0,
      "step": 12030
    },
    {
      "entropy": 1.982486742734909,
      "epoch": 0.1801503559708939,
      "grad_norm": 0.34255117177963257,
      "learning_rate": 4.686421555227706e-05,
      "loss": 0.2363,
      "mean_token_accuracy": 0.9440226208418607,
      "num_tokens": 89207254.0,
      "step": 12040
    },
    {
      "entropy": 1.9663990870118142,
      "epoch": 0.18029998251239795,
      "grad_norm": 0.33750784397125244,
      "learning_rate": 4.685839011422034e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9415697660297155,
      "num_tokens": 89281090.0,
      "step": 12050
    },
    {
      "entropy": 1.9928039580583572,
      "epoch": 0.18044960905390203,
      "grad_norm": 0.3189346492290497,
      "learning_rate": 4.685255963367591e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.9457709453999996,
      "num_tokens": 89354069.0,
      "step": 12060
    },
    {
      "entropy": 2.0083920061588287,
      "epoch": 0.18059923559540608,
      "grad_norm": 0.3330828845500946,
      "learning_rate": 4.6846724111991897e-05,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9401556111872196,
      "num_tokens": 89427942.0,
      "step": 12070
    },
    {
      "entropy": 2.0150293998420237,
      "epoch": 0.18074886213691016,
      "grad_norm": 0.307551771402359,
      "learning_rate": 4.684088355051756e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9449098668992519,
      "num_tokens": 89501309.0,
      "step": 12080
    },
    {
      "entropy": 1.9826707653701305,
      "epoch": 0.1808984886784142,
      "grad_norm": 0.3512353301048279,
      "learning_rate": 4.683503795060335e-05,
      "loss": 0.2299,
      "mean_token_accuracy": 0.94310795776546,
      "num_tokens": 89576985.0,
      "step": 12090
    },
    {
      "entropy": 2.0480178661644457,
      "epoch": 0.1810481152199183,
      "grad_norm": 0.29993006587028503,
      "learning_rate": 4.682918731360088e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9390263468027115,
      "num_tokens": 89654107.0,
      "step": 12100
    },
    {
      "entropy": 2.0381418362259867,
      "epoch": 0.18119774176142234,
      "grad_norm": 0.3122093379497528,
      "learning_rate": 4.6823331640862886e-05,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9458162236958743,
      "num_tokens": 89726204.0,
      "step": 12110
    },
    {
      "entropy": 1.9979759395122527,
      "epoch": 0.18134736830292641,
      "grad_norm": 0.33283668756484985,
      "learning_rate": 4.681747093374335e-05,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9441422950476408,
      "num_tokens": 89800794.0,
      "step": 12120
    },
    {
      "entropy": 1.9659617491066457,
      "epoch": 0.1814969948444305,
      "grad_norm": 0.38408419489860535,
      "learning_rate": 4.681160519359735e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9414928331971169,
      "num_tokens": 89873039.0,
      "step": 12130
    },
    {
      "entropy": 1.9444603882730007,
      "epoch": 0.18164662138593454,
      "grad_norm": 0.3329567611217499,
      "learning_rate": 4.680573442178114e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.94475935138762,
      "num_tokens": 89945730.0,
      "step": 12140
    },
    {
      "entropy": 1.9573444165289402,
      "epoch": 0.18179624792743862,
      "grad_norm": 0.3912147581577301,
      "learning_rate": 4.679985861965217e-05,
      "loss": 0.2321,
      "mean_token_accuracy": 0.9435878794640302,
      "num_tokens": 90019389.0,
      "step": 12150
    },
    {
      "entropy": 1.9586571998894216,
      "epoch": 0.18194587446894267,
      "grad_norm": 0.38354218006134033,
      "learning_rate": 4.6793977788569e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.946517676860094,
      "num_tokens": 90094285.0,
      "step": 12160
    },
    {
      "entropy": 1.97623944953084,
      "epoch": 0.18209550101044675,
      "grad_norm": 0.37047767639160156,
      "learning_rate": 4.678809192989141e-05,
      "loss": 0.2536,
      "mean_token_accuracy": 0.9409733712673187,
      "num_tokens": 90169038.0,
      "step": 12170
    },
    {
      "entropy": 2.041138318926096,
      "epoch": 0.1822451275519508,
      "grad_norm": 0.3336089849472046,
      "learning_rate": 4.678220104498031e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.942755876109004,
      "num_tokens": 90250646.0,
      "step": 12180
    },
    {
      "entropy": 2.015570829808712,
      "epoch": 0.18239475409345487,
      "grad_norm": 0.37503889203071594,
      "learning_rate": 4.677630513519777e-05,
      "loss": 0.1945,
      "mean_token_accuracy": 0.9520297717303038,
      "num_tokens": 90324523.0,
      "step": 12190
    },
    {
      "entropy": 1.973142673075199,
      "epoch": 0.18254438063495892,
      "grad_norm": 0.41232702136039734,
      "learning_rate": 4.677040420190703e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9427297707647085,
      "num_tokens": 90397984.0,
      "step": 12200
    },
    {
      "entropy": 1.9758772172033787,
      "epoch": 0.182694007176463,
      "grad_norm": 0.3938673734664917,
      "learning_rate": 4.676449824647249e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9453857235610486,
      "num_tokens": 90472863.0,
      "step": 12210
    },
    {
      "entropy": 2.016880201548338,
      "epoch": 0.18284363371796705,
      "grad_norm": 0.4802177846431732,
      "learning_rate": 4.675858727025973e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.9390620745718479,
      "num_tokens": 90545328.0,
      "step": 12220
    },
    {
      "entropy": 2.0198921017348765,
      "epoch": 0.18299326025947113,
      "grad_norm": 0.33742478489875793,
      "learning_rate": 4.675267127463545e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9443189572542906,
      "num_tokens": 90621653.0,
      "step": 12230
    },
    {
      "entropy": 2.04875762462616,
      "epoch": 0.18314288680097518,
      "grad_norm": 0.3153846263885498,
      "learning_rate": 4.6746750260967525e-05,
      "loss": 0.241,
      "mean_token_accuracy": 0.939516383036971,
      "num_tokens": 90696348.0,
      "step": 12240
    },
    {
      "entropy": 2.0584877356886864,
      "epoch": 0.18329251334247926,
      "grad_norm": 0.2935430705547333,
      "learning_rate": 4.6740824230625025e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9445522993803024,
      "num_tokens": 90772986.0,
      "step": 12250
    },
    {
      "entropy": 1.9970536909997463,
      "epoch": 0.1834421398839833,
      "grad_norm": 0.3429434895515442,
      "learning_rate": 4.673489318497815e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9447935916483402,
      "num_tokens": 90847550.0,
      "step": 12260
    },
    {
      "entropy": 1.9761030197143554,
      "epoch": 0.18359176642548738,
      "grad_norm": 0.3029574751853943,
      "learning_rate": 4.6728957125398256e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9419239275157452,
      "num_tokens": 90921388.0,
      "step": 12270
    },
    {
      "entropy": 2.002379209548235,
      "epoch": 0.18374139296699146,
      "grad_norm": 0.31647494435310364,
      "learning_rate": 4.672301605325786e-05,
      "loss": 0.2626,
      "mean_token_accuracy": 0.9385421011596918,
      "num_tokens": 90994944.0,
      "step": 12280
    },
    {
      "entropy": 2.0049819447100163,
      "epoch": 0.1838910195084955,
      "grad_norm": 0.3294638395309448,
      "learning_rate": 4.671706996993064e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9447358623147011,
      "num_tokens": 91072590.0,
      "step": 12290
    },
    {
      "entropy": 1.9556850589811803,
      "epoch": 0.1840406460499996,
      "grad_norm": 0.3292301297187805,
      "learning_rate": 4.671111887679146e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9431035924702883,
      "num_tokens": 91146085.0,
      "step": 12300
    },
    {
      "entropy": 1.9493032641708852,
      "epoch": 0.18419027259150364,
      "grad_norm": 0.4046098291873932,
      "learning_rate": 4.670516277521631e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9407251317054033,
      "num_tokens": 91219831.0,
      "step": 12310
    },
    {
      "entropy": 1.9774758778512478,
      "epoch": 0.18433989913300772,
      "grad_norm": 0.2884744703769684,
      "learning_rate": 4.669920166658233e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.941375371068716,
      "num_tokens": 91294385.0,
      "step": 12320
    },
    {
      "entropy": 1.9768357135355472,
      "epoch": 0.18448952567451177,
      "grad_norm": 0.35058748722076416,
      "learning_rate": 4.669323555226784e-05,
      "loss": 0.2005,
      "mean_token_accuracy": 0.9459309797734022,
      "num_tokens": 91366030.0,
      "step": 12330
    },
    {
      "entropy": 1.9743297010660172,
      "epoch": 0.18463915221601584,
      "grad_norm": 0.3337935507297516,
      "learning_rate": 4.6687264433652325e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.943428598344326,
      "num_tokens": 91438181.0,
      "step": 12340
    },
    {
      "entropy": 1.9773857183754444,
      "epoch": 0.1847887787575199,
      "grad_norm": 0.384075790643692,
      "learning_rate": 4.668128831211641e-05,
      "loss": 0.2389,
      "mean_token_accuracy": 0.9412962321192027,
      "num_tokens": 91514040.0,
      "step": 12350
    },
    {
      "entropy": 1.986565286666155,
      "epoch": 0.18493840529902397,
      "grad_norm": 0.25894656777381897,
      "learning_rate": 4.667530718904187e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9439631383866072,
      "num_tokens": 91590875.0,
      "step": 12360
    },
    {
      "entropy": 1.9711171582341194,
      "epoch": 0.18508803184052802,
      "grad_norm": 0.31875500082969666,
      "learning_rate": 4.666932106581166e-05,
      "loss": 0.235,
      "mean_token_accuracy": 0.9451153051108122,
      "num_tokens": 91666768.0,
      "step": 12370
    },
    {
      "entropy": 1.9550660729408265,
      "epoch": 0.1852376583820321,
      "grad_norm": 0.2813469469547272,
      "learning_rate": 4.666332994380987e-05,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9487713694572448,
      "num_tokens": 91743694.0,
      "step": 12380
    },
    {
      "entropy": 1.9945204332470894,
      "epoch": 0.18538728492353615,
      "grad_norm": 0.3674592673778534,
      "learning_rate": 4.665733382442175e-05,
      "loss": 0.2448,
      "mean_token_accuracy": 0.9395047053694725,
      "num_tokens": 91817221.0,
      "step": 12390
    },
    {
      "entropy": 2.0199562132358553,
      "epoch": 0.18553691146504023,
      "grad_norm": 0.38970842957496643,
      "learning_rate": 4.6651332709033715e-05,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9402485869824886,
      "num_tokens": 91891831.0,
      "step": 12400
    },
    {
      "entropy": 2.012060023844242,
      "epoch": 0.18568653800654428,
      "grad_norm": 0.3685074746608734,
      "learning_rate": 4.664532659903333e-05,
      "loss": 0.2489,
      "mean_token_accuracy": 0.9402780160307884,
      "num_tokens": 91964973.0,
      "step": 12410
    },
    {
      "entropy": 1.9883493572473525,
      "epoch": 0.18583616454804835,
      "grad_norm": 0.32115596532821655,
      "learning_rate": 4.663931549580932e-05,
      "loss": 0.2035,
      "mean_token_accuracy": 0.9494829922914505,
      "num_tokens": 92041187.0,
      "step": 12420
    },
    {
      "entropy": 2.0159843400120736,
      "epoch": 0.18598579108955243,
      "grad_norm": 0.4011954367160797,
      "learning_rate": 4.663329940075155e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.9392575826495886,
      "num_tokens": 92113701.0,
      "step": 12430
    },
    {
      "entropy": 2.0084626637399197,
      "epoch": 0.18613541763105648,
      "grad_norm": 0.2720877528190613,
      "learning_rate": 4.6627278315251046e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9442315146327018,
      "num_tokens": 92187595.0,
      "step": 12440
    },
    {
      "entropy": 2.0480816535651685,
      "epoch": 0.18628504417256056,
      "grad_norm": 0.35714223980903625,
      "learning_rate": 4.662125224070001e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9465075585991144,
      "num_tokens": 92260465.0,
      "step": 12450
    },
    {
      "entropy": 2.0112148486077785,
      "epoch": 0.1864346707140646,
      "grad_norm": 0.35225775837898254,
      "learning_rate": 4.6615221178491754e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9486066609621048,
      "num_tokens": 92336835.0,
      "step": 12460
    },
    {
      "entropy": 2.012115205079317,
      "epoch": 0.1865842972555687,
      "grad_norm": 0.40558138489723206,
      "learning_rate": 4.660918513002078e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9446737866848707,
      "num_tokens": 92409251.0,
      "step": 12470
    },
    {
      "entropy": 2.001447109133005,
      "epoch": 0.18673392379707274,
      "grad_norm": 0.39400357007980347,
      "learning_rate": 4.660314409668272e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9460970275104046,
      "num_tokens": 92483828.0,
      "step": 12480
    },
    {
      "entropy": 2.0200071588158606,
      "epoch": 0.18688355033857681,
      "grad_norm": 0.3507217466831207,
      "learning_rate": 4.6597098079874376e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9435345999896526,
      "num_tokens": 92556122.0,
      "step": 12490
    },
    {
      "entropy": 2.024462301284075,
      "epoch": 0.18703317688008086,
      "grad_norm": 0.359611451625824,
      "learning_rate": 4.65910470809937e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9451525513082742,
      "num_tokens": 92632267.0,
      "step": 12500
    },
    {
      "entropy": 2.0418856598436834,
      "epoch": 0.18718280342158494,
      "grad_norm": 0.416637122631073,
      "learning_rate": 4.658499110143978e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9408710427582264,
      "num_tokens": 92703385.0,
      "step": 12510
    },
    {
      "entropy": 2.0599921569228172,
      "epoch": 0.187332429963089,
      "grad_norm": 0.3596418499946594,
      "learning_rate": 4.657893014261286e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9438908442854881,
      "num_tokens": 92775553.0,
      "step": 12520
    },
    {
      "entropy": 2.033356899023056,
      "epoch": 0.18748205650459307,
      "grad_norm": 0.267794132232666,
      "learning_rate": 4.657286420591436e-05,
      "loss": 0.2006,
      "mean_token_accuracy": 0.9488066412508488,
      "num_tokens": 92851494.0,
      "step": 12530
    },
    {
      "entropy": 1.994048361480236,
      "epoch": 0.18763168304609712,
      "grad_norm": 0.28815969824790955,
      "learning_rate": 4.656679329274682e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9466773256659508,
      "num_tokens": 92929310.0,
      "step": 12540
    },
    {
      "entropy": 1.986451756209135,
      "epoch": 0.1877813095876012,
      "grad_norm": 0.3258649408817291,
      "learning_rate": 4.656071740451395e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9459570970386266,
      "num_tokens": 93005383.0,
      "step": 12550
    },
    {
      "entropy": 1.9851609982550145,
      "epoch": 0.18793093612910527,
      "grad_norm": 0.412160724401474,
      "learning_rate": 4.65546365426206e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.944441294670105,
      "num_tokens": 93079874.0,
      "step": 12560
    },
    {
      "entropy": 1.9879626117646694,
      "epoch": 0.18808056267060932,
      "grad_norm": 0.5187681913375854,
      "learning_rate": 4.654855070847277e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.948008393123746,
      "num_tokens": 93154115.0,
      "step": 12570
    },
    {
      "entropy": 1.993713741749525,
      "epoch": 0.1882301892121134,
      "grad_norm": 0.3736415207386017,
      "learning_rate": 4.654245990347762e-05,
      "loss": 0.2451,
      "mean_token_accuracy": 0.9392469093203545,
      "num_tokens": 93227654.0,
      "step": 12580
    },
    {
      "entropy": 1.9707159124314786,
      "epoch": 0.18837981575361745,
      "grad_norm": 0.3252852261066437,
      "learning_rate": 4.653636412904344e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9420193661004305,
      "num_tokens": 93301105.0,
      "step": 12590
    },
    {
      "entropy": 2.0169090069830418,
      "epoch": 0.18852944229512153,
      "grad_norm": 0.34150323271751404,
      "learning_rate": 4.65302633865797e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9447330333292484,
      "num_tokens": 93375384.0,
      "step": 12600
    },
    {
      "entropy": 1.9933415532112122,
      "epoch": 0.18867906883662558,
      "grad_norm": 0.3371177017688751,
      "learning_rate": 4.6524157677496985e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9466878648847341,
      "num_tokens": 93450330.0,
      "step": 12610
    },
    {
      "entropy": 1.9838598623871804,
      "epoch": 0.18882869537812966,
      "grad_norm": 0.3177426755428314,
      "learning_rate": 4.651804700320706e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9484531678259372,
      "num_tokens": 93523839.0,
      "step": 12620
    },
    {
      "entropy": 2.011392040550709,
      "epoch": 0.1889783219196337,
      "grad_norm": 0.343526691198349,
      "learning_rate": 4.6511931365122804e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9463164493441582,
      "num_tokens": 93598052.0,
      "step": 12630
    },
    {
      "entropy": 2.048951276391745,
      "epoch": 0.18912794846113778,
      "grad_norm": 0.36618557572364807,
      "learning_rate": 4.6505810764658267e-05,
      "loss": 0.2518,
      "mean_token_accuracy": 0.9396453212946654,
      "num_tokens": 93671792.0,
      "step": 12640
    },
    {
      "entropy": 2.046569278091192,
      "epoch": 0.18927757500264183,
      "grad_norm": 0.3630104064941406,
      "learning_rate": 4.649968520322864e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9469216849654913,
      "num_tokens": 93743419.0,
      "step": 12650
    },
    {
      "entropy": 1.998548550158739,
      "epoch": 0.1894272015441459,
      "grad_norm": 0.3913586139678955,
      "learning_rate": 4.649355468225026e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9418482434004545,
      "num_tokens": 93818566.0,
      "step": 12660
    },
    {
      "entropy": 2.0538141928613185,
      "epoch": 0.18957682808564996,
      "grad_norm": 0.37342703342437744,
      "learning_rate": 4.648741920314062e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9431826565414667,
      "num_tokens": 93893712.0,
      "step": 12670
    },
    {
      "entropy": 2.0663034223020076,
      "epoch": 0.18972645462715404,
      "grad_norm": 0.3052605390548706,
      "learning_rate": 4.6481278767318356e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9462787631899119,
      "num_tokens": 93968978.0,
      "step": 12680
    },
    {
      "entropy": 2.0582934468984604,
      "epoch": 0.1898760811686581,
      "grad_norm": 0.46919500827789307,
      "learning_rate": 4.6475133376203234e-05,
      "loss": 0.2354,
      "mean_token_accuracy": 0.9422364257276058,
      "num_tokens": 94042750.0,
      "step": 12690
    },
    {
      "entropy": 2.0404804304242132,
      "epoch": 0.19002570771016217,
      "grad_norm": 0.3673853576183319,
      "learning_rate": 4.646898303121618e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9459051191806793,
      "num_tokens": 94115821.0,
      "step": 12700
    },
    {
      "entropy": 2.0473834209144117,
      "epoch": 0.19017533425166624,
      "grad_norm": 0.34090012311935425,
      "learning_rate": 4.646282773377927e-05,
      "loss": 0.2107,
      "mean_token_accuracy": 0.9489753317087889,
      "num_tokens": 94189227.0,
      "step": 12710
    },
    {
      "entropy": 2.030070558935404,
      "epoch": 0.1903249607931703,
      "grad_norm": 0.38792163133621216,
      "learning_rate": 4.64566674853157e-05,
      "loss": 0.2438,
      "mean_token_accuracy": 0.9421743422746658,
      "num_tokens": 94264537.0,
      "step": 12720
    },
    {
      "entropy": 2.0587759517133235,
      "epoch": 0.19047458733467437,
      "grad_norm": 0.3498346507549286,
      "learning_rate": 4.6450502287249856e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9464053608477115,
      "num_tokens": 94337210.0,
      "step": 12730
    },
    {
      "entropy": 2.0429174192249775,
      "epoch": 0.19062421387617842,
      "grad_norm": 0.35026171803474426,
      "learning_rate": 4.644433214100722e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9421745900064706,
      "num_tokens": 94408998.0,
      "step": 12740
    },
    {
      "entropy": 2.0348124846816065,
      "epoch": 0.1907738404176825,
      "grad_norm": 0.34645387530326843,
      "learning_rate": 4.643815704801445e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9420465085655451,
      "num_tokens": 94482651.0,
      "step": 12750
    },
    {
      "entropy": 2.0275880500674246,
      "epoch": 0.19092346695918655,
      "grad_norm": 0.395126610994339,
      "learning_rate": 4.643197700969933e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9426661632955075,
      "num_tokens": 94560087.0,
      "step": 12760
    },
    {
      "entropy": 2.0455701306462286,
      "epoch": 0.19107309350069063,
      "grad_norm": 0.37439921498298645,
      "learning_rate": 4.6425792027490797e-05,
      "loss": 0.2592,
      "mean_token_accuracy": 0.9403974823653698,
      "num_tokens": 94632836.0,
      "step": 12770
    },
    {
      "entropy": 2.0700371719896795,
      "epoch": 0.19122272004219468,
      "grad_norm": 0.2971782982349396,
      "learning_rate": 4.6419602102818935e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9445210836827755,
      "num_tokens": 94707014.0,
      "step": 12780
    },
    {
      "entropy": 2.079044646024704,
      "epoch": 0.19137234658369875,
      "grad_norm": 0.30761635303497314,
      "learning_rate": 4.6413407237114965e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9456683035939932,
      "num_tokens": 94784026.0,
      "step": 12790
    },
    {
      "entropy": 2.0358204998075964,
      "epoch": 0.1915219731252028,
      "grad_norm": 0.3715766668319702,
      "learning_rate": 4.640720743181125e-05,
      "loss": 0.2025,
      "mean_token_accuracy": 0.9493161968886852,
      "num_tokens": 94859856.0,
      "step": 12800
    },
    {
      "entropy": 2.0614868752658366,
      "epoch": 0.19167159966670688,
      "grad_norm": 0.3233117163181305,
      "learning_rate": 4.6401002688341274e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9425906989723444,
      "num_tokens": 94930769.0,
      "step": 12810
    },
    {
      "entropy": 2.071055691689253,
      "epoch": 0.19182122620821093,
      "grad_norm": 0.297460675239563,
      "learning_rate": 4.6394793008139715e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9464649677276611,
      "num_tokens": 95008849.0,
      "step": 12820
    },
    {
      "entropy": 2.0118863292038442,
      "epoch": 0.191970852749715,
      "grad_norm": 0.35426774621009827,
      "learning_rate": 4.6388578392642344e-05,
      "loss": 0.2103,
      "mean_token_accuracy": 0.9499097801744938,
      "num_tokens": 95083466.0,
      "step": 12830
    },
    {
      "entropy": 2.0275502778589725,
      "epoch": 0.19212047929121906,
      "grad_norm": 0.40937238931655884,
      "learning_rate": 4.63823588432861e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9443226125091314,
      "num_tokens": 95158163.0,
      "step": 12840
    },
    {
      "entropy": 1.9942583583295346,
      "epoch": 0.19227010583272314,
      "grad_norm": 0.3694326877593994,
      "learning_rate": 4.637613436150904e-05,
      "loss": 0.2052,
      "mean_token_accuracy": 0.9458387911319732,
      "num_tokens": 95231564.0,
      "step": 12850
    },
    {
      "entropy": 1.9739273957908154,
      "epoch": 0.19241973237422721,
      "grad_norm": 0.2701924741268158,
      "learning_rate": 4.6369904948750385e-05,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9469009272754192,
      "num_tokens": 95305282.0,
      "step": 12860
    },
    {
      "entropy": 2.0230797596275805,
      "epoch": 0.19256935891573126,
      "grad_norm": 0.331013947725296,
      "learning_rate": 4.636367060645048e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9392116397619248,
      "num_tokens": 95377761.0,
      "step": 12870
    },
    {
      "entropy": 2.0178587026894093,
      "epoch": 0.19271898545723534,
      "grad_norm": 0.30451303720474243,
      "learning_rate": 4.6357431336050824e-05,
      "loss": 0.2373,
      "mean_token_accuracy": 0.9406913008540869,
      "num_tokens": 95450764.0,
      "step": 12880
    },
    {
      "entropy": 2.0127994656562804,
      "epoch": 0.1928686119987394,
      "grad_norm": 0.38137465715408325,
      "learning_rate": 4.635118713899404e-05,
      "loss": 0.2324,
      "mean_token_accuracy": 0.9411035474389792,
      "num_tokens": 95521611.0,
      "step": 12890
    },
    {
      "entropy": 1.9883999943733215,
      "epoch": 0.19301823854024347,
      "grad_norm": 0.40598559379577637,
      "learning_rate": 4.634493801672391e-05,
      "loss": 0.2325,
      "mean_token_accuracy": 0.9429035481065512,
      "num_tokens": 95597427.0,
      "step": 12900
    },
    {
      "entropy": 2.0049871653318405,
      "epoch": 0.19316786508174752,
      "grad_norm": 0.3310081660747528,
      "learning_rate": 4.633868397068533e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9457008853554726,
      "num_tokens": 95670397.0,
      "step": 12910
    },
    {
      "entropy": 1.9839185170829297,
      "epoch": 0.1933174916232516,
      "grad_norm": 0.3417423665523529,
      "learning_rate": 4.6332425002324345e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9397353518754243,
      "num_tokens": 95744611.0,
      "step": 12920
    },
    {
      "entropy": 1.9417737230658532,
      "epoch": 0.19346711816475565,
      "grad_norm": 0.3883616030216217,
      "learning_rate": 4.632616111308814e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9417427971959114,
      "num_tokens": 95820601.0,
      "step": 12930
    },
    {
      "entropy": 1.9597273781895637,
      "epoch": 0.19361674470625972,
      "grad_norm": 0.3107667863368988,
      "learning_rate": 4.6319892304425046e-05,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9458891075104475,
      "num_tokens": 95896030.0,
      "step": 12940
    },
    {
      "entropy": 2.0041957937180994,
      "epoch": 0.19376637124776377,
      "grad_norm": 0.30913451313972473,
      "learning_rate": 4.631361857778452e-05,
      "loss": 0.2564,
      "mean_token_accuracy": 0.9401714570820332,
      "num_tokens": 95969492.0,
      "step": 12950
    },
    {
      "entropy": 1.9882222041487694,
      "epoch": 0.19391599778926785,
      "grad_norm": 0.4230377674102783,
      "learning_rate": 4.630733993461715e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9444467082619667,
      "num_tokens": 96044684.0,
      "step": 12960
    },
    {
      "entropy": 1.9750345475971698,
      "epoch": 0.1940656243307719,
      "grad_norm": 0.3696315586566925,
      "learning_rate": 4.630105637637469e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9450843960046769,
      "num_tokens": 96123136.0,
      "step": 12970
    },
    {
      "entropy": 1.9689860306680202,
      "epoch": 0.19421525087227598,
      "grad_norm": 0.3193543553352356,
      "learning_rate": 4.629476790450999e-05,
      "loss": 0.2485,
      "mean_token_accuracy": 0.9359623327851295,
      "num_tokens": 96196828.0,
      "step": 12980
    },
    {
      "entropy": 2.0244121462106706,
      "epoch": 0.19436487741378006,
      "grad_norm": 0.36437466740608215,
      "learning_rate": 4.628847452047707e-05,
      "loss": 0.2521,
      "mean_token_accuracy": 0.9398603431880475,
      "num_tokens": 96268549.0,
      "step": 12990
    },
    {
      "entropy": 2.027615471184254,
      "epoch": 0.1945145039552841,
      "grad_norm": 0.33267271518707275,
      "learning_rate": 4.628217622573107e-05,
      "loss": 0.2169,
      "mean_token_accuracy": 0.9415725883096456,
      "num_tokens": 96342161.0,
      "step": 13000
    },
    {
      "epoch": 0.1945145039552841,
      "eval_entropy": 1.992415794301033,
      "eval_loss": 0.24962739646434784,
      "eval_mean_token_accuracy": 0.9442220282196998,
      "eval_num_tokens": 96342161.0,
      "eval_runtime": 544.4218,
      "eval_samples_per_second": 36.736,
      "eval_steps_per_second": 9.184,
      "step": 13000
    },
    {
      "entropy": 1.9892649523913861,
      "epoch": 0.19466413049678818,
      "grad_norm": 0.42806655168533325,
      "learning_rate": 4.627587302172825e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9446578934788704,
      "num_tokens": 96415995.0,
      "step": 13010
    },
    {
      "entropy": 2.0028172358870506,
      "epoch": 0.19481375703829223,
      "grad_norm": 0.3377649784088135,
      "learning_rate": 4.626956490992605e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9403172481805087,
      "num_tokens": 96488301.0,
      "step": 13020
    },
    {
      "entropy": 2.0120084576308725,
      "epoch": 0.1949633835797963,
      "grad_norm": 0.35164982080459595,
      "learning_rate": 4.6263251891783006e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9391670566052198,
      "num_tokens": 96562278.0,
      "step": 13030
    },
    {
      "entropy": 1.9583049647510051,
      "epoch": 0.19511301012130036,
      "grad_norm": 0.28708845376968384,
      "learning_rate": 4.62569339687588e-05,
      "loss": 0.1987,
      "mean_token_accuracy": 0.9480591140687465,
      "num_tokens": 96637269.0,
      "step": 13040
    },
    {
      "entropy": 1.9730226568877698,
      "epoch": 0.19526263666280444,
      "grad_norm": 0.3432137370109558,
      "learning_rate": 4.625061114231425e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.943403460457921,
      "num_tokens": 96712005.0,
      "step": 13050
    },
    {
      "entropy": 2.010793086141348,
      "epoch": 0.1954122632043085,
      "grad_norm": 0.29028448462486267,
      "learning_rate": 4.62442834139113e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9441841159015893,
      "num_tokens": 96787071.0,
      "step": 13060
    },
    {
      "entropy": 2.051914893090725,
      "epoch": 0.19556188974581257,
      "grad_norm": 0.30175355076789856,
      "learning_rate": 4.623795078501303e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9403054881840944,
      "num_tokens": 96858996.0,
      "step": 13070
    },
    {
      "entropy": 2.044176313281059,
      "epoch": 0.19571151628731662,
      "grad_norm": 0.3076680302619934,
      "learning_rate": 4.623161325708367e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9441388368606567,
      "num_tokens": 96934462.0,
      "step": 13080
    },
    {
      "entropy": 2.0268160715699195,
      "epoch": 0.1958611428288207,
      "grad_norm": 0.3417598009109497,
      "learning_rate": 4.6225270831588566e-05,
      "loss": 0.2104,
      "mean_token_accuracy": 0.9461489867419004,
      "num_tokens": 97009842.0,
      "step": 13090
    },
    {
      "entropy": 2.0121673926711083,
      "epoch": 0.19601076937032474,
      "grad_norm": 0.36367419362068176,
      "learning_rate": 4.621892350999419e-05,
      "loss": 0.2013,
      "mean_token_accuracy": 0.948818153142929,
      "num_tokens": 97086526.0,
      "step": 13100
    },
    {
      "entropy": 2.0380085848271845,
      "epoch": 0.19616039591182882,
      "grad_norm": 0.36796584725379944,
      "learning_rate": 4.621257129376817e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9436346258968115,
      "num_tokens": 97161419.0,
      "step": 13110
    },
    {
      "entropy": 2.028250318020582,
      "epoch": 0.19631002245333287,
      "grad_norm": 0.3301417827606201,
      "learning_rate": 4.6206214184379235e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9452169176191092,
      "num_tokens": 97234842.0,
      "step": 13120
    },
    {
      "entropy": 2.0282094910740853,
      "epoch": 0.19645964899483695,
      "grad_norm": 0.34328019618988037,
      "learning_rate": 4.619985218329728e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9430899232625961,
      "num_tokens": 97309630.0,
      "step": 13130
    },
    {
      "entropy": 2.0284452505409716,
      "epoch": 0.19660927553634103,
      "grad_norm": 0.32580190896987915,
      "learning_rate": 4.619348529199329e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9462859068065882,
      "num_tokens": 97383689.0,
      "step": 13140
    },
    {
      "entropy": 2.0305058397352695,
      "epoch": 0.19675890207784508,
      "grad_norm": 0.3518810272216797,
      "learning_rate": 4.618711351193943e-05,
      "loss": 0.2144,
      "mean_token_accuracy": 0.9489344019442797,
      "num_tokens": 97457261.0,
      "step": 13150
    },
    {
      "entropy": 2.0396322667598725,
      "epoch": 0.19690852861934915,
      "grad_norm": 0.34808966517448425,
      "learning_rate": 4.6180736844608944e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9473258048295975,
      "num_tokens": 97532004.0,
      "step": 13160
    },
    {
      "entropy": 2.0629574082791806,
      "epoch": 0.1970581551608532,
      "grad_norm": 0.282895565032959,
      "learning_rate": 4.617435529147625e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9454885926097631,
      "num_tokens": 97605922.0,
      "step": 13170
    },
    {
      "entropy": 2.0492070063948633,
      "epoch": 0.19720778170235728,
      "grad_norm": 0.3502848148345947,
      "learning_rate": 4.616796885401686e-05,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9461462937295437,
      "num_tokens": 97680663.0,
      "step": 13180
    },
    {
      "entropy": 2.0526459857821466,
      "epoch": 0.19735740824386133,
      "grad_norm": 0.4252983033657074,
      "learning_rate": 4.616157753370745e-05,
      "loss": 0.2402,
      "mean_token_accuracy": 0.9420736491680145,
      "num_tokens": 97753982.0,
      "step": 13190
    },
    {
      "entropy": 2.031171438097954,
      "epoch": 0.1975070347853654,
      "grad_norm": 0.39808085560798645,
      "learning_rate": 4.615518133202579e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9456831488758326,
      "num_tokens": 97828807.0,
      "step": 13200
    },
    {
      "entropy": 2.0179490357637406,
      "epoch": 0.19765666132686946,
      "grad_norm": 0.37608644366264343,
      "learning_rate": 4.614878025045081e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9439956180751323,
      "num_tokens": 97903240.0,
      "step": 13210
    },
    {
      "entropy": 2.0230360493063926,
      "epoch": 0.19780628786837354,
      "grad_norm": 0.3312515914440155,
      "learning_rate": 4.614237429046255e-05,
      "loss": 0.2056,
      "mean_token_accuracy": 0.9471672214567661,
      "num_tokens": 97978637.0,
      "step": 13220
    },
    {
      "entropy": 2.020954118669033,
      "epoch": 0.1979559144098776,
      "grad_norm": 0.33968374133110046,
      "learning_rate": 4.613596345354217e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9415618892759084,
      "num_tokens": 98050588.0,
      "step": 13230
    },
    {
      "entropy": 2.005578847974539,
      "epoch": 0.19810554095138166,
      "grad_norm": 0.31122902035713196,
      "learning_rate": 4.612954774117198e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9435742847621441,
      "num_tokens": 98126415.0,
      "step": 13240
    },
    {
      "entropy": 2.024767728149891,
      "epoch": 0.19825516749288571,
      "grad_norm": 0.3609143793582916,
      "learning_rate": 4.612312715483541e-05,
      "loss": 0.2569,
      "mean_token_accuracy": 0.9379268161952495,
      "num_tokens": 98197156.0,
      "step": 13250
    },
    {
      "entropy": 2.000527998059988,
      "epoch": 0.1984047940343898,
      "grad_norm": 0.3371099829673767,
      "learning_rate": 4.6116701696017e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9432230889797211,
      "num_tokens": 98273879.0,
      "step": 13260
    },
    {
      "entropy": 1.9830008544027806,
      "epoch": 0.19855442057589384,
      "grad_norm": 0.38208097219467163,
      "learning_rate": 4.6110271366202455e-05,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9472452659159899,
      "num_tokens": 98351084.0,
      "step": 13270
    },
    {
      "entropy": 2.0097381584346294,
      "epoch": 0.19870404711739792,
      "grad_norm": 0.29632511734962463,
      "learning_rate": 4.610383616687856e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9429416820406914,
      "num_tokens": 98425663.0,
      "step": 13280
    },
    {
      "entropy": 2.0016107082366945,
      "epoch": 0.198853673658902,
      "grad_norm": 0.37086182832717896,
      "learning_rate": 4.6097396099533255e-05,
      "loss": 0.1971,
      "mean_token_accuracy": 0.9506999772042036,
      "num_tokens": 98501723.0,
      "step": 13290
    },
    {
      "entropy": 1.9863453939557076,
      "epoch": 0.19900330020040605,
      "grad_norm": 0.378793329000473,
      "learning_rate": 4.609095116565559e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9500246290117502,
      "num_tokens": 98576940.0,
      "step": 13300
    },
    {
      "entropy": 1.9806398324668408,
      "epoch": 0.19915292674191012,
      "grad_norm": 0.36172476410865784,
      "learning_rate": 4.608450136673576e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.945240779966116,
      "num_tokens": 98650287.0,
      "step": 13310
    },
    {
      "entropy": 2.0126746609807014,
      "epoch": 0.19930255328341417,
      "grad_norm": 0.3847506642341614,
      "learning_rate": 4.6078046704265064e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9469317834824323,
      "num_tokens": 98723445.0,
      "step": 13320
    },
    {
      "entropy": 1.9976569883525372,
      "epoch": 0.19945217982491825,
      "grad_norm": 0.32241061329841614,
      "learning_rate": 4.6071587179735945e-05,
      "loss": 0.2088,
      "mean_token_accuracy": 0.9470632802695036,
      "num_tokens": 98796170.0,
      "step": 13330
    },
    {
      "entropy": 1.984074392169714,
      "epoch": 0.1996018063664223,
      "grad_norm": 0.32582202553749084,
      "learning_rate": 4.6065122794641954e-05,
      "loss": 0.201,
      "mean_token_accuracy": 0.9485153377056121,
      "num_tokens": 98873005.0,
      "step": 13340
    },
    {
      "entropy": 1.9937818862497807,
      "epoch": 0.19975143290792638,
      "grad_norm": 0.31795161962509155,
      "learning_rate": 4.605865355047778e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9431983504444361,
      "num_tokens": 98946320.0,
      "step": 13350
    },
    {
      "entropy": 1.982643087953329,
      "epoch": 0.19990105944943043,
      "grad_norm": 0.3697422444820404,
      "learning_rate": 4.6052179448739206e-05,
      "loss": 0.2128,
      "mean_token_accuracy": 0.9459359519183635,
      "num_tokens": 99021575.0,
      "step": 13360
    },
    {
      "entropy": 1.956934154778719,
      "epoch": 0.2000506859909345,
      "grad_norm": 0.3301470875740051,
      "learning_rate": 4.604570049092318e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9443233996629715,
      "num_tokens": 99096846.0,
      "step": 13370
    },
    {
      "entropy": 1.946332136541605,
      "epoch": 0.20020031253243856,
      "grad_norm": 0.38618722558021545,
      "learning_rate": 4.603921667852774e-05,
      "loss": 0.1992,
      "mean_token_accuracy": 0.9492907710373402,
      "num_tokens": 99171795.0,
      "step": 13380
    },
    {
      "entropy": 1.9673570565879346,
      "epoch": 0.20034993907394263,
      "grad_norm": 0.4620040953159332,
      "learning_rate": 4.603272801305206e-05,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9461702398955822,
      "num_tokens": 99245991.0,
      "step": 13390
    },
    {
      "entropy": 1.9973426215350627,
      "epoch": 0.20049956561544668,
      "grad_norm": 0.333280473947525,
      "learning_rate": 4.602623449599644e-05,
      "loss": 0.2043,
      "mean_token_accuracy": 0.9497626837342977,
      "num_tokens": 99319184.0,
      "step": 13400
    },
    {
      "entropy": 2.023459393531084,
      "epoch": 0.20064919215695076,
      "grad_norm": 0.35945624113082886,
      "learning_rate": 4.601973612886231e-05,
      "loss": 0.2357,
      "mean_token_accuracy": 0.9375539746135473,
      "num_tokens": 99393675.0,
      "step": 13410
    },
    {
      "entropy": 2.0211163640022276,
      "epoch": 0.2007988186984548,
      "grad_norm": 0.3786636292934418,
      "learning_rate": 4.601323291315218e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.945373747497797,
      "num_tokens": 99466708.0,
      "step": 13420
    },
    {
      "entropy": 2.0557154424488546,
      "epoch": 0.2009484452399589,
      "grad_norm": 0.3517566919326782,
      "learning_rate": 4.600672485036973e-05,
      "loss": 0.2522,
      "mean_token_accuracy": 0.9366081919521093,
      "num_tokens": 99539955.0,
      "step": 13430
    },
    {
      "entropy": 2.029182569682598,
      "epoch": 0.20109807178146297,
      "grad_norm": 0.31801098585128784,
      "learning_rate": 4.600021194201973e-05,
      "loss": 0.2127,
      "mean_token_accuracy": 0.947599996253848,
      "num_tokens": 99616424.0,
      "step": 13440
    },
    {
      "entropy": 2.0114041179418565,
      "epoch": 0.20124769832296702,
      "grad_norm": 0.29898601770401,
      "learning_rate": 4.5993694189608085e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9424703534692526,
      "num_tokens": 99693559.0,
      "step": 13450
    },
    {
      "entropy": 2.004046581685543,
      "epoch": 0.2013973248644711,
      "grad_norm": 0.2844940423965454,
      "learning_rate": 4.59871715946418e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9415409177541733,
      "num_tokens": 99766196.0,
      "step": 13460
    },
    {
      "entropy": 1.9924535147845746,
      "epoch": 0.20154695140597514,
      "grad_norm": 0.3602909445762634,
      "learning_rate": 4.598064415862904e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9447620656341315,
      "num_tokens": 99838637.0,
      "step": 13470
    },
    {
      "entropy": 2.0252523370087148,
      "epoch": 0.20169657794747922,
      "grad_norm": 0.31451576948165894,
      "learning_rate": 4.597411188307906e-05,
      "loss": 0.2517,
      "mean_token_accuracy": 0.9372601445764304,
      "num_tokens": 99911721.0,
      "step": 13480
    },
    {
      "entropy": 2.0222074188292027,
      "epoch": 0.20184620448898327,
      "grad_norm": 0.3806108832359314,
      "learning_rate": 4.596757476950221e-05,
      "loss": 0.2061,
      "mean_token_accuracy": 0.9478537701070309,
      "num_tokens": 99985032.0,
      "step": 13490
    },
    {
      "entropy": 2.0263987362384794,
      "epoch": 0.20199583103048735,
      "grad_norm": 0.33649447560310364,
      "learning_rate": 4.596103281941003e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9429521523416042,
      "num_tokens": 100057342.0,
      "step": 13500
    },
    {
      "entropy": 2.006431180983782,
      "epoch": 0.2021454575719914,
      "grad_norm": 0.3981854319572449,
      "learning_rate": 4.595448603431509e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9443598583340644,
      "num_tokens": 100132735.0,
      "step": 13510
    },
    {
      "entropy": 1.9652430541813373,
      "epoch": 0.20229508411349548,
      "grad_norm": 0.3931941092014313,
      "learning_rate": 4.5947934415731164e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9428030338138342,
      "num_tokens": 100205902.0,
      "step": 13520
    },
    {
      "entropy": 1.9791137717664242,
      "epoch": 0.20244471065499953,
      "grad_norm": 0.36476054787635803,
      "learning_rate": 4.594137796517308e-05,
      "loss": 0.2315,
      "mean_token_accuracy": 0.9434269558638334,
      "num_tokens": 100280315.0,
      "step": 13530
    },
    {
      "entropy": 2.003652787953615,
      "epoch": 0.2025943371965036,
      "grad_norm": 0.3404044210910797,
      "learning_rate": 4.593481668415681e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9437274042516947,
      "num_tokens": 100355765.0,
      "step": 13540
    },
    {
      "entropy": 2.0061951696872713,
      "epoch": 0.20274396373800765,
      "grad_norm": 0.2875403165817261,
      "learning_rate": 4.5928250574199436e-05,
      "loss": 0.2049,
      "mean_token_accuracy": 0.9470361057668925,
      "num_tokens": 100431129.0,
      "step": 13550
    },
    {
      "entropy": 2.029212936758995,
      "epoch": 0.20289359027951173,
      "grad_norm": 0.2941003739833832,
      "learning_rate": 4.592167963681917e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9400687415152789,
      "num_tokens": 100503907.0,
      "step": 13560
    },
    {
      "entropy": 2.038173444569111,
      "epoch": 0.2030432168210158,
      "grad_norm": 0.38190141320228577,
      "learning_rate": 4.591510387353531e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9434329379349947,
      "num_tokens": 100577285.0,
      "step": 13570
    },
    {
      "entropy": 2.0054588414728642,
      "epoch": 0.20319284336251986,
      "grad_norm": 0.26861098408699036,
      "learning_rate": 4.590852328586831e-05,
      "loss": 0.2088,
      "mean_token_accuracy": 0.9485218290239572,
      "num_tokens": 100653379.0,
      "step": 13580
    },
    {
      "entropy": 2.038307202607393,
      "epoch": 0.20334246990402394,
      "grad_norm": 0.4251239001750946,
      "learning_rate": 4.590193787533972e-05,
      "loss": 0.2393,
      "mean_token_accuracy": 0.937624192610383,
      "num_tokens": 100726169.0,
      "step": 13590
    },
    {
      "entropy": 2.022958455234766,
      "epoch": 0.203492096445528,
      "grad_norm": 0.3239625096321106,
      "learning_rate": 4.589534764347218e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9460875399410724,
      "num_tokens": 100803277.0,
      "step": 13600
    },
    {
      "entropy": 2.0323875240981577,
      "epoch": 0.20364172298703206,
      "grad_norm": 0.3710615336894989,
      "learning_rate": 4.588875259178949e-05,
      "loss": 0.2309,
      "mean_token_accuracy": 0.9419273141771555,
      "num_tokens": 100873353.0,
      "step": 13610
    },
    {
      "entropy": 2.015052127093077,
      "epoch": 0.20379134952853611,
      "grad_norm": 0.4614577889442444,
      "learning_rate": 4.588215272181654e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9371887370944023,
      "num_tokens": 100946619.0,
      "step": 13620
    },
    {
      "entropy": 2.0216915301978586,
      "epoch": 0.2039409760700402,
      "grad_norm": 0.33102545142173767,
      "learning_rate": 4.587554803507934e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9474774982780219,
      "num_tokens": 101023012.0,
      "step": 13630
    },
    {
      "entropy": 2.0015990279614924,
      "epoch": 0.20409060261154424,
      "grad_norm": 0.379935085773468,
      "learning_rate": 4.5868938533104994e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9469445217400789,
      "num_tokens": 101099582.0,
      "step": 13640
    },
    {
      "entropy": 1.9948401413857937,
      "epoch": 0.20424022915304832,
      "grad_norm": 0.3240996301174164,
      "learning_rate": 4.5862324217421745e-05,
      "loss": 0.2474,
      "mean_token_accuracy": 0.9417290586978198,
      "num_tokens": 101172142.0,
      "step": 13650
    },
    {
      "entropy": 2.0060148909687996,
      "epoch": 0.20438985569455237,
      "grad_norm": 0.32297807931900024,
      "learning_rate": 4.585570508955895e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9453066024929285,
      "num_tokens": 101247694.0,
      "step": 13660
    },
    {
      "entropy": 2.0866540536284446,
      "epoch": 0.20453948223605645,
      "grad_norm": 0.3442113995552063,
      "learning_rate": 4.584908115104707e-05,
      "loss": 0.261,
      "mean_token_accuracy": 0.9361680295318365,
      "num_tokens": 101319108.0,
      "step": 13670
    },
    {
      "entropy": 2.062214882671833,
      "epoch": 0.2046891087775605,
      "grad_norm": 0.33941927552223206,
      "learning_rate": 4.584245240341766e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9421812366694212,
      "num_tokens": 101394331.0,
      "step": 13680
    },
    {
      "entropy": 2.0513444118201734,
      "epoch": 0.20483873531906457,
      "grad_norm": 0.32651960849761963,
      "learning_rate": 4.583581884820341e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9456468645483256,
      "num_tokens": 101466364.0,
      "step": 13690
    },
    {
      "entropy": 2.0187342561781407,
      "epoch": 0.20498836186056862,
      "grad_norm": 0.38680827617645264,
      "learning_rate": 4.582918048693814e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.941051097586751,
      "num_tokens": 101540124.0,
      "step": 13700
    },
    {
      "entropy": 2.020453943312168,
      "epoch": 0.2051379884020727,
      "grad_norm": 0.3135911226272583,
      "learning_rate": 4.5822537321156714e-05,
      "loss": 0.2376,
      "mean_token_accuracy": 0.9436797488480806,
      "num_tokens": 101614131.0,
      "step": 13710
    },
    {
      "entropy": 2.0127654664218424,
      "epoch": 0.20528761494357678,
      "grad_norm": 0.36478978395462036,
      "learning_rate": 4.581588935239519e-05,
      "loss": 0.2065,
      "mean_token_accuracy": 0.9487535905092954,
      "num_tokens": 101687314.0,
      "step": 13720
    },
    {
      "entropy": 2.000778838992119,
      "epoch": 0.20543724148508083,
      "grad_norm": 0.3725127577781677,
      "learning_rate": 4.5809236582190675e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.945598254725337,
      "num_tokens": 101759078.0,
      "step": 13730
    },
    {
      "entropy": 2.0251752883195877,
      "epoch": 0.2055868680265849,
      "grad_norm": 0.3356815576553345,
      "learning_rate": 4.580257901208141e-05,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9374097749590874,
      "num_tokens": 101831559.0,
      "step": 13740
    },
    {
      "entropy": 2.0035163924098014,
      "epoch": 0.20573649456808896,
      "grad_norm": 0.35103824734687805,
      "learning_rate": 4.579591664360675e-05,
      "loss": 0.2107,
      "mean_token_accuracy": 0.945456625148654,
      "num_tokens": 101905878.0,
      "step": 13750
    },
    {
      "entropy": 1.991134136915207,
      "epoch": 0.20588612110959303,
      "grad_norm": 0.3809353709220886,
      "learning_rate": 4.5789249478307145e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9493240788578987,
      "num_tokens": 101980615.0,
      "step": 13760
    },
    {
      "entropy": 2.06479157358408,
      "epoch": 0.20603574765109708,
      "grad_norm": 0.3452010452747345,
      "learning_rate": 4.578257751772417e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9434987671673298,
      "num_tokens": 102054692.0,
      "step": 13770
    },
    {
      "entropy": 2.0690784364938737,
      "epoch": 0.20618537419260116,
      "grad_norm": 0.3188270032405853,
      "learning_rate": 4.5775900763400495e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9364047430455684,
      "num_tokens": 102127223.0,
      "step": 13780
    },
    {
      "entropy": 2.0650348074734213,
      "epoch": 0.2063350007341052,
      "grad_norm": 0.4257359206676483,
      "learning_rate": 4.5769219216879914e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.9428519736975431,
      "num_tokens": 102200180.0,
      "step": 13790
    },
    {
      "entropy": 2.052281577140093,
      "epoch": 0.2064846272756093,
      "grad_norm": 0.407749742269516,
      "learning_rate": 4.57625328797073e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9436328608542681,
      "num_tokens": 102273246.0,
      "step": 13800
    },
    {
      "entropy": 2.0550562888383865,
      "epoch": 0.20663425381711334,
      "grad_norm": 0.3340471684932709,
      "learning_rate": 4.575584175342868e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9412753332406283,
      "num_tokens": 102345233.0,
      "step": 13810
    },
    {
      "entropy": 2.0366390213370322,
      "epoch": 0.20678388035861742,
      "grad_norm": 0.3973354399204254,
      "learning_rate": 4.574914583959114e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.9421418771147728,
      "num_tokens": 102417374.0,
      "step": 13820
    },
    {
      "entropy": 2.0372855864465236,
      "epoch": 0.20693350690012147,
      "grad_norm": 0.2730023264884949,
      "learning_rate": 4.574244513974291e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9400602508336305,
      "num_tokens": 102494337.0,
      "step": 13830
    },
    {
      "entropy": 2.0125115908682347,
      "epoch": 0.20708313344162554,
      "grad_norm": 0.4082595407962799,
      "learning_rate": 4.5735739655433285e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9440697595477104,
      "num_tokens": 102568352.0,
      "step": 13840
    },
    {
      "entropy": 2.0475587539374827,
      "epoch": 0.2072327599831296,
      "grad_norm": 0.33937376737594604,
      "learning_rate": 4.5729029388212725e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9444423895329237,
      "num_tokens": 102642015.0,
      "step": 13850
    },
    {
      "entropy": 2.045319443196058,
      "epoch": 0.20738238652463367,
      "grad_norm": 0.3459959030151367,
      "learning_rate": 4.572231433963274e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9403028655797243,
      "num_tokens": 102713298.0,
      "step": 13860
    },
    {
      "entropy": 2.042854976654053,
      "epoch": 0.20753201306613775,
      "grad_norm": 0.4001365005970001,
      "learning_rate": 4.571559451124598e-05,
      "loss": 0.2442,
      "mean_token_accuracy": 0.9411084968596697,
      "num_tokens": 102785749.0,
      "step": 13870
    },
    {
      "entropy": 2.004806100577116,
      "epoch": 0.2076816396076418,
      "grad_norm": 0.3483239412307739,
      "learning_rate": 4.570886990460618e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9450951226055622,
      "num_tokens": 102858265.0,
      "step": 13880
    },
    {
      "entropy": 2.031186766922474,
      "epoch": 0.20783126614914588,
      "grad_norm": 0.3541969954967499,
      "learning_rate": 4.57021405212682e-05,
      "loss": 0.257,
      "mean_token_accuracy": 0.933038930222392,
      "num_tokens": 102932864.0,
      "step": 13890
    },
    {
      "entropy": 2.0399622060358524,
      "epoch": 0.20798089269064993,
      "grad_norm": 0.2990318834781647,
      "learning_rate": 4.5695406362787985e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9412774406373501,
      "num_tokens": 103003551.0,
      "step": 13900
    },
    {
      "entropy": 2.0386382013559343,
      "epoch": 0.208130519232154,
      "grad_norm": 0.3305821120738983,
      "learning_rate": 4.5688667430722594e-05,
      "loss": 0.2089,
      "mean_token_accuracy": 0.9443869732320309,
      "num_tokens": 103076773.0,
      "step": 13910
    },
    {
      "entropy": 2.0051053181290626,
      "epoch": 0.20828014577365805,
      "grad_norm": 0.33082348108291626,
      "learning_rate": 4.5681923726630185e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9427927855402232,
      "num_tokens": 103152837.0,
      "step": 13920
    },
    {
      "entropy": 2.0167103253304957,
      "epoch": 0.20842977231516213,
      "grad_norm": 0.410129189491272,
      "learning_rate": 4.567517525207004e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9435369193553924,
      "num_tokens": 103227863.0,
      "step": 13930
    },
    {
      "entropy": 2.0337758123874665,
      "epoch": 0.20857939885666618,
      "grad_norm": 0.3676702380180359,
      "learning_rate": 4.56684220086025e-05,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9445686791092157,
      "num_tokens": 103300551.0,
      "step": 13940
    },
    {
      "entropy": 2.0316509559750555,
      "epoch": 0.20872902539817026,
      "grad_norm": 0.32173505425453186,
      "learning_rate": 4.5661663997789046e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.943427013233304,
      "num_tokens": 103373926.0,
      "step": 13950
    },
    {
      "entropy": 2.028409276902676,
      "epoch": 0.2088786519396743,
      "grad_norm": 0.34493017196655273,
      "learning_rate": 4.565490122119225e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9404893536120653,
      "num_tokens": 103448077.0,
      "step": 13960
    },
    {
      "entropy": 2.0482101306319236,
      "epoch": 0.2090282784811784,
      "grad_norm": 0.37653908133506775,
      "learning_rate": 4.56481336803758e-05,
      "loss": 0.254,
      "mean_token_accuracy": 0.9387929130345583,
      "num_tokens": 103522070.0,
      "step": 13970
    },
    {
      "entropy": 2.061176474392414,
      "epoch": 0.20917790502268244,
      "grad_norm": 0.32369285821914673,
      "learning_rate": 4.564136137690445e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9469171587377787,
      "num_tokens": 103598005.0,
      "step": 13980
    },
    {
      "entropy": 2.059191434085369,
      "epoch": 0.20932753156418651,
      "grad_norm": 0.35450130701065063,
      "learning_rate": 4.563458431234409e-05,
      "loss": 0.2468,
      "mean_token_accuracy": 0.9392203651368618,
      "num_tokens": 103671656.0,
      "step": 13990
    },
    {
      "entropy": 2.017102116346359,
      "epoch": 0.20947715810569056,
      "grad_norm": 0.3949444890022278,
      "learning_rate": 4.562780248826168e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9415003594011069,
      "num_tokens": 103744642.0,
      "step": 14000
    },
    {
      "epoch": 0.20947715810569056,
      "eval_entropy": 2.0080390424013137,
      "eval_loss": 0.24956941604614258,
      "eval_mean_token_accuracy": 0.9442813966870308,
      "eval_num_tokens": 103744642.0,
      "eval_runtime": 545.3147,
      "eval_samples_per_second": 36.676,
      "eval_steps_per_second": 9.169,
      "step": 14000
    },
    {
      "entropy": 2.0273719415068627,
      "epoch": 0.20962678464719464,
      "grad_norm": 0.48935794830322266,
      "learning_rate": 4.562101590622533e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9418709628283978,
      "num_tokens": 103819933.0,
      "step": 14010
    },
    {
      "entropy": 2.0217023603618145,
      "epoch": 0.20977641118869872,
      "grad_norm": 0.3255375027656555,
      "learning_rate": 4.5614224567804206e-05,
      "loss": 0.2084,
      "mean_token_accuracy": 0.9497186955064535,
      "num_tokens": 103897382.0,
      "step": 14020
    },
    {
      "entropy": 2.0894786074757574,
      "epoch": 0.20992603773020277,
      "grad_norm": 0.36979830265045166,
      "learning_rate": 4.560742847456856e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9374333295971156,
      "num_tokens": 103968011.0,
      "step": 14030
    },
    {
      "entropy": 2.0647294141352175,
      "epoch": 0.21007566427170685,
      "grad_norm": 0.3802686035633087,
      "learning_rate": 4.56006276280898e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9426207568496465,
      "num_tokens": 104041370.0,
      "step": 14040
    },
    {
      "entropy": 2.032048621028662,
      "epoch": 0.2102252908132109,
      "grad_norm": 0.3462115526199341,
      "learning_rate": 4.5593822029940396e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9428571213036776,
      "num_tokens": 104116049.0,
      "step": 14050
    },
    {
      "entropy": 2.0698194071650504,
      "epoch": 0.21037491735471497,
      "grad_norm": 0.35200339555740356,
      "learning_rate": 4.5587011681693924e-05,
      "loss": 0.2112,
      "mean_token_accuracy": 0.943917902559042,
      "num_tokens": 104189171.0,
      "step": 14060
    },
    {
      "entropy": 2.062006044387817,
      "epoch": 0.21052454389621902,
      "grad_norm": 0.3327452540397644,
      "learning_rate": 4.558019658492505e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9459799945354461,
      "num_tokens": 104262940.0,
      "step": 14070
    },
    {
      "entropy": 2.1112869054079058,
      "epoch": 0.2106741704377231,
      "grad_norm": 0.3299127519130707,
      "learning_rate": 4.557337674120955e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9473522700369358,
      "num_tokens": 104336884.0,
      "step": 14080
    },
    {
      "entropy": 2.0866345278918743,
      "epoch": 0.21082379697922715,
      "grad_norm": 0.30508989095687866,
      "learning_rate": 4.5566552152124306e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.942463006079197,
      "num_tokens": 104410565.0,
      "step": 14090
    },
    {
      "entropy": 2.0396482326090335,
      "epoch": 0.21097342352073123,
      "grad_norm": 0.311389297246933,
      "learning_rate": 4.5559722819247255e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9408155672252179,
      "num_tokens": 104485578.0,
      "step": 14100
    },
    {
      "entropy": 2.059272352606058,
      "epoch": 0.21112305006223528,
      "grad_norm": 0.3001222610473633,
      "learning_rate": 4.5552888744157485e-05,
      "loss": 0.2024,
      "mean_token_accuracy": 0.9480536639690399,
      "num_tokens": 104562070.0,
      "step": 14110
    },
    {
      "entropy": 2.0498397432267668,
      "epoch": 0.21127267660373936,
      "grad_norm": 0.3171306252479553,
      "learning_rate": 4.554604992843514e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9457856170833111,
      "num_tokens": 104635469.0,
      "step": 14120
    },
    {
      "entropy": 2.0390916019678116,
      "epoch": 0.2114223031452434,
      "grad_norm": 0.40084949135780334,
      "learning_rate": 4.55392063736615e-05,
      "loss": 0.2021,
      "mean_token_accuracy": 0.950303491950035,
      "num_tokens": 104710055.0,
      "step": 14130
    },
    {
      "entropy": 2.0825776025652885,
      "epoch": 0.21157192968674748,
      "grad_norm": 0.4157717227935791,
      "learning_rate": 4.553235808141888e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9444057192653418,
      "num_tokens": 104780604.0,
      "step": 14140
    },
    {
      "entropy": 2.077703186869621,
      "epoch": 0.21172155622825156,
      "grad_norm": 0.3221442401409149,
      "learning_rate": 4.5525505053290756e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9470340929925442,
      "num_tokens": 104853872.0,
      "step": 14150
    },
    {
      "entropy": 2.047640284895897,
      "epoch": 0.2118711827697556,
      "grad_norm": 0.3039309084415436,
      "learning_rate": 4.551864729086166e-05,
      "loss": 0.2095,
      "mean_token_accuracy": 0.9490785293281079,
      "num_tokens": 104929782.0,
      "step": 14160
    },
    {
      "entropy": 2.044596544653177,
      "epoch": 0.2120208093112597,
      "grad_norm": 0.3463141620159149,
      "learning_rate": 4.551178479571724e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9413029212504626,
      "num_tokens": 105005706.0,
      "step": 14170
    },
    {
      "entropy": 1.9910026602447033,
      "epoch": 0.21217043585276374,
      "grad_norm": 0.3136058449745178,
      "learning_rate": 4.55049175694442e-05,
      "loss": 0.2077,
      "mean_token_accuracy": 0.9501286294311285,
      "num_tokens": 105080213.0,
      "step": 14180
    },
    {
      "entropy": 2.009447134286165,
      "epoch": 0.21232006239426782,
      "grad_norm": 0.32847824692726135,
      "learning_rate": 4.5498045613630396e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9423948977142572,
      "num_tokens": 105152601.0,
      "step": 14190
    },
    {
      "entropy": 2.008737666159868,
      "epoch": 0.21246968893577187,
      "grad_norm": 0.3503546118736267,
      "learning_rate": 4.549116892986473e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9447012137621641,
      "num_tokens": 105225433.0,
      "step": 14200
    },
    {
      "entropy": 1.9871891394257546,
      "epoch": 0.21261931547727594,
      "grad_norm": 0.31457802653312683,
      "learning_rate": 4.548428751973721e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.943184008821845,
      "num_tokens": 105300395.0,
      "step": 14210
    },
    {
      "entropy": 2.0030731901526453,
      "epoch": 0.21276894201878,
      "grad_norm": 0.2812349498271942,
      "learning_rate": 4.5477401384838946e-05,
      "loss": 0.2401,
      "mean_token_accuracy": 0.9453381665050984,
      "num_tokens": 105377221.0,
      "step": 14220
    },
    {
      "entropy": 1.994311522692442,
      "epoch": 0.21291856856028407,
      "grad_norm": 0.2985413074493408,
      "learning_rate": 4.547051052676214e-05,
      "loss": 0.2205,
      "mean_token_accuracy": 0.9466726612299681,
      "num_tokens": 105452687.0,
      "step": 14230
    },
    {
      "entropy": 2.0383806467056274,
      "epoch": 0.21306819510178812,
      "grad_norm": 0.33854004740715027,
      "learning_rate": 4.546361494710007e-05,
      "loss": 0.2314,
      "mean_token_accuracy": 0.9397301491349935,
      "num_tokens": 105526979.0,
      "step": 14240
    },
    {
      "entropy": 1.9875376045703887,
      "epoch": 0.2132178216432922,
      "grad_norm": 0.36132553219795227,
      "learning_rate": 4.545671464744713e-05,
      "loss": 0.2051,
      "mean_token_accuracy": 0.9500968258827924,
      "num_tokens": 105601120.0,
      "step": 14250
    },
    {
      "entropy": 1.9937491610646247,
      "epoch": 0.21336744818479625,
      "grad_norm": 0.4212442934513092,
      "learning_rate": 4.544980962939878e-05,
      "loss": 0.2344,
      "mean_token_accuracy": 0.9429334599524737,
      "num_tokens": 105674339.0,
      "step": 14260
    },
    {
      "entropy": 2.010219784080982,
      "epoch": 0.21351707472630033,
      "grad_norm": 0.3434756100177765,
      "learning_rate": 4.5442899894551586e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9454343423247338,
      "num_tokens": 105749181.0,
      "step": 14270
    },
    {
      "entropy": 2.003081592917442,
      "epoch": 0.21366670126780438,
      "grad_norm": 0.3519607186317444,
      "learning_rate": 4.543598544450321e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9445916220545769,
      "num_tokens": 105822684.0,
      "step": 14280
    },
    {
      "entropy": 2.02151248306036,
      "epoch": 0.21381632780930845,
      "grad_norm": 0.3664741814136505,
      "learning_rate": 4.542906628085238e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9424382109194994,
      "num_tokens": 105897841.0,
      "step": 14290
    },
    {
      "entropy": 2.0471510611474515,
      "epoch": 0.21396595435081253,
      "grad_norm": 0.3892284333705902,
      "learning_rate": 4.542214240519895e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9425520073622465,
      "num_tokens": 105970105.0,
      "step": 14300
    },
    {
      "entropy": 2.01874038875103,
      "epoch": 0.21411558089231658,
      "grad_norm": 0.3899373412132263,
      "learning_rate": 4.541521381914381e-05,
      "loss": 0.1973,
      "mean_token_accuracy": 0.9492669120430947,
      "num_tokens": 106044014.0,
      "step": 14310
    },
    {
      "entropy": 2.05056769028306,
      "epoch": 0.21426520743382066,
      "grad_norm": 0.30089476704597473,
      "learning_rate": 4.540828052428902e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9427152283489704,
      "num_tokens": 106116822.0,
      "step": 14320
    },
    {
      "entropy": 2.0359993636608125,
      "epoch": 0.2144148339753247,
      "grad_norm": 0.4225158393383026,
      "learning_rate": 4.5401342522237645e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9437825735658407,
      "num_tokens": 106188818.0,
      "step": 14330
    },
    {
      "entropy": 2.0393473096191883,
      "epoch": 0.2145644605168288,
      "grad_norm": 0.29121410846710205,
      "learning_rate": 4.539439981459388e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9438212994486094,
      "num_tokens": 106262658.0,
      "step": 14340
    },
    {
      "entropy": 2.0159406289458275,
      "epoch": 0.21471408705833284,
      "grad_norm": 0.3678544759750366,
      "learning_rate": 4.538745240296302e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.947970000654459,
      "num_tokens": 106336352.0,
      "step": 14350
    },
    {
      "entropy": 2.035090596228838,
      "epoch": 0.21486371359983691,
      "grad_norm": 0.36536234617233276,
      "learning_rate": 4.53805002889514e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9452924955636263,
      "num_tokens": 106411675.0,
      "step": 14360
    },
    {
      "entropy": 2.020190351456404,
      "epoch": 0.21501334014134096,
      "grad_norm": 0.46197736263275146,
      "learning_rate": 4.53735434741665e-05,
      "loss": 0.2479,
      "mean_token_accuracy": 0.9390276070684195,
      "num_tokens": 106485288.0,
      "step": 14370
    },
    {
      "entropy": 1.9986189514398576,
      "epoch": 0.21516296668284504,
      "grad_norm": 0.40906208753585815,
      "learning_rate": 4.536658196021685e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9424494810402393,
      "num_tokens": 106558833.0,
      "step": 14380
    },
    {
      "entropy": 2.0325156562030315,
      "epoch": 0.2153125932243491,
      "grad_norm": 0.3710268437862396,
      "learning_rate": 4.535961574871207e-05,
      "loss": 0.1952,
      "mean_token_accuracy": 0.9485281784087419,
      "num_tokens": 106636349.0,
      "step": 14390
    },
    {
      "entropy": 2.039134042710066,
      "epoch": 0.21546221976585317,
      "grad_norm": 0.42623192071914673,
      "learning_rate": 4.535264484126289e-05,
      "loss": 0.2532,
      "mean_token_accuracy": 0.9385594550520182,
      "num_tokens": 106710212.0,
      "step": 14400
    },
    {
      "entropy": 2.0437661364674566,
      "epoch": 0.21561184630735722,
      "grad_norm": 0.3585650324821472,
      "learning_rate": 4.534566923948109e-05,
      "loss": 0.2404,
      "mean_token_accuracy": 0.9394515540450812,
      "num_tokens": 106782549.0,
      "step": 14410
    },
    {
      "entropy": 2.0139872558414935,
      "epoch": 0.2157614728488613,
      "grad_norm": 0.31033241748809814,
      "learning_rate": 4.533868894497955e-05,
      "loss": 0.2,
      "mean_token_accuracy": 0.9479148518294096,
      "num_tokens": 106858473.0,
      "step": 14420
    },
    {
      "entropy": 2.0231525883078576,
      "epoch": 0.21591109939036535,
      "grad_norm": 0.39773106575012207,
      "learning_rate": 4.533170395937225e-05,
      "loss": 0.2601,
      "mean_token_accuracy": 0.9381564032286406,
      "num_tokens": 106930704.0,
      "step": 14430
    },
    {
      "entropy": 2.0354252994060515,
      "epoch": 0.21606072593186942,
      "grad_norm": 0.38997986912727356,
      "learning_rate": 4.532471428427424e-05,
      "loss": 0.2369,
      "mean_token_accuracy": 0.9402155447751284,
      "num_tokens": 107006177.0,
      "step": 14440
    },
    {
      "entropy": 1.9895979709923268,
      "epoch": 0.2162103524733735,
      "grad_norm": 0.27290162444114685,
      "learning_rate": 4.531771992130165e-05,
      "loss": 0.2035,
      "mean_token_accuracy": 0.9463789030909538,
      "num_tokens": 107082959.0,
      "step": 14450
    },
    {
      "entropy": 2.0247496142983437,
      "epoch": 0.21635997901487755,
      "grad_norm": 0.4424523413181305,
      "learning_rate": 4.5310720872071723e-05,
      "loss": 0.2488,
      "mean_token_accuracy": 0.939605812728405,
      "num_tokens": 107158075.0,
      "step": 14460
    },
    {
      "entropy": 2.035554326325655,
      "epoch": 0.21650960555638163,
      "grad_norm": 0.3559085428714752,
      "learning_rate": 4.5303717138202734e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.9397487282752991,
      "num_tokens": 107231961.0,
      "step": 14470
    },
    {
      "entropy": 2.002565947920084,
      "epoch": 0.21665923209788568,
      "grad_norm": 0.34464335441589355,
      "learning_rate": 4.529670872131409e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9457228995859623,
      "num_tokens": 107306940.0,
      "step": 14480
    },
    {
      "entropy": 1.9886301167309284,
      "epoch": 0.21680885863938976,
      "grad_norm": 0.3407740592956543,
      "learning_rate": 4.5289695623026256e-05,
      "loss": 0.2384,
      "mean_token_accuracy": 0.9402735892683267,
      "num_tokens": 107381814.0,
      "step": 14490
    },
    {
      "entropy": 1.9978798821568489,
      "epoch": 0.2169584851808938,
      "grad_norm": 0.42909935116767883,
      "learning_rate": 4.528267784496078e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9391152616590261,
      "num_tokens": 107455316.0,
      "step": 14500
    },
    {
      "entropy": 1.9780038617551328,
      "epoch": 0.21710811172239788,
      "grad_norm": 0.33966153860092163,
      "learning_rate": 4.527565538874031e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9435544602572918,
      "num_tokens": 107526737.0,
      "step": 14510
    },
    {
      "entropy": 1.9603234604001045,
      "epoch": 0.21725773826390193,
      "grad_norm": 0.29562997817993164,
      "learning_rate": 4.526862825598857e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9456202309578657,
      "num_tokens": 107600302.0,
      "step": 14520
    },
    {
      "entropy": 1.998859702795744,
      "epoch": 0.217407364805406,
      "grad_norm": 0.3347362279891968,
      "learning_rate": 4.526159644833032e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9415851477533579,
      "num_tokens": 107677545.0,
      "step": 14530
    },
    {
      "entropy": 2.019018518924713,
      "epoch": 0.21755699134691006,
      "grad_norm": 0.37746894359588623,
      "learning_rate": 4.525455996739148e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9434488203376532,
      "num_tokens": 107751636.0,
      "step": 14540
    },
    {
      "entropy": 1.9951883628964424,
      "epoch": 0.21770661788841414,
      "grad_norm": 0.36727237701416016,
      "learning_rate": 4.524751881479898e-05,
      "loss": 0.2033,
      "mean_token_accuracy": 0.9505205690860749,
      "num_tokens": 107825718.0,
      "step": 14550
    },
    {
      "entropy": 2.028523854166269,
      "epoch": 0.2178562444299182,
      "grad_norm": 0.3333470821380615,
      "learning_rate": 4.524047299218088e-05,
      "loss": 0.2008,
      "mean_token_accuracy": 0.9488738182932138,
      "num_tokens": 107900609.0,
      "step": 14560
    },
    {
      "entropy": 1.9822638213634491,
      "epoch": 0.21800587097142227,
      "grad_norm": 0.39034590125083923,
      "learning_rate": 4.523342250116628e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9428967472165823,
      "num_tokens": 107976715.0,
      "step": 14570
    },
    {
      "entropy": 1.9745465092360974,
      "epoch": 0.21815549751292634,
      "grad_norm": 0.30179017782211304,
      "learning_rate": 4.522636734338541e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9453186102211475,
      "num_tokens": 108051634.0,
      "step": 14580
    },
    {
      "entropy": 1.9934435322880746,
      "epoch": 0.2183051240544304,
      "grad_norm": 0.35103926062583923,
      "learning_rate": 4.521930752046951e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.9403260391205549,
      "num_tokens": 108125681.0,
      "step": 14590
    },
    {
      "entropy": 2.012093395739794,
      "epoch": 0.21845475059593447,
      "grad_norm": 0.3107030987739563,
      "learning_rate": 4.521224303405097e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9447121240198613,
      "num_tokens": 108197951.0,
      "step": 14600
    },
    {
      "entropy": 2.002424023300409,
      "epoch": 0.21860437713743852,
      "grad_norm": 0.3759273290634155,
      "learning_rate": 4.5205173885763195e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9482057135552168,
      "num_tokens": 108272325.0,
      "step": 14610
    },
    {
      "entropy": 2.008123317360878,
      "epoch": 0.2187540036789426,
      "grad_norm": 0.3698536455631256,
      "learning_rate": 4.519810007724073e-05,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9434138637036085,
      "num_tokens": 108347284.0,
      "step": 14620
    },
    {
      "entropy": 2.0290424823760986,
      "epoch": 0.21890363022044665,
      "grad_norm": 0.2920144498348236,
      "learning_rate": 4.519102161011914e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9413405280560255,
      "num_tokens": 108421060.0,
      "step": 14630
    },
    {
      "entropy": 1.9961930103600025,
      "epoch": 0.21905325676195073,
      "grad_norm": 0.4342615604400635,
      "learning_rate": 4.518393848603509e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.944616636261344,
      "num_tokens": 108496282.0,
      "step": 14640
    },
    {
      "entropy": 1.9991177491843701,
      "epoch": 0.21920288330345478,
      "grad_norm": 0.37027639150619507,
      "learning_rate": 4.517685070662635e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9444262117147446,
      "num_tokens": 108569810.0,
      "step": 14650
    },
    {
      "entropy": 2.0311176896095278,
      "epoch": 0.21935250984495885,
      "grad_norm": 0.3263057470321655,
      "learning_rate": 4.516975827353172e-05,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9493883866816759,
      "num_tokens": 108643556.0,
      "step": 14660
    },
    {
      "entropy": 2.0206560507416724,
      "epoch": 0.2195021363864629,
      "grad_norm": 0.3265742361545563,
      "learning_rate": 4.5162661188391105e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9442542865872383,
      "num_tokens": 108720282.0,
      "step": 14670
    },
    {
      "entropy": 2.020593335479498,
      "epoch": 0.21965176292796698,
      "grad_norm": 0.35405778884887695,
      "learning_rate": 4.515555945284547e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9472356382757425,
      "num_tokens": 108795078.0,
      "step": 14680
    },
    {
      "entropy": 1.9945539109408856,
      "epoch": 0.21980138946947103,
      "grad_norm": 0.3349688947200775,
      "learning_rate": 4.514845306853687e-05,
      "loss": 0.2142,
      "mean_token_accuracy": 0.9470288790762424,
      "num_tokens": 108871043.0,
      "step": 14690
    },
    {
      "entropy": 2.015487936139107,
      "epoch": 0.2199510160109751,
      "grad_norm": 0.2935735881328583,
      "learning_rate": 4.514134203710843e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9479257009923459,
      "num_tokens": 108945571.0,
      "step": 14700
    },
    {
      "entropy": 2.0221143081784247,
      "epoch": 0.22010064255247916,
      "grad_norm": 0.28863877058029175,
      "learning_rate": 4.5134226360204335e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9445856083184481,
      "num_tokens": 109018931.0,
      "step": 14710
    },
    {
      "entropy": 2.0221292711794376,
      "epoch": 0.22025026909398324,
      "grad_norm": 0.363619327545166,
      "learning_rate": 4.5127106039469865e-05,
      "loss": 0.215,
      "mean_token_accuracy": 0.9446193464100361,
      "num_tokens": 109091989.0,
      "step": 14720
    },
    {
      "entropy": 2.044370622932911,
      "epoch": 0.22039989563548731,
      "grad_norm": 0.32431307435035706,
      "learning_rate": 4.5119981076551365e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9402523253113031,
      "num_tokens": 109166996.0,
      "step": 14730
    },
    {
      "entropy": 2.046439695358276,
      "epoch": 0.22054952217699136,
      "grad_norm": 0.2999296486377716,
      "learning_rate": 4.5112851473096245e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9384118739515543,
      "num_tokens": 109241378.0,
      "step": 14740
    },
    {
      "entropy": 2.0250346310436727,
      "epoch": 0.22069914871849544,
      "grad_norm": 0.4018903374671936,
      "learning_rate": 4.5105717230753e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9417232543230056,
      "num_tokens": 109315486.0,
      "step": 14750
    },
    {
      "entropy": 1.9814502477645874,
      "epoch": 0.2208487752599995,
      "grad_norm": 0.2861822247505188,
      "learning_rate": 4.509857835117121e-05,
      "loss": 0.2282,
      "mean_token_accuracy": 0.9445540428161621,
      "num_tokens": 109390889.0,
      "step": 14760
    },
    {
      "entropy": 2.018892766535282,
      "epoch": 0.22099840180150357,
      "grad_norm": 0.27450478076934814,
      "learning_rate": 4.509143483600148e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9467836678028106,
      "num_tokens": 109468399.0,
      "step": 14770
    },
    {
      "entropy": 2.038572285324335,
      "epoch": 0.22114802834300762,
      "grad_norm": 0.3654242157936096,
      "learning_rate": 4.508428668689554e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9472150184214115,
      "num_tokens": 109540867.0,
      "step": 14780
    },
    {
      "entropy": 2.0337013743817804,
      "epoch": 0.2212976548845117,
      "grad_norm": 0.32926055788993835,
      "learning_rate": 4.507713390550617e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9450745921581983,
      "num_tokens": 109614102.0,
      "step": 14790
    },
    {
      "entropy": 2.020954838395119,
      "epoch": 0.22144728142601575,
      "grad_norm": 0.299465149641037,
      "learning_rate": 4.506997649348721e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9443256184458733,
      "num_tokens": 109686639.0,
      "step": 14800
    },
    {
      "entropy": 2.018161502480507,
      "epoch": 0.22159690796751982,
      "grad_norm": 0.40466201305389404,
      "learning_rate": 4.506281445249358e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9436618845909834,
      "num_tokens": 109765406.0,
      "step": 14810
    },
    {
      "entropy": 2.03743614628911,
      "epoch": 0.22174653450902387,
      "grad_norm": 0.27791842818260193,
      "learning_rate": 4.5055647784181284e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9436235826462507,
      "num_tokens": 109838262.0,
      "step": 14820
    },
    {
      "entropy": 2.0500128395855426,
      "epoch": 0.22189616105052795,
      "grad_norm": 0.31925252079963684,
      "learning_rate": 4.504847649020738e-05,
      "loss": 0.2257,
      "mean_token_accuracy": 0.944381145015359,
      "num_tokens": 109912208.0,
      "step": 14830
    },
    {
      "entropy": 2.054877780377865,
      "epoch": 0.222045787592032,
      "grad_norm": 0.3602094054222107,
      "learning_rate": 4.504130057222998e-05,
      "loss": 0.2278,
      "mean_token_accuracy": 0.9451314132660628,
      "num_tokens": 109984899.0,
      "step": 14840
    },
    {
      "entropy": 2.0549397341907025,
      "epoch": 0.22219541413353608,
      "grad_norm": 0.36160555481910706,
      "learning_rate": 4.50341200319083e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9428565960377455,
      "num_tokens": 110059659.0,
      "step": 14850
    },
    {
      "entropy": 2.01293528303504,
      "epoch": 0.22234504067504013,
      "grad_norm": 0.3723517954349518,
      "learning_rate": 4.502693487090261e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9437457606196403,
      "num_tokens": 110132942.0,
      "step": 14860
    },
    {
      "entropy": 2.040511917322874,
      "epoch": 0.2224946672165442,
      "grad_norm": 0.4029499888420105,
      "learning_rate": 4.5019745090874236e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9431281600147485,
      "num_tokens": 110209573.0,
      "step": 14870
    },
    {
      "entropy": 1.9904178090393543,
      "epoch": 0.22264429375804828,
      "grad_norm": 0.30335453152656555,
      "learning_rate": 4.5012550693485606e-05,
      "loss": 0.2087,
      "mean_token_accuracy": 0.9479824036359787,
      "num_tokens": 110282786.0,
      "step": 14880
    },
    {
      "entropy": 1.9474720314145089,
      "epoch": 0.22279392029955233,
      "grad_norm": 0.3769731819629669,
      "learning_rate": 4.500535168040017e-05,
      "loss": 0.2054,
      "mean_token_accuracy": 0.9461545970290899,
      "num_tokens": 110359163.0,
      "step": 14890
    },
    {
      "entropy": 1.9322262346744536,
      "epoch": 0.2229435468410564,
      "grad_norm": 0.34426093101501465,
      "learning_rate": 4.499814805328247e-05,
      "loss": 0.2149,
      "mean_token_accuracy": 0.9482953980565071,
      "num_tokens": 110435083.0,
      "step": 14900
    },
    {
      "entropy": 1.9813862100243569,
      "epoch": 0.22309317338256046,
      "grad_norm": 0.40497761964797974,
      "learning_rate": 4.499093981379813e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9443288538604975,
      "num_tokens": 110507961.0,
      "step": 14910
    },
    {
      "entropy": 2.014493740350008,
      "epoch": 0.22324279992406454,
      "grad_norm": 0.37459319829940796,
      "learning_rate": 4.49837269636138e-05,
      "loss": 0.2477,
      "mean_token_accuracy": 0.9367315895855427,
      "num_tokens": 110580474.0,
      "step": 14920
    },
    {
      "entropy": 2.00632803812623,
      "epoch": 0.2233924264655686,
      "grad_norm": 0.28861263394355774,
      "learning_rate": 4.497650950439723e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9422844409942627,
      "num_tokens": 110652856.0,
      "step": 14930
    },
    {
      "entropy": 1.997613750398159,
      "epoch": 0.22354205300707267,
      "grad_norm": 0.3124845027923584,
      "learning_rate": 4.496928743781723e-05,
      "loss": 0.2157,
      "mean_token_accuracy": 0.9445130914449692,
      "num_tokens": 110727797.0,
      "step": 14940
    },
    {
      "entropy": 2.027167608588934,
      "epoch": 0.22369167954857672,
      "grad_norm": 0.3353038728237152,
      "learning_rate": 4.496206076554366e-05,
      "loss": 0.2417,
      "mean_token_accuracy": 0.9426096089184284,
      "num_tokens": 110799840.0,
      "step": 14950
    },
    {
      "entropy": 2.0383143827319143,
      "epoch": 0.2238413060900808,
      "grad_norm": 0.3793546259403229,
      "learning_rate": 4.495482948924747e-05,
      "loss": 0.2239,
      "mean_token_accuracy": 0.9443628448992968,
      "num_tokens": 110872758.0,
      "step": 14960
    },
    {
      "entropy": 2.0226249001920222,
      "epoch": 0.22399093263158484,
      "grad_norm": 0.35126665234565735,
      "learning_rate": 4.494759361060064e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9447830632328987,
      "num_tokens": 110947932.0,
      "step": 14970
    },
    {
      "entropy": 2.0217545360326765,
      "epoch": 0.22414055917308892,
      "grad_norm": 0.42849671840667725,
      "learning_rate": 4.494035313127625e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9408001057803631,
      "num_tokens": 111020906.0,
      "step": 14980
    },
    {
      "entropy": 2.0074046552181244,
      "epoch": 0.22429018571459297,
      "grad_norm": 0.32058873772621155,
      "learning_rate": 4.4933108052948416e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9435962595045566,
      "num_tokens": 111092756.0,
      "step": 14990
    },
    {
      "entropy": 2.0074996441602706,
      "epoch": 0.22443981225609705,
      "grad_norm": 0.3518684208393097,
      "learning_rate": 4.492585837729235e-05,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9457818914204836,
      "num_tokens": 111167650.0,
      "step": 15000
    },
    {
      "epoch": 0.22443981225609705,
      "eval_entropy": 2.0174143748521804,
      "eval_loss": 0.24776865541934967,
      "eval_mean_token_accuracy": 0.9446574423193932,
      "eval_num_tokens": 111167650.0,
      "eval_runtime": 546.9813,
      "eval_samples_per_second": 36.564,
      "eval_steps_per_second": 9.141,
      "step": 15000
    },
    {
      "entropy": 2.012942238897085,
      "epoch": 0.2245894387976011,
      "grad_norm": 0.37047305703163147,
      "learning_rate": 4.491860410598428e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9441512532532215,
      "num_tokens": 111240820.0,
      "step": 15010
    },
    {
      "entropy": 2.0317082308232783,
      "epoch": 0.22473906533910518,
      "grad_norm": 0.31004947423934937,
      "learning_rate": 4.491134524070152e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9467091996222734,
      "num_tokens": 111315654.0,
      "step": 15020
    },
    {
      "entropy": 2.0063771970570086,
      "epoch": 0.22488869188060925,
      "grad_norm": 0.3441016972064972,
      "learning_rate": 4.4904081783122486e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9466823801398278,
      "num_tokens": 111389415.0,
      "step": 15030
    },
    {
      "entropy": 2.0139746218919754,
      "epoch": 0.2250383184221133,
      "grad_norm": 0.32765012979507446,
      "learning_rate": 4.489681373492659e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9435013346374035,
      "num_tokens": 111464089.0,
      "step": 15040
    },
    {
      "entropy": 1.9938028737902642,
      "epoch": 0.22518794496361738,
      "grad_norm": 0.27247363328933716,
      "learning_rate": 4.488954109779432e-05,
      "loss": 0.2098,
      "mean_token_accuracy": 0.947207548469305,
      "num_tokens": 111540008.0,
      "step": 15050
    },
    {
      "entropy": 2.0233231499791144,
      "epoch": 0.22533757150512143,
      "grad_norm": 0.43771183490753174,
      "learning_rate": 4.488226387340728e-05,
      "loss": 0.2345,
      "mean_token_accuracy": 0.942552775144577,
      "num_tokens": 111611082.0,
      "step": 15060
    },
    {
      "entropy": 2.0256434842944144,
      "epoch": 0.2254871980466255,
      "grad_norm": 0.3521353304386139,
      "learning_rate": 4.487498206344806e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9430926218628883,
      "num_tokens": 111683690.0,
      "step": 15070
    },
    {
      "entropy": 1.981017790734768,
      "epoch": 0.22563682458812956,
      "grad_norm": 0.2681540250778198,
      "learning_rate": 4.486769566960036e-05,
      "loss": 0.2092,
      "mean_token_accuracy": 0.9438813894987106,
      "num_tokens": 111759958.0,
      "step": 15080
    },
    {
      "entropy": 1.9519768051803112,
      "epoch": 0.22578645112963364,
      "grad_norm": 0.31493079662323,
      "learning_rate": 4.4860404693548916e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9446296352893114,
      "num_tokens": 111835018.0,
      "step": 15090
    },
    {
      "entropy": 1.996109177172184,
      "epoch": 0.2259360776711377,
      "grad_norm": 0.396827757358551,
      "learning_rate": 4.485310913697953e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.942341884970665,
      "num_tokens": 111907119.0,
      "step": 15100
    },
    {
      "entropy": 2.000981708616018,
      "epoch": 0.22608570421264176,
      "grad_norm": 0.3713546097278595,
      "learning_rate": 4.4845809001579074e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9471950579434634,
      "num_tokens": 111980244.0,
      "step": 15110
    },
    {
      "entropy": 2.024167613685131,
      "epoch": 0.2262353307541458,
      "grad_norm": 0.33345866203308105,
      "learning_rate": 4.483850428903547e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9453145645558834,
      "num_tokens": 112051740.0,
      "step": 15120
    },
    {
      "entropy": 2.015564624220133,
      "epoch": 0.2263849572956499,
      "grad_norm": 0.31970494985580444,
      "learning_rate": 4.483119500103768e-05,
      "loss": 0.2064,
      "mean_token_accuracy": 0.945020142570138,
      "num_tokens": 112127905.0,
      "step": 15130
    },
    {
      "entropy": 2.0034637093544005,
      "epoch": 0.22653458383715394,
      "grad_norm": 0.31423550844192505,
      "learning_rate": 4.482388113927576e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9474483519792557,
      "num_tokens": 112201671.0,
      "step": 15140
    },
    {
      "entropy": 1.9893664948642253,
      "epoch": 0.22668421037865802,
      "grad_norm": 0.349835604429245,
      "learning_rate": 4.48165627054408e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9444744683802128,
      "num_tokens": 112276969.0,
      "step": 15150
    },
    {
      "entropy": 2.0122391514480116,
      "epoch": 0.2268338369201621,
      "grad_norm": 0.40789559483528137,
      "learning_rate": 4.480923970122495e-05,
      "loss": 0.2381,
      "mean_token_accuracy": 0.9406100045889616,
      "num_tokens": 112348870.0,
      "step": 15160
    },
    {
      "entropy": 2.0072805628180506,
      "epoch": 0.22698346346166615,
      "grad_norm": 0.3554311990737915,
      "learning_rate": 4.4801912128321414e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9441691931337118,
      "num_tokens": 112422180.0,
      "step": 15170
    },
    {
      "entropy": 2.0480201341211797,
      "epoch": 0.22713309000317022,
      "grad_norm": 0.2843068540096283,
      "learning_rate": 4.4794579988424466e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.9463415015488863,
      "num_tokens": 112496651.0,
      "step": 15180
    },
    {
      "entropy": 2.02738618478179,
      "epoch": 0.22728271654467427,
      "grad_norm": 0.3361741304397583,
      "learning_rate": 4.478724328322943e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9482717584818602,
      "num_tokens": 112571282.0,
      "step": 15190
    },
    {
      "entropy": 2.002488683909178,
      "epoch": 0.22743234308617835,
      "grad_norm": 0.4443276524543762,
      "learning_rate": 4.477990201443266e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9418461881577969,
      "num_tokens": 112645406.0,
      "step": 15200
    },
    {
      "entropy": 2.0331130675971507,
      "epoch": 0.2275819696276824,
      "grad_norm": 0.37163543701171875,
      "learning_rate": 4.477255618373161e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.94352303519845,
      "num_tokens": 112718870.0,
      "step": 15210
    },
    {
      "entropy": 2.0063043043017386,
      "epoch": 0.22773159616918648,
      "grad_norm": 0.3073780834674835,
      "learning_rate": 4.476520579282477e-05,
      "loss": 0.199,
      "mean_token_accuracy": 0.9514647215604782,
      "num_tokens": 112795738.0,
      "step": 15220
    },
    {
      "entropy": 2.012467374652624,
      "epoch": 0.22788122271069053,
      "grad_norm": 0.332012414932251,
      "learning_rate": 4.475785084341166e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9413779832422733,
      "num_tokens": 112869347.0,
      "step": 15230
    },
    {
      "entropy": 2.0126412525773048,
      "epoch": 0.2280308492521946,
      "grad_norm": 0.3271101415157318,
      "learning_rate": 4.475049133719289e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9442832361906767,
      "num_tokens": 112944687.0,
      "step": 15240
    },
    {
      "entropy": 2.007618512958288,
      "epoch": 0.22818047579369866,
      "grad_norm": 0.3432691693305969,
      "learning_rate": 4.4743127275870116e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.948285736888647,
      "num_tokens": 113019129.0,
      "step": 15250
    },
    {
      "entropy": 2.021186326444149,
      "epoch": 0.22833010233520273,
      "grad_norm": 0.43205055594444275,
      "learning_rate": 4.473575866114602e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9421476304531098,
      "num_tokens": 113094468.0,
      "step": 15260
    },
    {
      "entropy": 2.0146180599927903,
      "epoch": 0.22847972887670678,
      "grad_norm": 0.2635376751422882,
      "learning_rate": 4.472838549472437e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9405811082571744,
      "num_tokens": 113166788.0,
      "step": 15270
    },
    {
      "entropy": 1.970995730906725,
      "epoch": 0.22862935541821086,
      "grad_norm": 0.3192756772041321,
      "learning_rate": 4.4721007778309966e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9490535497665405,
      "num_tokens": 113241371.0,
      "step": 15280
    },
    {
      "entropy": 2.0057829186320304,
      "epoch": 0.2287789819597149,
      "grad_norm": 0.34637805819511414,
      "learning_rate": 4.471362551360867e-05,
      "loss": 0.2103,
      "mean_token_accuracy": 0.9471637140959501,
      "num_tokens": 113318042.0,
      "step": 15290
    },
    {
      "entropy": 1.962550550699234,
      "epoch": 0.228928608501219,
      "grad_norm": 0.3373715579509735,
      "learning_rate": 4.4706238702327403e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9427593577653169,
      "num_tokens": 113393855.0,
      "step": 15300
    },
    {
      "entropy": 1.9506757736206055,
      "epoch": 0.22907823504272307,
      "grad_norm": 0.44020876288414,
      "learning_rate": 4.469884734617411e-05,
      "loss": 0.201,
      "mean_token_accuracy": 0.9504586450755597,
      "num_tokens": 113470528.0,
      "step": 15310
    },
    {
      "entropy": 1.9786839924752713,
      "epoch": 0.22922786158422712,
      "grad_norm": 0.45927894115448,
      "learning_rate": 4.469145144685781e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9459199849516153,
      "num_tokens": 113545045.0,
      "step": 15320
    },
    {
      "entropy": 1.985029573738575,
      "epoch": 0.2293774881257312,
      "grad_norm": 0.3427533805370331,
      "learning_rate": 4.4684051006088575e-05,
      "loss": 0.2034,
      "mean_token_accuracy": 0.9476940434426069,
      "num_tokens": 113618152.0,
      "step": 15330
    },
    {
      "entropy": 1.9877336949110032,
      "epoch": 0.22952711466723524,
      "grad_norm": 0.4131075441837311,
      "learning_rate": 4.467664602557751e-05,
      "loss": 0.2383,
      "mean_token_accuracy": 0.9390629705041647,
      "num_tokens": 113690716.0,
      "step": 15340
    },
    {
      "entropy": 1.9985576778650285,
      "epoch": 0.22967674120873932,
      "grad_norm": 0.3161776065826416,
      "learning_rate": 4.4669236507036776e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9456408929079771,
      "num_tokens": 113765518.0,
      "step": 15350
    },
    {
      "entropy": 2.0003026209771635,
      "epoch": 0.22982636775024337,
      "grad_norm": 0.3550056517124176,
      "learning_rate": 4.4661822452179594e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9440422609448433,
      "num_tokens": 113840565.0,
      "step": 15360
    },
    {
      "entropy": 2.025747472792864,
      "epoch": 0.22997599429174745,
      "grad_norm": 0.3500961363315582,
      "learning_rate": 4.465440386272022e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9430900417268276,
      "num_tokens": 113914731.0,
      "step": 15370
    },
    {
      "entropy": 2.0125860802829267,
      "epoch": 0.2301256208332515,
      "grad_norm": 0.32893988490104675,
      "learning_rate": 4.464698074037396e-05,
      "loss": 0.2283,
      "mean_token_accuracy": 0.943072447180748,
      "num_tokens": 113988657.0,
      "step": 15380
    },
    {
      "entropy": 2.012650966644287,
      "epoch": 0.23027524737475558,
      "grad_norm": 0.39278852939605713,
      "learning_rate": 4.463955308685719e-05,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9394141204655171,
      "num_tokens": 114063127.0,
      "step": 15390
    },
    {
      "entropy": 2.0306603901088236,
      "epoch": 0.23042487391625963,
      "grad_norm": 0.32136571407318115,
      "learning_rate": 4.463212090388729e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9420533444732427,
      "num_tokens": 114137488.0,
      "step": 15400
    },
    {
      "entropy": 1.984863667935133,
      "epoch": 0.2305745004577637,
      "grad_norm": 0.30267465114593506,
      "learning_rate": 4.462468419318273e-05,
      "loss": 0.2236,
      "mean_token_accuracy": 0.9452044025063515,
      "num_tokens": 114212397.0,
      "step": 15410
    },
    {
      "entropy": 2.0077911049127577,
      "epoch": 0.23072412699926775,
      "grad_norm": 0.35751578211784363,
      "learning_rate": 4.461724295646301e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9420890953391791,
      "num_tokens": 114283689.0,
      "step": 15420
    },
    {
      "entropy": 1.9952733471989632,
      "epoch": 0.23087375354077183,
      "grad_norm": 0.37443336844444275,
      "learning_rate": 4.460979719544866e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9466187212616205,
      "num_tokens": 114356048.0,
      "step": 15430
    },
    {
      "entropy": 1.9643361032009126,
      "epoch": 0.23102338008227588,
      "grad_norm": 0.3619861304759979,
      "learning_rate": 4.4602346911861296e-05,
      "loss": 0.1879,
      "mean_token_accuracy": 0.9491075027734042,
      "num_tokens": 114432972.0,
      "step": 15440
    },
    {
      "entropy": 1.995201399922371,
      "epoch": 0.23117300662377996,
      "grad_norm": 0.3272418975830078,
      "learning_rate": 4.459489210742354e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9431777350604534,
      "num_tokens": 114504861.0,
      "step": 15450
    },
    {
      "entropy": 2.0347736589610577,
      "epoch": 0.23132263316528404,
      "grad_norm": 0.30950263142585754,
      "learning_rate": 4.458743278385907e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9462115079164505,
      "num_tokens": 114580705.0,
      "step": 15460
    },
    {
      "entropy": 2.0100708961486817,
      "epoch": 0.23147225970678809,
      "grad_norm": 0.3298996388912201,
      "learning_rate": 4.457996894289263e-05,
      "loss": 0.206,
      "mean_token_accuracy": 0.9478289760649204,
      "num_tokens": 114656412.0,
      "step": 15470
    },
    {
      "entropy": 1.989882143586874,
      "epoch": 0.23162188624829216,
      "grad_norm": 0.3745694160461426,
      "learning_rate": 4.457250058624999e-05,
      "loss": 0.2491,
      "mean_token_accuracy": 0.9389148846268653,
      "num_tokens": 114727618.0,
      "step": 15480
    },
    {
      "entropy": 2.004741308093071,
      "epoch": 0.2317715127897962,
      "grad_norm": 0.3305400013923645,
      "learning_rate": 4.456502771565796e-05,
      "loss": 0.2028,
      "mean_token_accuracy": 0.9482883840799332,
      "num_tokens": 114803744.0,
      "step": 15490
    },
    {
      "entropy": 1.985539823770523,
      "epoch": 0.2319211393313003,
      "grad_norm": 0.2776624262332916,
      "learning_rate": 4.455755033284441e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9452421743422746,
      "num_tokens": 114878305.0,
      "step": 15500
    },
    {
      "entropy": 2.0420316599309443,
      "epoch": 0.23207076587280434,
      "grad_norm": 0.3426123857498169,
      "learning_rate": 4.455006843953823e-05,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9434906840324402,
      "num_tokens": 114953541.0,
      "step": 15510
    },
    {
      "entropy": 2.0307010404765604,
      "epoch": 0.23222039241430842,
      "grad_norm": 0.3252795338630676,
      "learning_rate": 4.454258203746938e-05,
      "loss": 0.2313,
      "mean_token_accuracy": 0.9415312763303518,
      "num_tokens": 115027963.0,
      "step": 15520
    },
    {
      "entropy": 2.0189708456397057,
      "epoch": 0.23237001895581247,
      "grad_norm": 0.3352615535259247,
      "learning_rate": 4.453509112836884e-05,
      "loss": 0.24,
      "mean_token_accuracy": 0.9406589701771736,
      "num_tokens": 115100268.0,
      "step": 15530
    },
    {
      "entropy": 2.033125193417072,
      "epoch": 0.23251964549731655,
      "grad_norm": 0.32187867164611816,
      "learning_rate": 4.452759571396864e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9452978666871786,
      "num_tokens": 115175484.0,
      "step": 15540
    },
    {
      "entropy": 2.05086784735322,
      "epoch": 0.2326692720388206,
      "grad_norm": 0.38895705342292786,
      "learning_rate": 4.452009579600187e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9441924814134837,
      "num_tokens": 115248366.0,
      "step": 15550
    },
    {
      "entropy": 2.039254383742809,
      "epoch": 0.23281889858032467,
      "grad_norm": 0.3700357675552368,
      "learning_rate": 4.451259137620263e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.946529433131218,
      "num_tokens": 115323547.0,
      "step": 15560
    },
    {
      "entropy": 2.0355828039348127,
      "epoch": 0.23296852512182872,
      "grad_norm": 0.362968385219574,
      "learning_rate": 4.4505082456306074e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9431021708995104,
      "num_tokens": 115397627.0,
      "step": 15570
    },
    {
      "entropy": 2.057919367402792,
      "epoch": 0.2331181516633328,
      "grad_norm": 0.2936594486236572,
      "learning_rate": 4.4497569038048396e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9424905151128768,
      "num_tokens": 115471366.0,
      "step": 15580
    },
    {
      "entropy": 2.0321060746908186,
      "epoch": 0.23326777820483688,
      "grad_norm": 0.4062644839286804,
      "learning_rate": 4.449005112316685e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9460125006735325,
      "num_tokens": 115545041.0,
      "step": 15590
    },
    {
      "entropy": 2.041969918459654,
      "epoch": 0.23341740474634093,
      "grad_norm": 0.3537786900997162,
      "learning_rate": 4.44825287133997e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9420095339417458,
      "num_tokens": 115618091.0,
      "step": 15600
    },
    {
      "entropy": 2.0278504341840744,
      "epoch": 0.233567031287845,
      "grad_norm": 0.343595415353775,
      "learning_rate": 4.447500181048626e-05,
      "loss": 0.2183,
      "mean_token_accuracy": 0.9424794338643551,
      "num_tokens": 115692572.0,
      "step": 15610
    },
    {
      "entropy": 2.0476527832448483,
      "epoch": 0.23371665782934906,
      "grad_norm": 0.3992612361907959,
      "learning_rate": 4.4467470416166876e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9465295031666756,
      "num_tokens": 115764858.0,
      "step": 15620
    },
    {
      "entropy": 2.0857450872659684,
      "epoch": 0.23386628437085313,
      "grad_norm": 0.33826524019241333,
      "learning_rate": 4.445993453218296e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9450020175427198,
      "num_tokens": 115840642.0,
      "step": 15630
    },
    {
      "entropy": 2.093594118207693,
      "epoch": 0.23401591091235718,
      "grad_norm": 0.3504444360733032,
      "learning_rate": 4.4452394160276914e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9479228254407644,
      "num_tokens": 115915890.0,
      "step": 15640
    },
    {
      "entropy": 2.0980848394334317,
      "epoch": 0.23416553745386126,
      "grad_norm": 0.39364513754844666,
      "learning_rate": 4.444484930219224e-05,
      "loss": 0.2399,
      "mean_token_accuracy": 0.942791672796011,
      "num_tokens": 115988657.0,
      "step": 15650
    },
    {
      "entropy": 2.1216987788677217,
      "epoch": 0.2343151639953653,
      "grad_norm": 0.3813497722148895,
      "learning_rate": 4.443729995967342e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9420614931732416,
      "num_tokens": 116058457.0,
      "step": 15660
    },
    {
      "entropy": 2.09001642242074,
      "epoch": 0.2344647905368694,
      "grad_norm": 0.3150218427181244,
      "learning_rate": 4.4429746134466004e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9393909320235252,
      "num_tokens": 116133822.0,
      "step": 15670
    },
    {
      "entropy": 2.043114899843931,
      "epoch": 0.23461441707837344,
      "grad_norm": 0.30269700288772583,
      "learning_rate": 4.442218782831657e-05,
      "loss": 0.2118,
      "mean_token_accuracy": 0.9457640029489994,
      "num_tokens": 116209675.0,
      "step": 15680
    },
    {
      "entropy": 2.059490706026554,
      "epoch": 0.23476404361987752,
      "grad_norm": 0.2580171227455139,
      "learning_rate": 4.4414625042972734e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9483754806220531,
      "num_tokens": 116283696.0,
      "step": 15690
    },
    {
      "entropy": 2.0796605609357357,
      "epoch": 0.23491367016138157,
      "grad_norm": 0.44307324290275574,
      "learning_rate": 4.440705778018314e-05,
      "loss": 0.2477,
      "mean_token_accuracy": 0.938773638382554,
      "num_tokens": 116355017.0,
      "step": 15700
    },
    {
      "entropy": 2.0547694697976113,
      "epoch": 0.23506329670288564,
      "grad_norm": 0.3357575535774231,
      "learning_rate": 4.439948604169748e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9463300593197346,
      "num_tokens": 116430040.0,
      "step": 15710
    },
    {
      "entropy": 2.1021420657634735,
      "epoch": 0.2352129232443897,
      "grad_norm": 0.3328273594379425,
      "learning_rate": 4.439190982926648e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9411369491368532,
      "num_tokens": 116502028.0,
      "step": 15720
    },
    {
      "entropy": 2.0631238766014577,
      "epoch": 0.23536254978589377,
      "grad_norm": 0.3535475432872772,
      "learning_rate": 4.438432914464188e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9472492534667254,
      "num_tokens": 116575955.0,
      "step": 15730
    },
    {
      "entropy": 2.070156579464674,
      "epoch": 0.23551217632739785,
      "grad_norm": 0.3684746026992798,
      "learning_rate": 4.4376743989576486e-05,
      "loss": 0.2364,
      "mean_token_accuracy": 0.9411379583179951,
      "num_tokens": 116649624.0,
      "step": 15740
    },
    {
      "entropy": 2.037500633299351,
      "epoch": 0.2356618028689019,
      "grad_norm": 0.35305026173591614,
      "learning_rate": 4.4369154365824114e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9428451918065548,
      "num_tokens": 116723254.0,
      "step": 15750
    },
    {
      "entropy": 2.047383426874876,
      "epoch": 0.23581142941040598,
      "grad_norm": 0.37458449602127075,
      "learning_rate": 4.436156027513961e-05,
      "loss": 0.2061,
      "mean_token_accuracy": 0.9489231254905462,
      "num_tokens": 116797033.0,
      "step": 15760
    },
    {
      "entropy": 2.0384680651128293,
      "epoch": 0.23596105595191003,
      "grad_norm": 0.33996808528900146,
      "learning_rate": 4.435396171927887e-05,
      "loss": 0.2025,
      "mean_token_accuracy": 0.9509435724467039,
      "num_tokens": 116869641.0,
      "step": 15770
    },
    {
      "entropy": 2.0476046822965146,
      "epoch": 0.2361106824934141,
      "grad_norm": 0.33860573172569275,
      "learning_rate": 4.434635869999882e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9495259393006563,
      "num_tokens": 116943636.0,
      "step": 15780
    },
    {
      "entropy": 2.0439771845936776,
      "epoch": 0.23626030903491815,
      "grad_norm": 0.33408668637275696,
      "learning_rate": 4.4338751219057414e-05,
      "loss": 0.2323,
      "mean_token_accuracy": 0.9447633545845747,
      "num_tokens": 117016617.0,
      "step": 15790
    },
    {
      "entropy": 2.0760328866541387,
      "epoch": 0.23640993557642223,
      "grad_norm": 0.3129129707813263,
      "learning_rate": 4.433113927821362e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.9417223058640957,
      "num_tokens": 117091042.0,
      "step": 15800
    },
    {
      "entropy": 2.0507302597165107,
      "epoch": 0.23655956211792628,
      "grad_norm": 0.38783541321754456,
      "learning_rate": 4.432352287922748e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9432470206171274,
      "num_tokens": 117166568.0,
      "step": 15810
    },
    {
      "entropy": 2.0711214542388916,
      "epoch": 0.23670918865943036,
      "grad_norm": 0.3301248252391815,
      "learning_rate": 4.431590202386002e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9427740078419447,
      "num_tokens": 117243022.0,
      "step": 15820
    },
    {
      "entropy": 2.0793760634958742,
      "epoch": 0.2368588152009344,
      "grad_norm": 0.3130905032157898,
      "learning_rate": 4.4308276713873335e-05,
      "loss": 0.2291,
      "mean_token_accuracy": 0.9445065405219794,
      "num_tokens": 117316059.0,
      "step": 15830
    },
    {
      "entropy": 2.090775099396706,
      "epoch": 0.23700844174243849,
      "grad_norm": 0.3165552020072937,
      "learning_rate": 4.430064695103052e-05,
      "loss": 0.2159,
      "mean_token_accuracy": 0.9448061212897301,
      "num_tokens": 117389091.0,
      "step": 15840
    },
    {
      "entropy": 2.0690905950963496,
      "epoch": 0.23715806828394254,
      "grad_norm": 0.3243985176086426,
      "learning_rate": 4.4293012737095705e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9462750602513552,
      "num_tokens": 117460937.0,
      "step": 15850
    },
    {
      "entropy": 2.053806984424591,
      "epoch": 0.2373076948254466,
      "grad_norm": 0.36149829626083374,
      "learning_rate": 4.4285374073834077e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9425987333059311,
      "num_tokens": 117531851.0,
      "step": 15860
    },
    {
      "entropy": 2.0383010409772395,
      "epoch": 0.23745732136695066,
      "grad_norm": 0.32097554206848145,
      "learning_rate": 4.42777309630118e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9430879011750222,
      "num_tokens": 117609917.0,
      "step": 15870
    },
    {
      "entropy": 2.0740928016602993,
      "epoch": 0.23760694790845474,
      "grad_norm": 0.3941027820110321,
      "learning_rate": 4.427008340639614e-05,
      "loss": 0.2149,
      "mean_token_accuracy": 0.9488260477781296,
      "num_tokens": 117685225.0,
      "step": 15880
    },
    {
      "entropy": 2.073116146028042,
      "epoch": 0.23775657444995882,
      "grad_norm": 0.3424692451953888,
      "learning_rate": 4.4262431405755314e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9461569461971522,
      "num_tokens": 117759559.0,
      "step": 15890
    },
    {
      "entropy": 2.0798732958734036,
      "epoch": 0.23790620099146287,
      "grad_norm": 0.42008957266807556,
      "learning_rate": 4.425477496285861e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.9454585824161768,
      "num_tokens": 117832399.0,
      "step": 15900
    },
    {
      "entropy": 2.0575854554772377,
      "epoch": 0.23805582753296695,
      "grad_norm": 0.35241132974624634,
      "learning_rate": 4.424711407947633e-05,
      "loss": 0.2311,
      "mean_token_accuracy": 0.9424953121691942,
      "num_tokens": 117906704.0,
      "step": 15910
    },
    {
      "entropy": 2.0935396663844585,
      "epoch": 0.238205454074471,
      "grad_norm": 0.40505465865135193,
      "learning_rate": 4.423944875737982e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9435701709240675,
      "num_tokens": 117977995.0,
      "step": 15920
    },
    {
      "entropy": 2.1041508346796034,
      "epoch": 0.23835508061597507,
      "grad_norm": 0.27820131182670593,
      "learning_rate": 4.423177899834143e-05,
      "loss": 0.2457,
      "mean_token_accuracy": 0.940861701965332,
      "num_tokens": 118050373.0,
      "step": 15930
    },
    {
      "entropy": 2.068365006893873,
      "epoch": 0.23850470715747912,
      "grad_norm": 0.302327960729599,
      "learning_rate": 4.422410480413454e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9455043748021126,
      "num_tokens": 118126341.0,
      "step": 15940
    },
    {
      "entropy": 2.0691476710140706,
      "epoch": 0.2386543336989832,
      "grad_norm": 0.31645458936691284,
      "learning_rate": 4.421642617653356e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9467107065021991,
      "num_tokens": 118204674.0,
      "step": 15950
    },
    {
      "entropy": 2.087872290611267,
      "epoch": 0.23880396024048725,
      "grad_norm": 0.3942251205444336,
      "learning_rate": 4.420874311731394e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.9445725582540035,
      "num_tokens": 118275557.0,
      "step": 15960
    },
    {
      "entropy": 2.0585035145282746,
      "epoch": 0.23895358678199133,
      "grad_norm": 0.3205868899822235,
      "learning_rate": 4.420105562825213e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9473403446376324,
      "num_tokens": 118352338.0,
      "step": 15970
    },
    {
      "entropy": 2.0751403607428074,
      "epoch": 0.23910321332349538,
      "grad_norm": 0.397442102432251,
      "learning_rate": 4.419336371112561e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9462721731513739,
      "num_tokens": 118429252.0,
      "step": 15980
    },
    {
      "entropy": 2.1158520355820656,
      "epoch": 0.23925283986499946,
      "grad_norm": 0.335366427898407,
      "learning_rate": 4.4185667367712895e-05,
      "loss": 0.2326,
      "mean_token_accuracy": 0.94274448081851,
      "num_tokens": 118504478.0,
      "step": 15990
    },
    {
      "entropy": 2.1069047853350638,
      "epoch": 0.2394024664065035,
      "grad_norm": 0.4520975947380066,
      "learning_rate": 4.417796659979352e-05,
      "loss": 0.2441,
      "mean_token_accuracy": 0.9418787810951471,
      "num_tokens": 118577149.0,
      "step": 16000
    },
    {
      "epoch": 0.2394024664065035,
      "eval_entropy": 2.098721252131462,
      "eval_loss": 0.24755480885505676,
      "eval_mean_token_accuracy": 0.944678509402275,
      "eval_num_tokens": 118577149.0,
      "eval_runtime": 543.513,
      "eval_samples_per_second": 36.798,
      "eval_steps_per_second": 9.199,
      "step": 16000
    },
    {
      "entropy": 2.0848411075770854,
      "epoch": 0.23955209294800758,
      "grad_norm": 0.4328034818172455,
      "learning_rate": 4.417026140914802e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9453351046890021,
      "num_tokens": 118649926.0,
      "step": 16010
    },
    {
      "entropy": 2.0452876202762127,
      "epoch": 0.23970171948951163,
      "grad_norm": 0.352972149848938,
      "learning_rate": 4.4162551797558e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9448449041694402,
      "num_tokens": 118724618.0,
      "step": 16020
    },
    {
      "entropy": 2.069345071911812,
      "epoch": 0.2398513460310157,
      "grad_norm": 0.3966101408004761,
      "learning_rate": 4.415483776680604e-05,
      "loss": 0.2499,
      "mean_token_accuracy": 0.9425683658570051,
      "num_tokens": 118796079.0,
      "step": 16030
    },
    {
      "entropy": 2.0619590900838376,
      "epoch": 0.2400009725725198,
      "grad_norm": 0.36089709401130676,
      "learning_rate": 4.414711931867576e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9465872455388308,
      "num_tokens": 118869909.0,
      "step": 16040
    },
    {
      "entropy": 2.0692064471542837,
      "epoch": 0.24015059911402384,
      "grad_norm": 0.2963952124118805,
      "learning_rate": 4.413939645495182e-05,
      "loss": 0.2104,
      "mean_token_accuracy": 0.9475876275449991,
      "num_tokens": 118945988.0,
      "step": 16050
    },
    {
      "entropy": 2.058154482394457,
      "epoch": 0.24030022565552792,
      "grad_norm": 0.30522504448890686,
      "learning_rate": 4.4131669177419865e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9439670585095883,
      "num_tokens": 119021646.0,
      "step": 16060
    },
    {
      "entropy": 2.0585036404430865,
      "epoch": 0.24044985219703197,
      "grad_norm": 0.3717178404331207,
      "learning_rate": 4.412393748786658e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9466731593012809,
      "num_tokens": 119093720.0,
      "step": 16070
    },
    {
      "entropy": 2.0416114769876,
      "epoch": 0.24059947873853604,
      "grad_norm": 0.27685752511024475,
      "learning_rate": 4.4116201388079684e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.941658328846097,
      "num_tokens": 119168171.0,
      "step": 16080
    },
    {
      "entropy": 2.0600788913667203,
      "epoch": 0.2407491052800401,
      "grad_norm": 0.31977570056915283,
      "learning_rate": 4.410846087984789e-05,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9483227115124464,
      "num_tokens": 119241908.0,
      "step": 16090
    },
    {
      "entropy": 2.0905535601079466,
      "epoch": 0.24089873182154417,
      "grad_norm": 0.3567586839199066,
      "learning_rate": 4.4100715964960944e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9422936372458934,
      "num_tokens": 119315026.0,
      "step": 16100
    },
    {
      "entropy": 2.065988294035196,
      "epoch": 0.24104835836304822,
      "grad_norm": 0.32850468158721924,
      "learning_rate": 4.409296664520961e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9438053708523512,
      "num_tokens": 119389410.0,
      "step": 16110
    },
    {
      "entropy": 2.0821424767374994,
      "epoch": 0.2411979849045523,
      "grad_norm": 0.42206600308418274,
      "learning_rate": 4.408521292238566e-05,
      "loss": 0.2098,
      "mean_token_accuracy": 0.9485210493206978,
      "num_tokens": 119464709.0,
      "step": 16120
    },
    {
      "entropy": 2.0478326573967935,
      "epoch": 0.24134761144605635,
      "grad_norm": 0.42451903223991394,
      "learning_rate": 4.4077454798281905e-05,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9468190506100654,
      "num_tokens": 119537451.0,
      "step": 16130
    },
    {
      "entropy": 2.074319223314524,
      "epoch": 0.24149723798756043,
      "grad_norm": 0.4473297894001007,
      "learning_rate": 4.4069692274692155e-05,
      "loss": 0.243,
      "mean_token_accuracy": 0.9423326823860407,
      "num_tokens": 119608619.0,
      "step": 16140
    },
    {
      "entropy": 2.0335396267473698,
      "epoch": 0.24164686452906448,
      "grad_norm": 0.33964160084724426,
      "learning_rate": 4.406192535341124e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9463571567088366,
      "num_tokens": 119684314.0,
      "step": 16150
    },
    {
      "entropy": 2.03813733831048,
      "epoch": 0.24179649107056855,
      "grad_norm": 0.2849193215370178,
      "learning_rate": 4.4054154036235016e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9464765787124634,
      "num_tokens": 119759183.0,
      "step": 16160
    },
    {
      "entropy": 2.0337743066251277,
      "epoch": 0.24194611761207263,
      "grad_norm": 0.33442673087120056,
      "learning_rate": 4.404637832496034e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9459697607904672,
      "num_tokens": 119833426.0,
      "step": 16170
    },
    {
      "entropy": 2.055902273207903,
      "epoch": 0.24209574415357668,
      "grad_norm": 0.3602216839790344,
      "learning_rate": 4.4038598221385115e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9417982868850231,
      "num_tokens": 119906636.0,
      "step": 16180
    },
    {
      "entropy": 2.07627956867218,
      "epoch": 0.24224537069508076,
      "grad_norm": 0.443600594997406,
      "learning_rate": 4.403081372730821e-05,
      "loss": 0.2352,
      "mean_token_accuracy": 0.9413743548095226,
      "num_tokens": 119982633.0,
      "step": 16190
    },
    {
      "entropy": 2.041438663750887,
      "epoch": 0.2423949972365848,
      "grad_norm": 0.29614707827568054,
      "learning_rate": 4.402302484452957e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9437682565301657,
      "num_tokens": 120057007.0,
      "step": 16200
    },
    {
      "entropy": 2.0192115172743796,
      "epoch": 0.24254462377808889,
      "grad_norm": 0.3380019962787628,
      "learning_rate": 4.40152315748501e-05,
      "loss": 0.2072,
      "mean_token_accuracy": 0.9471330963075161,
      "num_tokens": 120129796.0,
      "step": 16210
    },
    {
      "entropy": 2.016851223260164,
      "epoch": 0.24269425031959294,
      "grad_norm": 0.3314051926136017,
      "learning_rate": 4.400743392007176e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.9428550433367491,
      "num_tokens": 120200554.0,
      "step": 16220
    },
    {
      "entropy": 1.9930514819920062,
      "epoch": 0.242843876861097,
      "grad_norm": 0.3363197147846222,
      "learning_rate": 4.399963188199748e-05,
      "loss": 0.2112,
      "mean_token_accuracy": 0.9477915689349174,
      "num_tokens": 120273992.0,
      "step": 16230
    },
    {
      "entropy": 2.0037550680339336,
      "epoch": 0.24299350340260106,
      "grad_norm": 0.3375464379787445,
      "learning_rate": 4.399182546243127e-05,
      "loss": 0.2239,
      "mean_token_accuracy": 0.9422389108687639,
      "num_tokens": 120348151.0,
      "step": 16240
    },
    {
      "entropy": 1.9866594195365905,
      "epoch": 0.24314312994410514,
      "grad_norm": 0.3421008586883545,
      "learning_rate": 4.398401466317808e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9468516822904348,
      "num_tokens": 120424637.0,
      "step": 16250
    },
    {
      "entropy": 2.0045946575701237,
      "epoch": 0.2432927564856092,
      "grad_norm": 0.39425644278526306,
      "learning_rate": 4.397619948604391e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9434769116342068,
      "num_tokens": 120497875.0,
      "step": 16260
    },
    {
      "entropy": 2.0337354809045793,
      "epoch": 0.24344238302711327,
      "grad_norm": 0.40602436661720276,
      "learning_rate": 4.3968379932835786e-05,
      "loss": 0.2558,
      "mean_token_accuracy": 0.9360252488404512,
      "num_tokens": 120568843.0,
      "step": 16270
    },
    {
      "entropy": 1.9917780749499798,
      "epoch": 0.24359200956861732,
      "grad_norm": 0.32842880487442017,
      "learning_rate": 4.396055600536171e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9452771190553904,
      "num_tokens": 120644025.0,
      "step": 16280
    },
    {
      "entropy": 1.9718476995825767,
      "epoch": 0.2437416361101214,
      "grad_norm": 0.3771713078022003,
      "learning_rate": 4.395272770543072e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9456279583275318,
      "num_tokens": 120718577.0,
      "step": 16290
    },
    {
      "entropy": 2.002423678338528,
      "epoch": 0.24389126265162545,
      "grad_norm": 0.39539510011672974,
      "learning_rate": 4.394489503485285e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.9407203998416662,
      "num_tokens": 120793611.0,
      "step": 16300
    },
    {
      "entropy": 2.012669153511524,
      "epoch": 0.24404088919312952,
      "grad_norm": 0.3061200976371765,
      "learning_rate": 4.3937057995439176e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.9413558680564165,
      "num_tokens": 120869946.0,
      "step": 16310
    },
    {
      "entropy": 2.052753484249115,
      "epoch": 0.2441905157346336,
      "grad_norm": 0.3230930268764496,
      "learning_rate": 4.392921658900174e-05,
      "loss": 0.2504,
      "mean_token_accuracy": 0.9360945120453834,
      "num_tokens": 120941861.0,
      "step": 16320
    },
    {
      "entropy": 2.026337506622076,
      "epoch": 0.24434014227613765,
      "grad_norm": 0.39909130334854126,
      "learning_rate": 4.3921370817353605e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9437697723507881,
      "num_tokens": 121014794.0,
      "step": 16330
    },
    {
      "entropy": 2.0071354769170284,
      "epoch": 0.24448976881764173,
      "grad_norm": 0.3305225968360901,
      "learning_rate": 4.391352068230888e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9416888732463121,
      "num_tokens": 121091143.0,
      "step": 16340
    },
    {
      "entropy": 2.036068270355463,
      "epoch": 0.24463939535914578,
      "grad_norm": 0.3151331841945648,
      "learning_rate": 4.390566618568264e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9463795971125364,
      "num_tokens": 121165791.0,
      "step": 16350
    },
    {
      "entropy": 2.0497440725564955,
      "epoch": 0.24478902190064986,
      "grad_norm": 0.3441990315914154,
      "learning_rate": 4.3897807329290995e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9438645161688328,
      "num_tokens": 121240097.0,
      "step": 16360
    },
    {
      "entropy": 2.0152134865522386,
      "epoch": 0.2449386484421539,
      "grad_norm": 0.3641666769981384,
      "learning_rate": 4.388994411495104e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.944090685993433,
      "num_tokens": 121313149.0,
      "step": 16370
    },
    {
      "entropy": 2.018006507307291,
      "epoch": 0.24508827498365798,
      "grad_norm": 0.2816300690174103,
      "learning_rate": 4.3882076544480895e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9424852773547172,
      "num_tokens": 121386154.0,
      "step": 16380
    },
    {
      "entropy": 2.019013747572899,
      "epoch": 0.24523790152516203,
      "grad_norm": 0.3590509295463562,
      "learning_rate": 4.3874204619699684e-05,
      "loss": 0.2535,
      "mean_token_accuracy": 0.9339492313563824,
      "num_tokens": 121459193.0,
      "step": 16390
    },
    {
      "entropy": 1.9696059748530388,
      "epoch": 0.2453875280666661,
      "grad_norm": 0.3834046423435211,
      "learning_rate": 4.386632834242753e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9442971590906382,
      "num_tokens": 121531655.0,
      "step": 16400
    },
    {
      "entropy": 1.96181361079216,
      "epoch": 0.24553715460817016,
      "grad_norm": 0.3191194534301758,
      "learning_rate": 4.3858447714485574e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9467709362506866,
      "num_tokens": 121605420.0,
      "step": 16410
    },
    {
      "entropy": 1.9572921447455882,
      "epoch": 0.24568678114967424,
      "grad_norm": 0.39426347613334656,
      "learning_rate": 4.385056273769596e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9477232247591019,
      "num_tokens": 121681949.0,
      "step": 16420
    },
    {
      "entropy": 2.0002415224909784,
      "epoch": 0.2458364076911783,
      "grad_norm": 0.2768111824989319,
      "learning_rate": 4.384267341388183e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9430888962000609,
      "num_tokens": 121755880.0,
      "step": 16430
    },
    {
      "entropy": 2.0337524503469466,
      "epoch": 0.24598603423268237,
      "grad_norm": 0.36334654688835144,
      "learning_rate": 4.3834779744867335e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9378060389310121,
      "num_tokens": 121828391.0,
      "step": 16440
    },
    {
      "entropy": 2.0571872130036355,
      "epoch": 0.24613566077418642,
      "grad_norm": 0.3307816684246063,
      "learning_rate": 4.382688173247763e-05,
      "loss": 0.2336,
      "mean_token_accuracy": 0.9402903161942959,
      "num_tokens": 121901243.0,
      "step": 16450
    },
    {
      "entropy": 2.053033521771431,
      "epoch": 0.2462852873156905,
      "grad_norm": 0.3435797393321991,
      "learning_rate": 4.381897937853889e-05,
      "loss": 0.2238,
      "mean_token_accuracy": 0.943028011918068,
      "num_tokens": 121975299.0,
      "step": 16460
    },
    {
      "entropy": 2.0084967061877252,
      "epoch": 0.24643491385719457,
      "grad_norm": 0.3818540573120117,
      "learning_rate": 4.381107268487827e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9409181106835603,
      "num_tokens": 122049289.0,
      "step": 16470
    },
    {
      "entropy": 2.047669021040201,
      "epoch": 0.24658454039869862,
      "grad_norm": 0.32403117418289185,
      "learning_rate": 4.3803161653323926e-05,
      "loss": 0.2548,
      "mean_token_accuracy": 0.9364870361983776,
      "num_tokens": 122120549.0,
      "step": 16480
    },
    {
      "entropy": 1.9855361193418504,
      "epoch": 0.2467341669402027,
      "grad_norm": 0.30268576741218567,
      "learning_rate": 4.3795246285705056e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.9451680358499288,
      "num_tokens": 122195401.0,
      "step": 16490
    },
    {
      "entropy": 1.9707822889089583,
      "epoch": 0.24688379348170675,
      "grad_norm": 0.3267269432544708,
      "learning_rate": 4.3787326583851806e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9465970400720835,
      "num_tokens": 122268489.0,
      "step": 16500
    },
    {
      "entropy": 2.0079424642026424,
      "epoch": 0.24703342002321083,
      "grad_norm": 0.3960433006286621,
      "learning_rate": 4.377940254959537e-05,
      "loss": 0.2265,
      "mean_token_accuracy": 0.942170899733901,
      "num_tokens": 122340322.0,
      "step": 16510
    },
    {
      "entropy": 2.048304967582226,
      "epoch": 0.24718304656471488,
      "grad_norm": 0.3772333562374115,
      "learning_rate": 4.377147418476791e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9434278637170792,
      "num_tokens": 122416783.0,
      "step": 16520
    },
    {
      "entropy": 2.017482512444258,
      "epoch": 0.24733267310621895,
      "grad_norm": 0.3865550458431244,
      "learning_rate": 4.376354149120263e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9457809492945671,
      "num_tokens": 122491568.0,
      "step": 16530
    },
    {
      "entropy": 1.9773899637162686,
      "epoch": 0.247482299647723,
      "grad_norm": 0.39673933386802673,
      "learning_rate": 4.375560447073369e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9469282567501068,
      "num_tokens": 122564461.0,
      "step": 16540
    },
    {
      "entropy": 1.9578679554164409,
      "epoch": 0.24763192618922708,
      "grad_norm": 0.3386728763580322,
      "learning_rate": 4.3747663125196274e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9476192023605108,
      "num_tokens": 122638600.0,
      "step": 16550
    },
    {
      "entropy": 1.990031461417675,
      "epoch": 0.24778155273073113,
      "grad_norm": 0.3364439904689789,
      "learning_rate": 4.3739717456426574e-05,
      "loss": 0.2228,
      "mean_token_accuracy": 0.94636568762362,
      "num_tokens": 122714984.0,
      "step": 16560
    },
    {
      "entropy": 2.0045017920434476,
      "epoch": 0.2479311792722352,
      "grad_norm": 0.3133964538574219,
      "learning_rate": 4.3731767466261744e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9434027221053839,
      "num_tokens": 122788247.0,
      "step": 16570
    },
    {
      "entropy": 1.9610675513744353,
      "epoch": 0.24808080581373926,
      "grad_norm": 0.3368903696537018,
      "learning_rate": 4.372381315653998e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9473176315426827,
      "num_tokens": 122862257.0,
      "step": 16580
    },
    {
      "entropy": 1.9492383748292923,
      "epoch": 0.24823043235524334,
      "grad_norm": 0.3312985897064209,
      "learning_rate": 4.371585452910047e-05,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9479431383311748,
      "num_tokens": 122938846.0,
      "step": 16590
    },
    {
      "entropy": 1.971950925141573,
      "epoch": 0.2483800588967474,
      "grad_norm": 0.3543524146080017,
      "learning_rate": 4.370789158578337e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9431917991489172,
      "num_tokens": 123011438.0,
      "step": 16600
    },
    {
      "entropy": 1.9573254011571408,
      "epoch": 0.24852968543825146,
      "grad_norm": 0.3092285096645355,
      "learning_rate": 4.3699924328429873e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9475264042615891,
      "num_tokens": 123087341.0,
      "step": 16610
    },
    {
      "entropy": 2.0000997729599477,
      "epoch": 0.24867931197975554,
      "grad_norm": 0.33425843715667725,
      "learning_rate": 4.369195275888214e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9441487450152636,
      "num_tokens": 123162687.0,
      "step": 16620
    },
    {
      "entropy": 2.0022789508104326,
      "epoch": 0.2488289385212596,
      "grad_norm": 0.3099813163280487,
      "learning_rate": 4.368397687898333e-05,
      "loss": 0.2075,
      "mean_token_accuracy": 0.9505322974175214,
      "num_tokens": 123238702.0,
      "step": 16630
    },
    {
      "entropy": 2.0081651136279106,
      "epoch": 0.24897856506276367,
      "grad_norm": 0.31102094054222107,
      "learning_rate": 4.367599669057762e-05,
      "loss": 0.2306,
      "mean_token_accuracy": 0.9449191451072693,
      "num_tokens": 123310061.0,
      "step": 16640
    },
    {
      "entropy": 2.0100230783224107,
      "epoch": 0.24912819160426772,
      "grad_norm": 0.32776451110839844,
      "learning_rate": 4.3668012195510175e-05,
      "loss": 0.21,
      "mean_token_accuracy": 0.9471614200621843,
      "num_tokens": 123386702.0,
      "step": 16650
    },
    {
      "entropy": 2.0042002387344837,
      "epoch": 0.2492778181457718,
      "grad_norm": 0.38518717885017395,
      "learning_rate": 4.366002339562715e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.94392872415483,
      "num_tokens": 123463652.0,
      "step": 16660
    },
    {
      "entropy": 1.975849761813879,
      "epoch": 0.24942744468727585,
      "grad_norm": 0.3566230535507202,
      "learning_rate": 4.3652030292775695e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9478672806173563,
      "num_tokens": 123538573.0,
      "step": 16670
    },
    {
      "entropy": 1.989323589205742,
      "epoch": 0.24957707122877992,
      "grad_norm": 0.3418055772781372,
      "learning_rate": 4.3644032888803955e-05,
      "loss": 0.2317,
      "mean_token_accuracy": 0.9450258005410432,
      "num_tokens": 123611256.0,
      "step": 16680
    },
    {
      "entropy": 2.035205653309822,
      "epoch": 0.24972669777028397,
      "grad_norm": 0.363689661026001,
      "learning_rate": 4.363603118556107e-05,
      "loss": 0.2218,
      "mean_token_accuracy": 0.9423216570168733,
      "num_tokens": 123683517.0,
      "step": 16690
    },
    {
      "entropy": 2.039001078903675,
      "epoch": 0.24987632431178805,
      "grad_norm": 0.38119035959243774,
      "learning_rate": 4.3628025184897194e-05,
      "loss": 0.2435,
      "mean_token_accuracy": 0.9373656831681728,
      "num_tokens": 123756380.0,
      "step": 16700
    },
    {
      "entropy": 2.0349832609295846,
      "epoch": 0.2500259508532921,
      "grad_norm": 0.2823590636253357,
      "learning_rate": 4.362001488866343e-05,
      "loss": 0.2229,
      "mean_token_accuracy": 0.941704834997654,
      "num_tokens": 123829670.0,
      "step": 16710
    },
    {
      "entropy": 2.0058689691126346,
      "epoch": 0.2501755773947962,
      "grad_norm": 0.2992836833000183,
      "learning_rate": 4.361200029871192e-05,
      "loss": 0.2142,
      "mean_token_accuracy": 0.9481477294117212,
      "num_tokens": 123904601.0,
      "step": 16720
    },
    {
      "entropy": 2.0063904598355293,
      "epoch": 0.25032520393630026,
      "grad_norm": 0.41628551483154297,
      "learning_rate": 4.360398141689577e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9430948473513127,
      "num_tokens": 123977071.0,
      "step": 16730
    },
    {
      "entropy": 2.020273266732693,
      "epoch": 0.2504748304778043,
      "grad_norm": 0.3421095609664917,
      "learning_rate": 4.35959582450691e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9441402390599251,
      "num_tokens": 124049079.0,
      "step": 16740
    },
    {
      "entropy": 2.043296295404434,
      "epoch": 0.25062445701930836,
      "grad_norm": 0.3016813099384308,
      "learning_rate": 4.3587930785087e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9424254439771176,
      "num_tokens": 124123113.0,
      "step": 16750
    },
    {
      "entropy": 2.0543952472507954,
      "epoch": 0.25077408356081243,
      "grad_norm": 0.3363136947154999,
      "learning_rate": 4.3579899038805564e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9461530242115259,
      "num_tokens": 124196196.0,
      "step": 16760
    },
    {
      "entropy": 2.0375195115804674,
      "epoch": 0.2509237101023165,
      "grad_norm": 0.3310266137123108,
      "learning_rate": 4.357186300808188e-05,
      "loss": 0.2039,
      "mean_token_accuracy": 0.9482848644256592,
      "num_tokens": 124269633.0,
      "step": 16770
    },
    {
      "entropy": 1.9885814964771271,
      "epoch": 0.2510733366438206,
      "grad_norm": 0.294413298368454,
      "learning_rate": 4.3563822694774e-05,
      "loss": 0.2058,
      "mean_token_accuracy": 0.9478947155177593,
      "num_tokens": 124347201.0,
      "step": 16780
    },
    {
      "entropy": 2.026625461131334,
      "epoch": 0.2512229631853246,
      "grad_norm": 0.3412422239780426,
      "learning_rate": 4.3555778100741024e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.9442304644733668,
      "num_tokens": 124419940.0,
      "step": 16790
    },
    {
      "entropy": 2.031872060149908,
      "epoch": 0.2513725897268287,
      "grad_norm": 0.34379133582115173,
      "learning_rate": 4.354772922784297e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9434406265616417,
      "num_tokens": 124496337.0,
      "step": 16800
    },
    {
      "entropy": 2.0488209024071695,
      "epoch": 0.25152221626833277,
      "grad_norm": 0.411844402551651,
      "learning_rate": 4.35396760779409e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9420457955449819,
      "num_tokens": 124570583.0,
      "step": 16810
    },
    {
      "entropy": 2.015605629980564,
      "epoch": 0.25167184280983684,
      "grad_norm": 0.3768039047718048,
      "learning_rate": 4.3531618652896836e-05,
      "loss": 0.2076,
      "mean_token_accuracy": 0.9490985263139009,
      "num_tokens": 124644753.0,
      "step": 16820
    },
    {
      "entropy": 2.0118858821690084,
      "epoch": 0.25182146935134087,
      "grad_norm": 0.4580540060997009,
      "learning_rate": 4.3523556954573806e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9407936118543148,
      "num_tokens": 124719139.0,
      "step": 16830
    },
    {
      "entropy": 2.0249043092131616,
      "epoch": 0.25197109589284494,
      "grad_norm": 0.32576364278793335,
      "learning_rate": 4.351549098483582e-05,
      "loss": 0.2307,
      "mean_token_accuracy": 0.9403771858662366,
      "num_tokens": 124791060.0,
      "step": 16840
    },
    {
      "entropy": 2.0139253407716753,
      "epoch": 0.252120722434349,
      "grad_norm": 0.5852076411247253,
      "learning_rate": 4.350742074554787e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9467729043215514,
      "num_tokens": 124865922.0,
      "step": 16850
    },
    {
      "entropy": 2.0327835284173488,
      "epoch": 0.2522703489758531,
      "grad_norm": 0.34613946080207825,
      "learning_rate": 4.349934623857592e-05,
      "loss": 0.2222,
      "mean_token_accuracy": 0.9439053174108267,
      "num_tokens": 124938374.0,
      "step": 16860
    },
    {
      "entropy": 2.016629195213318,
      "epoch": 0.2524199755173571,
      "grad_norm": 0.4179118573665619,
      "learning_rate": 4.349126746578697e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9457638803869486,
      "num_tokens": 125012719.0,
      "step": 16870
    },
    {
      "entropy": 2.0316185727715492,
      "epoch": 0.2525696020588612,
      "grad_norm": 0.3678561747074127,
      "learning_rate": 4.348318442904897e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.947494312748313,
      "num_tokens": 125086126.0,
      "step": 16880
    },
    {
      "entropy": 2.0064745537936686,
      "epoch": 0.2527192286003653,
      "grad_norm": 0.28348076343536377,
      "learning_rate": 4.347509713023083e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9481040451675653,
      "num_tokens": 125162906.0,
      "step": 16890
    },
    {
      "entropy": 2.0510476611554624,
      "epoch": 0.25286885514186935,
      "grad_norm": 0.4033825993537903,
      "learning_rate": 4.346700557120253e-05,
      "loss": 0.2374,
      "mean_token_accuracy": 0.9422494497150182,
      "num_tokens": 125239022.0,
      "step": 16900
    },
    {
      "entropy": 2.0089678950607777,
      "epoch": 0.25301848168337343,
      "grad_norm": 0.34085607528686523,
      "learning_rate": 4.345890975383494e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9421025216579437,
      "num_tokens": 125314501.0,
      "step": 16910
    },
    {
      "entropy": 2.01661573946476,
      "epoch": 0.25316810822487745,
      "grad_norm": 0.35542815923690796,
      "learning_rate": 4.345080967999997e-05,
      "loss": 0.2353,
      "mean_token_accuracy": 0.9416692964732647,
      "num_tokens": 125385794.0,
      "step": 16920
    },
    {
      "entropy": 2.052581109851599,
      "epoch": 0.25331773476638153,
      "grad_norm": 0.3307231366634369,
      "learning_rate": 4.344270535157051e-05,
      "loss": 0.2482,
      "mean_token_accuracy": 0.9417580153793097,
      "num_tokens": 125459401.0,
      "step": 16930
    },
    {
      "entropy": 2.026107481867075,
      "epoch": 0.2534673613078856,
      "grad_norm": 0.42462819814682007,
      "learning_rate": 4.343459677042041e-05,
      "loss": 0.2041,
      "mean_token_accuracy": 0.9483846720308066,
      "num_tokens": 125535691.0,
      "step": 16940
    },
    {
      "entropy": 2.0067538641393186,
      "epoch": 0.2536169878493897,
      "grad_norm": 0.39572760462760925,
      "learning_rate": 4.342648393842453e-05,
      "loss": 0.2347,
      "mean_token_accuracy": 0.942656633630395,
      "num_tokens": 125608043.0,
      "step": 16950
    },
    {
      "entropy": 2.0076343685388567,
      "epoch": 0.2537666143908937,
      "grad_norm": 0.3775143325328827,
      "learning_rate": 4.34183668574587e-05,
      "loss": 0.215,
      "mean_token_accuracy": 0.9483071684837341,
      "num_tokens": 125679476.0,
      "step": 16960
    },
    {
      "entropy": 2.0084052227437494,
      "epoch": 0.2539162409323978,
      "grad_norm": 0.37910330295562744,
      "learning_rate": 4.341024552939972e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.94336854852736,
      "num_tokens": 125751129.0,
      "step": 16970
    },
    {
      "entropy": 2.036190717667341,
      "epoch": 0.25406586747390186,
      "grad_norm": 0.40339210629463196,
      "learning_rate": 4.340211995612541e-05,
      "loss": 0.2291,
      "mean_token_accuracy": 0.9453953098505735,
      "num_tokens": 125823456.0,
      "step": 16980
    },
    {
      "entropy": 2.0309246577322484,
      "epoch": 0.25421549401540594,
      "grad_norm": 0.3322254717350006,
      "learning_rate": 4.339399013951452e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9456148259341717,
      "num_tokens": 125898482.0,
      "step": 16990
    },
    {
      "entropy": 2.0426239393651486,
      "epoch": 0.25436512055690996,
      "grad_norm": 0.49731385707855225,
      "learning_rate": 4.338585608144683e-05,
      "loss": 0.2283,
      "mean_token_accuracy": 0.944075546041131,
      "num_tokens": 125970887.0,
      "step": 17000
    },
    {
      "epoch": 0.25436512055690996,
      "eval_entropy": 2.02679338157177,
      "eval_loss": 0.2466127872467041,
      "eval_mean_token_accuracy": 0.9448320381641387,
      "eval_num_tokens": 125970887.0,
      "eval_runtime": 543.27,
      "eval_samples_per_second": 36.814,
      "eval_steps_per_second": 9.204,
      "step": 17000
    },
    {
      "entropy": 2.0168982595205307,
      "epoch": 0.25451474709841404,
      "grad_norm": 0.3718942105770111,
      "learning_rate": 4.3377717783803066e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9435641348361969,
      "num_tokens": 126047886.0,
      "step": 17010
    },
    {
      "entropy": 2.0054264828562736,
      "epoch": 0.2546643736399181,
      "grad_norm": 0.44374823570251465,
      "learning_rate": 4.336957524846494e-05,
      "loss": 0.2381,
      "mean_token_accuracy": 0.9443205490708351,
      "num_tokens": 126124676.0,
      "step": 17020
    },
    {
      "entropy": 2.001351786404848,
      "epoch": 0.2548140001814222,
      "grad_norm": 0.3329358696937561,
      "learning_rate": 4.336142847731518e-05,
      "loss": 0.2029,
      "mean_token_accuracy": 0.9477333903312684,
      "num_tokens": 126198569.0,
      "step": 17030
    },
    {
      "entropy": 1.9631784930825233,
      "epoch": 0.2549636267229263,
      "grad_norm": 0.39166826009750366,
      "learning_rate": 4.3353277472237444e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9449137799441815,
      "num_tokens": 126273572.0,
      "step": 17040
    },
    {
      "entropy": 2.0052668027579785,
      "epoch": 0.2551132532644303,
      "grad_norm": 0.32352301478385925,
      "learning_rate": 4.334512223511638e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9435946058481932,
      "num_tokens": 126348967.0,
      "step": 17050
    },
    {
      "entropy": 1.9778543323278428,
      "epoch": 0.2552628798059344,
      "grad_norm": 0.3239785432815552,
      "learning_rate": 4.333696276783765e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9451056253165007,
      "num_tokens": 126421948.0,
      "step": 17060
    },
    {
      "entropy": 1.9993440955877304,
      "epoch": 0.25541250634743845,
      "grad_norm": 0.32256850600242615,
      "learning_rate": 4.332879907228784e-05,
      "loss": 0.2186,
      "mean_token_accuracy": 0.9425697054713964,
      "num_tokens": 126496428.0,
      "step": 17070
    },
    {
      "entropy": 1.9978416383266449,
      "epoch": 0.25556213288894253,
      "grad_norm": 0.313690185546875,
      "learning_rate": 4.332063115035456e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9494311794638634,
      "num_tokens": 126570848.0,
      "step": 17080
    },
    {
      "entropy": 2.0182333581149576,
      "epoch": 0.25571175943044655,
      "grad_norm": 0.3785570561885834,
      "learning_rate": 4.331245900392639e-05,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9434557281434536,
      "num_tokens": 126643932.0,
      "step": 17090
    },
    {
      "entropy": 1.9853479161858558,
      "epoch": 0.25586138597195063,
      "grad_norm": 0.3464750647544861,
      "learning_rate": 4.3304282634892836e-05,
      "loss": 0.2324,
      "mean_token_accuracy": 0.9406619675457477,
      "num_tokens": 126717843.0,
      "step": 17100
    },
    {
      "entropy": 1.9790563829243184,
      "epoch": 0.2560110125134547,
      "grad_norm": 0.3426876366138458,
      "learning_rate": 4.3296102045144456e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9472858447581529,
      "num_tokens": 126793079.0,
      "step": 17110
    },
    {
      "entropy": 2.0121135517954825,
      "epoch": 0.2561606390549588,
      "grad_norm": 0.35031428933143616,
      "learning_rate": 4.3287917236572734e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9414006408303977,
      "num_tokens": 126867589.0,
      "step": 17120
    },
    {
      "entropy": 1.9980814643204212,
      "epoch": 0.2563102655964628,
      "grad_norm": 0.4419938623905182,
      "learning_rate": 4.327972821107014e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9458811666816473,
      "num_tokens": 126941560.0,
      "step": 17130
    },
    {
      "entropy": 1.9779719687998294,
      "epoch": 0.2564598921379669,
      "grad_norm": 0.3605317175388336,
      "learning_rate": 4.3271534970530134e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9439295940101147,
      "num_tokens": 127016032.0,
      "step": 17140
    },
    {
      "entropy": 2.010127256065607,
      "epoch": 0.25660951867947096,
      "grad_norm": 0.2917308807373047,
      "learning_rate": 4.326333751684713e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.947640473023057,
      "num_tokens": 127088247.0,
      "step": 17150
    },
    {
      "entropy": 2.0435455590486526,
      "epoch": 0.25675914522097504,
      "grad_norm": 0.3274212181568146,
      "learning_rate": 4.325513585191652e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9439743306487799,
      "num_tokens": 127162723.0,
      "step": 17160
    },
    {
      "entropy": 2.03381944745779,
      "epoch": 0.25690877176247906,
      "grad_norm": 0.37538138031959534,
      "learning_rate": 4.3246929977634694e-05,
      "loss": 0.2223,
      "mean_token_accuracy": 0.9429311230778694,
      "num_tokens": 127237384.0,
      "step": 17170
    },
    {
      "entropy": 2.0363868504762648,
      "epoch": 0.25705839830398314,
      "grad_norm": 0.3380146026611328,
      "learning_rate": 4.3238719895898985e-05,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9434455685317517,
      "num_tokens": 127312529.0,
      "step": 17180
    },
    {
      "entropy": 2.039004324376583,
      "epoch": 0.2572080248454872,
      "grad_norm": 0.351798951625824,
      "learning_rate": 4.323050560860769e-05,
      "loss": 0.2019,
      "mean_token_accuracy": 0.9469769861549139,
      "num_tokens": 127386450.0,
      "step": 17190
    },
    {
      "entropy": 2.0314171835780144,
      "epoch": 0.2573576513869913,
      "grad_norm": 0.32466062903404236,
      "learning_rate": 4.322228711766014e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.942680211365223,
      "num_tokens": 127461005.0,
      "step": 17200
    },
    {
      "entropy": 2.022487550973892,
      "epoch": 0.25750727792849537,
      "grad_norm": 0.36314094066619873,
      "learning_rate": 4.3214064424956563e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9414778728038072,
      "num_tokens": 127533899.0,
      "step": 17210
    },
    {
      "entropy": 1.9910138800740242,
      "epoch": 0.2576569044699994,
      "grad_norm": 0.3252077102661133,
      "learning_rate": 4.320583753239821e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9486230961978436,
      "num_tokens": 127607851.0,
      "step": 17220
    },
    {
      "entropy": 1.9792562119662762,
      "epoch": 0.25780653101150347,
      "grad_norm": 0.4847642481327057,
      "learning_rate": 4.319760644188727e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9450223345309496,
      "num_tokens": 127681918.0,
      "step": 17230
    },
    {
      "entropy": 1.9889018058776855,
      "epoch": 0.25795615755300755,
      "grad_norm": 0.32753607630729675,
      "learning_rate": 4.318937115532693e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9469874978065491,
      "num_tokens": 127753352.0,
      "step": 17240
    },
    {
      "entropy": 2.0024723291397093,
      "epoch": 0.2581057840945116,
      "grad_norm": 0.3816656768321991,
      "learning_rate": 4.318113167462132e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9446849081665277,
      "num_tokens": 127824985.0,
      "step": 17250
    },
    {
      "entropy": 1.9502695381641388,
      "epoch": 0.25825541063601565,
      "grad_norm": 0.3409341871738434,
      "learning_rate": 4.317288800167557e-05,
      "loss": 0.2028,
      "mean_token_accuracy": 0.9492498282343149,
      "num_tokens": 127900533.0,
      "step": 17260
    },
    {
      "entropy": 1.9487469665706159,
      "epoch": 0.2584050371775197,
      "grad_norm": 0.32847729325294495,
      "learning_rate": 4.3164640138395764e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9460049189627171,
      "num_tokens": 127974650.0,
      "step": 17270
    },
    {
      "entropy": 1.952818927168846,
      "epoch": 0.2585546637190238,
      "grad_norm": 0.3086678087711334,
      "learning_rate": 4.315638808668894e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9491973910480738,
      "num_tokens": 128049316.0,
      "step": 17280
    },
    {
      "entropy": 2.0158708497881888,
      "epoch": 0.2587042902605279,
      "grad_norm": 0.34382888674736023,
      "learning_rate": 4.314813184846313e-05,
      "loss": 0.2469,
      "mean_token_accuracy": 0.93866643384099,
      "num_tokens": 128120506.0,
      "step": 17290
    },
    {
      "entropy": 2.022077728062868,
      "epoch": 0.2588539168020319,
      "grad_norm": 0.30771294236183167,
      "learning_rate": 4.3139871425627325e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9458696868270635,
      "num_tokens": 128195750.0,
      "step": 17300
    },
    {
      "entropy": 1.9928697161376476,
      "epoch": 0.259003543343536,
      "grad_norm": 0.3136599659919739,
      "learning_rate": 4.3131606820091474e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9444810140877962,
      "num_tokens": 128272874.0,
      "step": 17310
    },
    {
      "entropy": 2.008440475910902,
      "epoch": 0.25915316988504006,
      "grad_norm": 0.308378130197525,
      "learning_rate": 4.3123338033766506e-05,
      "loss": 0.2453,
      "mean_token_accuracy": 0.9369776047766208,
      "num_tokens": 128344606.0,
      "step": 17320
    },
    {
      "entropy": 2.0219190023839473,
      "epoch": 0.25930279642654414,
      "grad_norm": 0.35981300473213196,
      "learning_rate": 4.3115065068564316e-05,
      "loss": 0.2295,
      "mean_token_accuracy": 0.942860359326005,
      "num_tokens": 128417462.0,
      "step": 17330
    },
    {
      "entropy": 1.9899151727557183,
      "epoch": 0.2594524229680482,
      "grad_norm": 0.31166768074035645,
      "learning_rate": 4.310678792639775e-05,
      "loss": 0.2072,
      "mean_token_accuracy": 0.947108582407236,
      "num_tokens": 128491221.0,
      "step": 17340
    },
    {
      "entropy": 1.977298940718174,
      "epoch": 0.25960204950955224,
      "grad_norm": 0.3761517107486725,
      "learning_rate": 4.309850660918063e-05,
      "loss": 0.22,
      "mean_token_accuracy": 0.9459284603595733,
      "num_tokens": 128566158.0,
      "step": 17350
    },
    {
      "entropy": 1.9889166578650475,
      "epoch": 0.2597516760510563,
      "grad_norm": 0.35321831703186035,
      "learning_rate": 4.309022111882775e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9435504451394081,
      "num_tokens": 128638405.0,
      "step": 17360
    },
    {
      "entropy": 1.964811161905527,
      "epoch": 0.2599013025925604,
      "grad_norm": 0.35165607929229736,
      "learning_rate": 4.308193145725486e-05,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9455436542630196,
      "num_tokens": 128712216.0,
      "step": 17370
    },
    {
      "entropy": 1.9711891077458858,
      "epoch": 0.26005092913406447,
      "grad_norm": 0.3585442304611206,
      "learning_rate": 4.307363762637869e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.942692095413804,
      "num_tokens": 128788556.0,
      "step": 17380
    },
    {
      "entropy": 1.9759587064385413,
      "epoch": 0.2602005556755685,
      "grad_norm": 0.40561795234680176,
      "learning_rate": 4.3065339628116905e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9418573338538409,
      "num_tokens": 128863574.0,
      "step": 17390
    },
    {
      "entropy": 1.9541731216013432,
      "epoch": 0.26035018221707257,
      "grad_norm": 0.3849547505378723,
      "learning_rate": 4.305703746438815e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9432085048407316,
      "num_tokens": 128938799.0,
      "step": 17400
    },
    {
      "entropy": 1.9686829060316087,
      "epoch": 0.26049980875857665,
      "grad_norm": 0.35497909784317017,
      "learning_rate": 4.304873113711204e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9426066383719445,
      "num_tokens": 129010561.0,
      "step": 17410
    },
    {
      "entropy": 1.975174593180418,
      "epoch": 0.2606494353000807,
      "grad_norm": 0.3264389634132385,
      "learning_rate": 4.304042064820915e-05,
      "loss": 0.205,
      "mean_token_accuracy": 0.9479788269847631,
      "num_tokens": 129084874.0,
      "step": 17420
    },
    {
      "entropy": 1.9707801088690757,
      "epoch": 0.26079906184158475,
      "grad_norm": 0.41133826971054077,
      "learning_rate": 4.3032105999601e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.9441683806478978,
      "num_tokens": 129159563.0,
      "step": 17430
    },
    {
      "entropy": 1.9593721844255925,
      "epoch": 0.2609486883830888,
      "grad_norm": 0.4234388768672943,
      "learning_rate": 4.302378719321009e-05,
      "loss": 0.2418,
      "mean_token_accuracy": 0.9383463028818368,
      "num_tokens": 129232793.0,
      "step": 17440
    },
    {
      "entropy": 1.9605794072151184,
      "epoch": 0.2610983149245929,
      "grad_norm": 0.4137575626373291,
      "learning_rate": 4.301546423095989e-05,
      "loss": 0.2022,
      "mean_token_accuracy": 0.9507377818226814,
      "num_tokens": 129307007.0,
      "step": 17450
    },
    {
      "entropy": 1.9872503340244294,
      "epoch": 0.261247941466097,
      "grad_norm": 0.2910176217556,
      "learning_rate": 4.3007137114774785e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.945542873814702,
      "num_tokens": 129382829.0,
      "step": 17460
    },
    {
      "entropy": 1.9626220516860484,
      "epoch": 0.26139756800760106,
      "grad_norm": 0.29842138290405273,
      "learning_rate": 4.2998805846580184e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9435526240617037,
      "num_tokens": 129459384.0,
      "step": 17470
    },
    {
      "entropy": 1.961517696082592,
      "epoch": 0.2615471945491051,
      "grad_norm": 0.3537742793560028,
      "learning_rate": 4.2990470428302405e-05,
      "loss": 0.2207,
      "mean_token_accuracy": 0.9435574550181627,
      "num_tokens": 129535566.0,
      "step": 17480
    },
    {
      "entropy": 2.001833839714527,
      "epoch": 0.26169682109060916,
      "grad_norm": 0.35447975993156433,
      "learning_rate": 4.2982130861868755e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9445947904139758,
      "num_tokens": 129609625.0,
      "step": 17490
    },
    {
      "entropy": 1.9952963776886463,
      "epoch": 0.26184644763211323,
      "grad_norm": 0.37846505641937256,
      "learning_rate": 4.297378714920749e-05,
      "loss": 0.2153,
      "mean_token_accuracy": 0.9473113521933556,
      "num_tokens": 129685026.0,
      "step": 17500
    },
    {
      "entropy": 2.016940402984619,
      "epoch": 0.2619960741736173,
      "grad_norm": 0.3567720353603363,
      "learning_rate": 4.2965439292247814e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9449314881116152,
      "num_tokens": 129756995.0,
      "step": 17510
    },
    {
      "entropy": 1.9974692091345787,
      "epoch": 0.26214570071512133,
      "grad_norm": 0.36610057950019836,
      "learning_rate": 4.295708729291991e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9432786282151937,
      "num_tokens": 129830899.0,
      "step": 17520
    },
    {
      "entropy": 2.001037161052227,
      "epoch": 0.2622953272566254,
      "grad_norm": 0.3010680675506592,
      "learning_rate": 4.294873115315491e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9463255159556866,
      "num_tokens": 129906210.0,
      "step": 17530
    },
    {
      "entropy": 2.0504595056176185,
      "epoch": 0.2624449537981295,
      "grad_norm": 0.4136195182800293,
      "learning_rate": 4.2940370874884906e-05,
      "loss": 0.2523,
      "mean_token_accuracy": 0.9384560212492943,
      "num_tokens": 129978068.0,
      "step": 17540
    },
    {
      "entropy": 2.0636304400861265,
      "epoch": 0.26259458033963357,
      "grad_norm": 0.47486916184425354,
      "learning_rate": 4.293200646004292e-05,
      "loss": 0.242,
      "mean_token_accuracy": 0.9387151911854744,
      "num_tokens": 130052316.0,
      "step": 17550
    },
    {
      "entropy": 2.082205343991518,
      "epoch": 0.2627442068811376,
      "grad_norm": 0.40483376383781433,
      "learning_rate": 4.292363791056298e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9449768204241991,
      "num_tokens": 130125038.0,
      "step": 17560
    },
    {
      "entropy": 2.0309106297791004,
      "epoch": 0.26289383342264167,
      "grad_norm": 0.3578304648399353,
      "learning_rate": 4.291526522838002e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9455477550625802,
      "num_tokens": 130198234.0,
      "step": 17570
    },
    {
      "entropy": 2.0147901706397535,
      "epoch": 0.26304345996414574,
      "grad_norm": 0.3547304570674896,
      "learning_rate": 4.290688841542998e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.9418873999267816,
      "num_tokens": 130272004.0,
      "step": 17580
    },
    {
      "entropy": 2.0047319449484347,
      "epoch": 0.2631930865056498,
      "grad_norm": 0.34719160199165344,
      "learning_rate": 4.2898507473649706e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9467524759471416,
      "num_tokens": 130347702.0,
      "step": 17590
    },
    {
      "entropy": 2.0281103223562242,
      "epoch": 0.26334271304715384,
      "grad_norm": 0.3798537850379944,
      "learning_rate": 4.289012240497703e-05,
      "loss": 0.2037,
      "mean_token_accuracy": 0.9509062968194485,
      "num_tokens": 130421131.0,
      "step": 17600
    },
    {
      "entropy": 2.022486115992069,
      "epoch": 0.2634923395886579,
      "grad_norm": 0.44909578561782837,
      "learning_rate": 4.288173321135073e-05,
      "loss": 0.2338,
      "mean_token_accuracy": 0.9439417753368616,
      "num_tokens": 130495342.0,
      "step": 17610
    },
    {
      "entropy": 2.022466143220663,
      "epoch": 0.263641966130162,
      "grad_norm": 0.30635958909988403,
      "learning_rate": 4.287333989471052e-05,
      "loss": 0.2062,
      "mean_token_accuracy": 0.9496336128562689,
      "num_tokens": 130573953.0,
      "step": 17620
    },
    {
      "entropy": 2.0312775962054728,
      "epoch": 0.2637915926716661,
      "grad_norm": 0.3737289011478424,
      "learning_rate": 4.286494245699712e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.9403443302959203,
      "num_tokens": 130646781.0,
      "step": 17630
    },
    {
      "entropy": 2.0525998272001744,
      "epoch": 0.26394121921317015,
      "grad_norm": 0.38113486766815186,
      "learning_rate": 4.285654090015214e-05,
      "loss": 0.2326,
      "mean_token_accuracy": 0.940512691065669,
      "num_tokens": 130717668.0,
      "step": 17640
    },
    {
      "entropy": 2.029163409024477,
      "epoch": 0.2640908457546742,
      "grad_norm": 0.46075165271759033,
      "learning_rate": 4.2848135226118166e-05,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9450889442116022,
      "num_tokens": 130791338.0,
      "step": 17650
    },
    {
      "entropy": 1.9713542118668557,
      "epoch": 0.26424047229617825,
      "grad_norm": 0.2969851791858673,
      "learning_rate": 4.2839725436838754e-05,
      "loss": 0.2064,
      "mean_token_accuracy": 0.9488908022642135,
      "num_tokens": 130864093.0,
      "step": 17660
    },
    {
      "entropy": 1.9958363749086856,
      "epoch": 0.26439009883768233,
      "grad_norm": 0.291681706905365,
      "learning_rate": 4.2831311534258396e-05,
      "loss": 0.2006,
      "mean_token_accuracy": 0.948507721722126,
      "num_tokens": 130934749.0,
      "step": 17670
    },
    {
      "entropy": 2.024748580902815,
      "epoch": 0.2645397253791864,
      "grad_norm": 0.3385277986526489,
      "learning_rate": 4.282289352032252e-05,
      "loss": 0.2403,
      "mean_token_accuracy": 0.9381538767367601,
      "num_tokens": 131006490.0,
      "step": 17680
    },
    {
      "entropy": 2.04764044880867,
      "epoch": 0.26468935192069043,
      "grad_norm": 0.49041318893432617,
      "learning_rate": 4.281447139697754e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.942776982486248,
      "num_tokens": 131078439.0,
      "step": 17690
    },
    {
      "entropy": 2.03910723477602,
      "epoch": 0.2648389784621945,
      "grad_norm": 0.3468276560306549,
      "learning_rate": 4.280604516617078e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9486701887100935,
      "num_tokens": 131154427.0,
      "step": 17700
    },
    {
      "entropy": 2.037722332775593,
      "epoch": 0.2649886050036986,
      "grad_norm": 0.3466048538684845,
      "learning_rate": 4.279761482985056e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9459042120724916,
      "num_tokens": 131228821.0,
      "step": 17710
    },
    {
      "entropy": 2.0034206956624985,
      "epoch": 0.26513823154520266,
      "grad_norm": 0.2954052686691284,
      "learning_rate": 4.278918038996611e-05,
      "loss": 0.2061,
      "mean_token_accuracy": 0.946030754595995,
      "num_tokens": 131302554.0,
      "step": 17720
    },
    {
      "entropy": 1.9941191963851452,
      "epoch": 0.2652878580867067,
      "grad_norm": 0.36720165610313416,
      "learning_rate": 4.278074184846762e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9404993798583746,
      "num_tokens": 131376909.0,
      "step": 17730
    },
    {
      "entropy": 2.0092363335192203,
      "epoch": 0.26543748462821076,
      "grad_norm": 0.3861556947231293,
      "learning_rate": 4.277229920730623e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9456748113036155,
      "num_tokens": 131449664.0,
      "step": 17740
    },
    {
      "entropy": 2.019600921869278,
      "epoch": 0.26558711116971484,
      "grad_norm": 0.4270144999027252,
      "learning_rate": 4.276385246843404e-05,
      "loss": 0.2377,
      "mean_token_accuracy": 0.9397767677903175,
      "num_tokens": 131521087.0,
      "step": 17750
    },
    {
      "entropy": 2.046319001168013,
      "epoch": 0.2657367377112189,
      "grad_norm": 0.33829551935195923,
      "learning_rate": 4.2755401633804075e-05,
      "loss": 0.2037,
      "mean_token_accuracy": 0.9449324559420347,
      "num_tokens": 131597125.0,
      "step": 17760
    },
    {
      "entropy": 2.0328876174986363,
      "epoch": 0.265886364252723,
      "grad_norm": 0.31292375922203064,
      "learning_rate": 4.274694670537033e-05,
      "loss": 0.2432,
      "mean_token_accuracy": 0.9418448995798826,
      "num_tokens": 131670323.0,
      "step": 17770
    },
    {
      "entropy": 2.0468028388917445,
      "epoch": 0.266035990794227,
      "grad_norm": 0.3730217516422272,
      "learning_rate": 4.273848768508771e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9489680986851454,
      "num_tokens": 131742813.0,
      "step": 17780
    },
    {
      "entropy": 2.0350642174482347,
      "epoch": 0.2661856173357311,
      "grad_norm": 0.34829699993133545,
      "learning_rate": 4.273002457491212e-05,
      "loss": 0.2132,
      "mean_token_accuracy": 0.9468727346509695,
      "num_tokens": 131818194.0,
      "step": 17790
    },
    {
      "entropy": 2.0639965884387492,
      "epoch": 0.2663352438772352,
      "grad_norm": 0.3347161114215851,
      "learning_rate": 4.272155737680037e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9400759331882,
      "num_tokens": 131889949.0,
      "step": 17800
    },
    {
      "entropy": 2.011091100424528,
      "epoch": 0.26648487041873925,
      "grad_norm": 0.3353201150894165,
      "learning_rate": 4.271308609271022e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.942489180713892,
      "num_tokens": 131965365.0,
      "step": 17810
    },
    {
      "entropy": 2.060721480101347,
      "epoch": 0.2666344969602433,
      "grad_norm": 0.3964075744152069,
      "learning_rate": 4.270461072460039e-05,
      "loss": 0.2621,
      "mean_token_accuracy": 0.9336626414209604,
      "num_tokens": 132036943.0,
      "step": 17820
    },
    {
      "entropy": 2.002261830866337,
      "epoch": 0.26678412350174735,
      "grad_norm": 0.38764169812202454,
      "learning_rate": 4.269613127443054e-05,
      "loss": 0.2204,
      "mean_token_accuracy": 0.9458250042051077,
      "num_tokens": 132111293.0,
      "step": 17830
    },
    {
      "entropy": 2.0242681615054607,
      "epoch": 0.2669337500432514,
      "grad_norm": 0.3036015033721924,
      "learning_rate": 4.268764774416126e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9468665599822998,
      "num_tokens": 132183344.0,
      "step": 17840
    },
    {
      "entropy": 2.0303299449384213,
      "epoch": 0.2670833765847555,
      "grad_norm": 0.3105604350566864,
      "learning_rate": 4.2679160135754086e-05,
      "loss": 0.2007,
      "mean_token_accuracy": 0.9504918597638607,
      "num_tokens": 132255943.0,
      "step": 17850
    },
    {
      "entropy": 2.0628513015806673,
      "epoch": 0.2672330031262595,
      "grad_norm": 0.33919087052345276,
      "learning_rate": 4.267066845117153e-05,
      "loss": 0.2416,
      "mean_token_accuracy": 0.9390486862510443,
      "num_tokens": 132326230.0,
      "step": 17860
    },
    {
      "entropy": 2.00702800899744,
      "epoch": 0.2673826296677636,
      "grad_norm": 0.298063725233078,
      "learning_rate": 4.2662172692377e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9463082946836948,
      "num_tokens": 132402308.0,
      "step": 17870
    },
    {
      "entropy": 1.9925678826868534,
      "epoch": 0.2675322562092677,
      "grad_norm": 0.341201514005661,
      "learning_rate": 4.265367286133487e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.948339295387268,
      "num_tokens": 132477042.0,
      "step": 17880
    },
    {
      "entropy": 2.002018129825592,
      "epoch": 0.26768188275077176,
      "grad_norm": 0.37781527638435364,
      "learning_rate": 4.264516896001046e-05,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9447626590728759,
      "num_tokens": 132551960.0,
      "step": 17890
    },
    {
      "entropy": 2.0631848059594633,
      "epoch": 0.26783150929227584,
      "grad_norm": 0.3190608322620392,
      "learning_rate": 4.263666099037001e-05,
      "loss": 0.2347,
      "mean_token_accuracy": 0.9401345096528531,
      "num_tokens": 132624821.0,
      "step": 17900
    },
    {
      "entropy": 2.0559956721961496,
      "epoch": 0.26798113583377986,
      "grad_norm": 0.30890265107154846,
      "learning_rate": 4.262814895438073e-05,
      "loss": 0.2143,
      "mean_token_accuracy": 0.9479293666779995,
      "num_tokens": 132698605.0,
      "step": 17910
    },
    {
      "entropy": 2.078419268131256,
      "epoch": 0.26813076237528394,
      "grad_norm": 0.32682353258132935,
      "learning_rate": 4.2619632854010725e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.943324676156044,
      "num_tokens": 132771628.0,
      "step": 17920
    },
    {
      "entropy": 2.0519722640514373,
      "epoch": 0.268280388916788,
      "grad_norm": 0.31532716751098633,
      "learning_rate": 4.261111269122911e-05,
      "loss": 0.1974,
      "mean_token_accuracy": 0.9496322073042393,
      "num_tokens": 132847151.0,
      "step": 17930
    },
    {
      "entropy": 1.9829540766775609,
      "epoch": 0.2684300154582921,
      "grad_norm": 0.3580506443977356,
      "learning_rate": 4.2602588468005874e-05,
      "loss": 0.2082,
      "mean_token_accuracy": 0.9499303434044123,
      "num_tokens": 132923703.0,
      "step": 17940
    },
    {
      "entropy": 2.0283055998384953,
      "epoch": 0.2685796419997961,
      "grad_norm": 0.47611668705940247,
      "learning_rate": 4.259406018631197e-05,
      "loss": 0.234,
      "mean_token_accuracy": 0.9415302246809005,
      "num_tokens": 132998478.0,
      "step": 17950
    },
    {
      "entropy": 2.040881954878569,
      "epoch": 0.2687292685413002,
      "grad_norm": 0.30716952681541443,
      "learning_rate": 4.258552784811929e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9376135040074587,
      "num_tokens": 133070068.0,
      "step": 17960
    },
    {
      "entropy": 2.043821970373392,
      "epoch": 0.26887889508280427,
      "grad_norm": 0.4280816614627838,
      "learning_rate": 4.2576991455400655e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9450179640203714,
      "num_tokens": 133145191.0,
      "step": 17970
    },
    {
      "entropy": 2.0277457468211653,
      "epoch": 0.26902852162430835,
      "grad_norm": 0.35947221517562866,
      "learning_rate": 4.2568451010129853e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9440630115568638,
      "num_tokens": 133217884.0,
      "step": 17980
    },
    {
      "entropy": 2.0370886988937853,
      "epoch": 0.26917814816581237,
      "grad_norm": 0.2640123665332794,
      "learning_rate": 4.255990651428156e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9457904122769832,
      "num_tokens": 133292399.0,
      "step": 17990
    },
    {
      "entropy": 2.0545321866869926,
      "epoch": 0.26932777470731645,
      "grad_norm": 0.40209871530532837,
      "learning_rate": 4.2551357969831437e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.945240993052721,
      "num_tokens": 133369999.0,
      "step": 18000
    },
    {
      "epoch": 0.26932777470731645,
      "eval_entropy": 2.032932476854324,
      "eval_loss": 0.2456362247467041,
      "eval_mean_token_accuracy": 0.9449632157921791,
      "eval_num_tokens": 133369999.0,
      "eval_runtime": 543.1241,
      "eval_samples_per_second": 36.824,
      "eval_steps_per_second": 9.206,
      "step": 18000
    },
    {
      "entropy": 2.0408055007457735,
      "epoch": 0.2694774012488205,
      "grad_norm": 0.3894859850406647,
      "learning_rate": 4.254280537875605e-05,
      "loss": 0.224,
      "mean_token_accuracy": 0.945162595808506,
      "num_tokens": 133444279.0,
      "step": 18010
    },
    {
      "entropy": 2.040571527183056,
      "epoch": 0.2696270277903246,
      "grad_norm": 0.3294428884983063,
      "learning_rate": 4.253424874303291e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9471924025565386,
      "num_tokens": 133516898.0,
      "step": 18020
    },
    {
      "entropy": 2.025993775576353,
      "epoch": 0.2697766543318286,
      "grad_norm": 0.2972036600112915,
      "learning_rate": 4.2525688064640464e-05,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9465685278177262,
      "num_tokens": 133590262.0,
      "step": 18030
    },
    {
      "entropy": 2.0547771252691747,
      "epoch": 0.2699262808733327,
      "grad_norm": 0.3184581995010376,
      "learning_rate": 4.251712334555809e-05,
      "loss": 0.2077,
      "mean_token_accuracy": 0.9474600825458765,
      "num_tokens": 133664812.0,
      "step": 18040
    },
    {
      "entropy": 2.072434861212969,
      "epoch": 0.2700759074148368,
      "grad_norm": 0.401377409696579,
      "learning_rate": 4.2508554587766105e-05,
      "loss": 0.1998,
      "mean_token_accuracy": 0.9504003830254077,
      "num_tokens": 133742279.0,
      "step": 18050
    },
    {
      "entropy": 2.0875652648508547,
      "epoch": 0.27022553395634086,
      "grad_norm": 0.3234099745750427,
      "learning_rate": 4.249998179324577e-05,
      "loss": 0.2458,
      "mean_token_accuracy": 0.9395138043910265,
      "num_tokens": 133815216.0,
      "step": 18060
    },
    {
      "entropy": 2.0594569988548757,
      "epoch": 0.27037516049784494,
      "grad_norm": 0.2818845510482788,
      "learning_rate": 4.249140496397925e-05,
      "loss": 0.2098,
      "mean_token_accuracy": 0.9482460029423236,
      "num_tokens": 133888788.0,
      "step": 18070
    },
    {
      "entropy": 2.0344219408929347,
      "epoch": 0.27052478703934896,
      "grad_norm": 0.34061676263809204,
      "learning_rate": 4.248282410194967e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9489198680967093,
      "num_tokens": 133962592.0,
      "step": 18080
    },
    {
      "entropy": 2.0163093373179435,
      "epoch": 0.27067441358085303,
      "grad_norm": 0.4497212767601013,
      "learning_rate": 4.2474239209141075e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9442314021289349,
      "num_tokens": 134036851.0,
      "step": 18090
    },
    {
      "entropy": 2.017015743255615,
      "epoch": 0.2708240401223571,
      "grad_norm": 0.3502942621707916,
      "learning_rate": 4.246565028753845e-05,
      "loss": 0.2075,
      "mean_token_accuracy": 0.9471006710082293,
      "num_tokens": 134110631.0,
      "step": 18100
    },
    {
      "entropy": 2.0393383227288724,
      "epoch": 0.2709736666638612,
      "grad_norm": 0.37357452511787415,
      "learning_rate": 4.2457057339127705e-05,
      "loss": 0.2237,
      "mean_token_accuracy": 0.9457779925316572,
      "num_tokens": 134184413.0,
      "step": 18110
    },
    {
      "entropy": 2.0620221897959707,
      "epoch": 0.2711232932053652,
      "grad_norm": 0.3612636923789978,
      "learning_rate": 4.244846036589567e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9476705510169268,
      "num_tokens": 134258957.0,
      "step": 18120
    },
    {
      "entropy": 2.0866783775389193,
      "epoch": 0.2712729197468693,
      "grad_norm": 0.294169157743454,
      "learning_rate": 4.243985936983013e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9419513683766126,
      "num_tokens": 134333576.0,
      "step": 18130
    },
    {
      "entropy": 2.037560198456049,
      "epoch": 0.27142254628837337,
      "grad_norm": 0.3890853524208069,
      "learning_rate": 4.243125435291979e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9462948691099882,
      "num_tokens": 134407346.0,
      "step": 18140
    },
    {
      "entropy": 2.0153480976819993,
      "epoch": 0.27157217282987745,
      "grad_norm": 0.2998529374599457,
      "learning_rate": 4.242264531715428e-05,
      "loss": 0.2004,
      "mean_token_accuracy": 0.9478793989866972,
      "num_tokens": 134483911.0,
      "step": 18150
    },
    {
      "entropy": 2.03355066254735,
      "epoch": 0.27172179937138147,
      "grad_norm": 0.4071115255355835,
      "learning_rate": 4.241403226452416e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9447440791875124,
      "num_tokens": 134558075.0,
      "step": 18160
    },
    {
      "entropy": 2.03444909080863,
      "epoch": 0.27187142591288554,
      "grad_norm": 0.29406195878982544,
      "learning_rate": 4.240541519702092e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9405380580574274,
      "num_tokens": 134633426.0,
      "step": 18170
    },
    {
      "entropy": 2.06948067471385,
      "epoch": 0.2720210524543896,
      "grad_norm": 0.32778263092041016,
      "learning_rate": 4.2396794116636976e-05,
      "loss": 0.2009,
      "mean_token_accuracy": 0.948231928795576,
      "num_tokens": 134707687.0,
      "step": 18180
    },
    {
      "entropy": 2.0545158363878726,
      "epoch": 0.2721706789958937,
      "grad_norm": 0.3893728256225586,
      "learning_rate": 4.23881690253657e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9450697984546423,
      "num_tokens": 134780320.0,
      "step": 18190
    },
    {
      "entropy": 2.0420808009803295,
      "epoch": 0.2723203055373978,
      "grad_norm": 0.40593183040618896,
      "learning_rate": 4.237953992520133e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9433854795992375,
      "num_tokens": 134853971.0,
      "step": 18200
    },
    {
      "entropy": 2.0449700109660625,
      "epoch": 0.2724699320789018,
      "grad_norm": 0.3766489624977112,
      "learning_rate": 4.237090681813909e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.9448163963854312,
      "num_tokens": 134927865.0,
      "step": 18210
    },
    {
      "entropy": 2.0795838005840777,
      "epoch": 0.2726195586204059,
      "grad_norm": 0.43642085790634155,
      "learning_rate": 4.23622697061751e-05,
      "loss": 0.2316,
      "mean_token_accuracy": 0.94608996771276,
      "num_tokens": 135001022.0,
      "step": 18220
    },
    {
      "entropy": 2.0937659993767737,
      "epoch": 0.27276918516190996,
      "grad_norm": 0.2979442775249481,
      "learning_rate": 4.235362859130642e-05,
      "loss": 0.2174,
      "mean_token_accuracy": 0.9440000209957361,
      "num_tokens": 135078012.0,
      "step": 18230
    },
    {
      "entropy": 2.0923197217285634,
      "epoch": 0.27291881170341403,
      "grad_norm": 0.44040948152542114,
      "learning_rate": 4.234498347553103e-05,
      "loss": 0.2419,
      "mean_token_accuracy": 0.9399584282189608,
      "num_tokens": 135151736.0,
      "step": 18240
    },
    {
      "entropy": 2.0501592487096785,
      "epoch": 0.27306843824491805,
      "grad_norm": 0.36849233508110046,
      "learning_rate": 4.233633436084783e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9465037569403648,
      "num_tokens": 135227001.0,
      "step": 18250
    },
    {
      "entropy": 2.032037995010614,
      "epoch": 0.27321806478642213,
      "grad_norm": 0.3081227242946625,
      "learning_rate": 4.232768124925664e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9465609356760979,
      "num_tokens": 135302010.0,
      "step": 18260
    },
    {
      "entropy": 2.041413481533527,
      "epoch": 0.2733676913279262,
      "grad_norm": 0.3574448227882385,
      "learning_rate": 4.231902414275824e-05,
      "loss": 0.212,
      "mean_token_accuracy": 0.9455951485782862,
      "num_tokens": 135375800.0,
      "step": 18270
    },
    {
      "entropy": 2.0589251168072225,
      "epoch": 0.2735173178694303,
      "grad_norm": 0.341811865568161,
      "learning_rate": 4.2310363043354276e-05,
      "loss": 0.2125,
      "mean_token_accuracy": 0.9483492769300937,
      "num_tokens": 135450070.0,
      "step": 18280
    },
    {
      "entropy": 2.110490904003382,
      "epoch": 0.2736669444109343,
      "grad_norm": 0.35117724537849426,
      "learning_rate": 4.230169795304737e-05,
      "loss": 0.241,
      "mean_token_accuracy": 0.9425245705991984,
      "num_tokens": 135521763.0,
      "step": 18290
    },
    {
      "entropy": 2.0464411430060863,
      "epoch": 0.2738165709524384,
      "grad_norm": 0.4113922119140625,
      "learning_rate": 4.229302887384103e-05,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9458016745746136,
      "num_tokens": 135594041.0,
      "step": 18300
    },
    {
      "entropy": 2.062289177626371,
      "epoch": 0.27396619749394246,
      "grad_norm": 0.5518506765365601,
      "learning_rate": 4.2284355807739714e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.941171970590949,
      "num_tokens": 135668386.0,
      "step": 18310
    },
    {
      "entropy": 2.078730411082506,
      "epoch": 0.27411582403544654,
      "grad_norm": 0.36489707231521606,
      "learning_rate": 4.227567875674878e-05,
      "loss": 0.2504,
      "mean_token_accuracy": 0.9422944277524948,
      "num_tokens": 135742681.0,
      "step": 18320
    },
    {
      "entropy": 2.1214917838573455,
      "epoch": 0.2742654505769506,
      "grad_norm": 0.36626705527305603,
      "learning_rate": 4.2266997722874525e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9441429376602173,
      "num_tokens": 135817804.0,
      "step": 18330
    },
    {
      "entropy": 2.09757549315691,
      "epoch": 0.27441507711845464,
      "grad_norm": 0.3245505392551422,
      "learning_rate": 4.225831270812414e-05,
      "loss": 0.2428,
      "mean_token_accuracy": 0.9402012165635825,
      "num_tokens": 135892332.0,
      "step": 18340
    },
    {
      "entropy": 2.079206131398678,
      "epoch": 0.2745647036599587,
      "grad_norm": 0.38551098108291626,
      "learning_rate": 4.2249623714505785e-05,
      "loss": 0.2348,
      "mean_token_accuracy": 0.9424356084316969,
      "num_tokens": 135964909.0,
      "step": 18350
    },
    {
      "entropy": 2.0848313689231874,
      "epoch": 0.2747143302014628,
      "grad_norm": 0.4020768404006958,
      "learning_rate": 4.224093074402849e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9437496196478605,
      "num_tokens": 136036134.0,
      "step": 18360
    },
    {
      "entropy": 2.0565269708633425,
      "epoch": 0.2748639567429669,
      "grad_norm": 0.3413435220718384,
      "learning_rate": 4.223223379870222e-05,
      "loss": 0.228,
      "mean_token_accuracy": 0.9418129827827215,
      "num_tokens": 136111376.0,
      "step": 18370
    },
    {
      "entropy": 2.062454446405172,
      "epoch": 0.2750135832844709,
      "grad_norm": 0.3518320322036743,
      "learning_rate": 4.222353288053788e-05,
      "loss": 0.2082,
      "mean_token_accuracy": 0.9461835123598575,
      "num_tokens": 136184642.0,
      "step": 18380
    },
    {
      "entropy": 2.07863834425807,
      "epoch": 0.275163209825975,
      "grad_norm": 0.4114658236503601,
      "learning_rate": 4.2214827991547266e-05,
      "loss": 0.2464,
      "mean_token_accuracy": 0.9394934225827456,
      "num_tokens": 136255251.0,
      "step": 18390
    },
    {
      "entropy": 2.062755359709263,
      "epoch": 0.27531283636747905,
      "grad_norm": 0.42160865664482117,
      "learning_rate": 4.220611913374311e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9423040580004454,
      "num_tokens": 136327673.0,
      "step": 18400
    },
    {
      "entropy": 2.052606646716595,
      "epoch": 0.27546246290898313,
      "grad_norm": 0.44263756275177,
      "learning_rate": 4.219740630913904e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.941598741710186,
      "num_tokens": 136399657.0,
      "step": 18410
    },
    {
      "entropy": 2.0223962023854254,
      "epoch": 0.27561208945048715,
      "grad_norm": 0.3220590353012085,
      "learning_rate": 4.2188689519749636e-05,
      "loss": 0.2248,
      "mean_token_accuracy": 0.9472240276634694,
      "num_tokens": 136474705.0,
      "step": 18420
    },
    {
      "entropy": 2.0413773842155933,
      "epoch": 0.27576171599199123,
      "grad_norm": 0.3709854483604431,
      "learning_rate": 4.217996876759037e-05,
      "loss": 0.2101,
      "mean_token_accuracy": 0.9452756538987159,
      "num_tokens": 136549061.0,
      "step": 18430
    },
    {
      "entropy": 2.0624384708702563,
      "epoch": 0.2759113425334953,
      "grad_norm": 0.2936898469924927,
      "learning_rate": 4.217124405467762e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9419232547283173,
      "num_tokens": 136626400.0,
      "step": 18440
    },
    {
      "entropy": 2.0403049543499945,
      "epoch": 0.2760609690749994,
      "grad_norm": 0.37017953395843506,
      "learning_rate": 4.21625153830287e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9463361319154501,
      "num_tokens": 136703152.0,
      "step": 18450
    },
    {
      "entropy": 2.0422281220555307,
      "epoch": 0.2762105956165034,
      "grad_norm": 0.3478231728076935,
      "learning_rate": 4.215378275466185e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9414387229830027,
      "num_tokens": 136775701.0,
      "step": 18460
    },
    {
      "entropy": 2.027935326844454,
      "epoch": 0.2763602221580075,
      "grad_norm": 0.3442736864089966,
      "learning_rate": 4.21450461715962e-05,
      "loss": 0.1898,
      "mean_token_accuracy": 0.9508050210773945,
      "num_tokens": 136850589.0,
      "step": 18470
    },
    {
      "entropy": 1.9965142816305161,
      "epoch": 0.27650984869951156,
      "grad_norm": 0.2997758090496063,
      "learning_rate": 4.2136305635851796e-05,
      "loss": 0.2033,
      "mean_token_accuracy": 0.9486421305686236,
      "num_tokens": 136926048.0,
      "step": 18480
    },
    {
      "entropy": 1.9957321412861346,
      "epoch": 0.27665947524101564,
      "grad_norm": 0.31858164072036743,
      "learning_rate": 4.2127561149449605e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.942990842834115,
      "num_tokens": 137002238.0,
      "step": 18490
    },
    {
      "entropy": 2.0288641512393952,
      "epoch": 0.2768091017825197,
      "grad_norm": 0.4636212885379791,
      "learning_rate": 4.211881271441153e-05,
      "loss": 0.2407,
      "mean_token_accuracy": 0.9436327964067459,
      "num_tokens": 137076318.0,
      "step": 18500
    },
    {
      "entropy": 2.0751604810357094,
      "epoch": 0.27695872832402374,
      "grad_norm": 0.3495842516422272,
      "learning_rate": 4.2110060332760335e-05,
      "loss": 0.2663,
      "mean_token_accuracy": 0.9383338689804077,
      "num_tokens": 137149302.0,
      "step": 18510
    },
    {
      "entropy": 2.077486325800419,
      "epoch": 0.2771083548655278,
      "grad_norm": 0.38012605905532837,
      "learning_rate": 4.2101304006519735e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9448123019188642,
      "num_tokens": 137223791.0,
      "step": 18520
    },
    {
      "entropy": 2.0725693568587302,
      "epoch": 0.2772579814070319,
      "grad_norm": 0.3547988831996918,
      "learning_rate": 4.209254373771437e-05,
      "loss": 0.2074,
      "mean_token_accuracy": 0.9450439121574163,
      "num_tokens": 137296985.0,
      "step": 18530
    },
    {
      "entropy": 2.0524710163474085,
      "epoch": 0.277407607948536,
      "grad_norm": 0.37117061018943787,
      "learning_rate": 4.208377952836974e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.9486215464770794,
      "num_tokens": 137371610.0,
      "step": 18540
    },
    {
      "entropy": 2.0582168929278852,
      "epoch": 0.27755723449004,
      "grad_norm": 0.36203354597091675,
      "learning_rate": 4.207501138051231e-05,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9440382104367018,
      "num_tokens": 137447076.0,
      "step": 18550
    },
    {
      "entropy": 2.025244905799627,
      "epoch": 0.27770686103154407,
      "grad_norm": 0.3644031584262848,
      "learning_rate": 4.2066239296169414e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9462102502584457,
      "num_tokens": 137522208.0,
      "step": 18560
    },
    {
      "entropy": 2.0489648029208185,
      "epoch": 0.27785648757304815,
      "grad_norm": 0.35710328817367554,
      "learning_rate": 4.205746327736932e-05,
      "loss": 0.2322,
      "mean_token_accuracy": 0.9408591013401747,
      "num_tokens": 137593526.0,
      "step": 18570
    },
    {
      "entropy": 2.0361582323908807,
      "epoch": 0.2780061141145522,
      "grad_norm": 0.38068822026252747,
      "learning_rate": 4.20486833261412e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9474812593311072,
      "num_tokens": 137667469.0,
      "step": 18580
    },
    {
      "entropy": 2.0413844235241414,
      "epoch": 0.27815574065605625,
      "grad_norm": 0.3494391143321991,
      "learning_rate": 4.2039899444515147e-05,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9445367429405451,
      "num_tokens": 137742002.0,
      "step": 18590
    },
    {
      "entropy": 2.0514455653727053,
      "epoch": 0.2783053671975603,
      "grad_norm": 0.35180479288101196,
      "learning_rate": 4.2031111634522133e-05,
      "loss": 0.2501,
      "mean_token_accuracy": 0.9396644573658705,
      "num_tokens": 137814318.0,
      "step": 18600
    },
    {
      "entropy": 2.043489322066307,
      "epoch": 0.2784549937390644,
      "grad_norm": 0.3107980489730835,
      "learning_rate": 4.2022319898194054e-05,
      "loss": 0.225,
      "mean_token_accuracy": 0.9448707785457373,
      "num_tokens": 137892345.0,
      "step": 18610
    },
    {
      "entropy": 2.038166705518961,
      "epoch": 0.2786046202805685,
      "grad_norm": 0.3189834654331207,
      "learning_rate": 4.201352423756373e-05,
      "loss": 0.2249,
      "mean_token_accuracy": 0.9463358268141746,
      "num_tokens": 137967334.0,
      "step": 18620
    },
    {
      "entropy": 2.0123340927064417,
      "epoch": 0.27875424682207256,
      "grad_norm": 0.35150328278541565,
      "learning_rate": 4.2004724654664874e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9487344469875097,
      "num_tokens": 138039431.0,
      "step": 18630
    },
    {
      "entropy": 2.0151193886995316,
      "epoch": 0.2789038733635766,
      "grad_norm": 0.3368518352508545,
      "learning_rate": 4.199592115153209e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9443730127066374,
      "num_tokens": 138114386.0,
      "step": 18640
    },
    {
      "entropy": 2.016951708495617,
      "epoch": 0.27905349990508066,
      "grad_norm": 0.38006192445755005,
      "learning_rate": 4.198711373020092e-05,
      "loss": 0.2429,
      "mean_token_accuracy": 0.9414054803550244,
      "num_tokens": 138187157.0,
      "step": 18650
    },
    {
      "entropy": 2.0137404680252073,
      "epoch": 0.27920312644658474,
      "grad_norm": 0.3471709191799164,
      "learning_rate": 4.1978302392707795e-05,
      "loss": 0.2467,
      "mean_token_accuracy": 0.9362635936588048,
      "num_tokens": 138261052.0,
      "step": 18660
    },
    {
      "entropy": 2.0529631271958353,
      "epoch": 0.2793527529880888,
      "grad_norm": 0.36627161502838135,
      "learning_rate": 4.1969487141090044e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9440194319933652,
      "num_tokens": 138334014.0,
      "step": 18670
    },
    {
      "entropy": 2.0277971342206,
      "epoch": 0.27950237952959284,
      "grad_norm": 0.3198646008968353,
      "learning_rate": 4.1960667977385906e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9469820514321328,
      "num_tokens": 138408076.0,
      "step": 18680
    },
    {
      "entropy": 2.0425813741981984,
      "epoch": 0.2796520060710969,
      "grad_norm": 0.39036881923675537,
      "learning_rate": 4.195184490363455e-05,
      "loss": 0.2127,
      "mean_token_accuracy": 0.9463781978935003,
      "num_tokens": 138484654.0,
      "step": 18690
    },
    {
      "entropy": 2.0550139889121057,
      "epoch": 0.279801632612601,
      "grad_norm": 0.4529547095298767,
      "learning_rate": 4.194301792187602e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9451748415827751,
      "num_tokens": 138557827.0,
      "step": 18700
    },
    {
      "entropy": 2.0668136782944204,
      "epoch": 0.27995125915410507,
      "grad_norm": 0.3962840437889099,
      "learning_rate": 4.1934187034151264e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9478241514414549,
      "num_tokens": 138629774.0,
      "step": 18710
    },
    {
      "entropy": 2.0423460379242897,
      "epoch": 0.2801008856956091,
      "grad_norm": 0.3541772961616516,
      "learning_rate": 4.192535224250213e-05,
      "loss": 0.2213,
      "mean_token_accuracy": 0.94769446849823,
      "num_tokens": 138703439.0,
      "step": 18720
    },
    {
      "entropy": 2.0338203951716425,
      "epoch": 0.28025051223711317,
      "grad_norm": 0.3149760961532593,
      "learning_rate": 4.191651354897141e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.945298059284687,
      "num_tokens": 138778466.0,
      "step": 18730
    },
    {
      "entropy": 2.042553113400936,
      "epoch": 0.28040013877861725,
      "grad_norm": 0.43838343024253845,
      "learning_rate": 4.190767095560273e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9468694549053908,
      "num_tokens": 138854163.0,
      "step": 18740
    },
    {
      "entropy": 2.0193994611501696,
      "epoch": 0.2805497653201213,
      "grad_norm": 0.3140980899333954,
      "learning_rate": 4.1898824464440686e-05,
      "loss": 0.2029,
      "mean_token_accuracy": 0.9491469334810972,
      "num_tokens": 138928915.0,
      "step": 18750
    },
    {
      "entropy": 2.026965218037367,
      "epoch": 0.28069939186162535,
      "grad_norm": 0.38596779108047485,
      "learning_rate": 4.188997407753073e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9461881373077631,
      "num_tokens": 139000526.0,
      "step": 18760
    },
    {
      "entropy": 2.0746323578059673,
      "epoch": 0.2808490184031294,
      "grad_norm": 0.35778096318244934,
      "learning_rate": 4.1881119796919225e-05,
      "loss": 0.2024,
      "mean_token_accuracy": 0.9499216876924038,
      "num_tokens": 139077104.0,
      "step": 18770
    },
    {
      "entropy": 2.0472045801579952,
      "epoch": 0.2809986449446335,
      "grad_norm": 0.31523221731185913,
      "learning_rate": 4.1872261624653455e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9433153495192528,
      "num_tokens": 139152257.0,
      "step": 18780
    },
    {
      "entropy": 2.0438988633453845,
      "epoch": 0.2811482714861376,
      "grad_norm": 0.3384609818458557,
      "learning_rate": 4.186339956278157e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9469064172357321,
      "num_tokens": 139224851.0,
      "step": 18790
    },
    {
      "entropy": 2.0316933512687685,
      "epoch": 0.28129789802764166,
      "grad_norm": 0.34056273102760315,
      "learning_rate": 4.185453361335264e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9431099519133568,
      "num_tokens": 139299366.0,
      "step": 18800
    },
    {
      "entropy": 2.0515247009694577,
      "epoch": 0.2814475245691457,
      "grad_norm": 0.36333808302879333,
      "learning_rate": 4.184566377841664e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9453124329447746,
      "num_tokens": 139372178.0,
      "step": 18810
    },
    {
      "entropy": 2.084352830797434,
      "epoch": 0.28159715111064976,
      "grad_norm": 0.4016410708427429,
      "learning_rate": 4.183679006002441e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.946783883869648,
      "num_tokens": 139444690.0,
      "step": 18820
    },
    {
      "entropy": 2.064262942224741,
      "epoch": 0.28174677765215383,
      "grad_norm": 0.3707505762577057,
      "learning_rate": 4.182791246022774e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9474179904907942,
      "num_tokens": 139521546.0,
      "step": 18830
    },
    {
      "entropy": 2.061519002914429,
      "epoch": 0.2818964041936579,
      "grad_norm": 0.3917161524295807,
      "learning_rate": 4.1819030981079275e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9425201997160911,
      "num_tokens": 139593497.0,
      "step": 18840
    },
    {
      "entropy": 1.9889956191182137,
      "epoch": 0.28204603073516193,
      "grad_norm": 0.3814077377319336,
      "learning_rate": 4.181014562463257e-05,
      "loss": 0.1945,
      "mean_token_accuracy": 0.9537968579679728,
      "num_tokens": 139668834.0,
      "step": 18850
    },
    {
      "entropy": 2.024089213460684,
      "epoch": 0.282195657276666,
      "grad_norm": 0.3451828956604004,
      "learning_rate": 4.180125639294209e-05,
      "loss": 0.2516,
      "mean_token_accuracy": 0.9386969532817602,
      "num_tokens": 139740687.0,
      "step": 18860
    },
    {
      "entropy": 2.044738408178091,
      "epoch": 0.2823452838181701,
      "grad_norm": 0.31049394607543945,
      "learning_rate": 4.179236328806317e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9456838808953762,
      "num_tokens": 139814297.0,
      "step": 18870
    },
    {
      "entropy": 2.023719023913145,
      "epoch": 0.28249491035967417,
      "grad_norm": 0.34184572100639343,
      "learning_rate": 4.178346631205205e-05,
      "loss": 0.2166,
      "mean_token_accuracy": 0.9460842613130808,
      "num_tokens": 139886065.0,
      "step": 18880
    },
    {
      "entropy": 2.013763380795717,
      "epoch": 0.2826445369011782,
      "grad_norm": 0.3594011664390564,
      "learning_rate": 4.177456546696588e-05,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9472302999347448,
      "num_tokens": 139958509.0,
      "step": 18890
    },
    {
      "entropy": 2.0229734420776366,
      "epoch": 0.28279416344268227,
      "grad_norm": 0.36598172783851624,
      "learning_rate": 4.176566075486269e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9460081171244383,
      "num_tokens": 140032110.0,
      "step": 18900
    },
    {
      "entropy": 2.0380223214626314,
      "epoch": 0.28294378998418634,
      "grad_norm": 0.38212352991104126,
      "learning_rate": 4.175675217780141e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9459437381476163,
      "num_tokens": 140106692.0,
      "step": 18910
    },
    {
      "entropy": 2.0251020193099976,
      "epoch": 0.2830934165256904,
      "grad_norm": 0.39991524815559387,
      "learning_rate": 4.174783973784186e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9422227840870618,
      "num_tokens": 140180597.0,
      "step": 18920
    },
    {
      "entropy": 2.0390100069344044,
      "epoch": 0.2832430430671945,
      "grad_norm": 0.3565160930156708,
      "learning_rate": 4.173892343704475e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9423737730830908,
      "num_tokens": 140251960.0,
      "step": 18930
    },
    {
      "entropy": 1.9707071863114833,
      "epoch": 0.2833926696086985,
      "grad_norm": 0.3760647773742676,
      "learning_rate": 4.173000327747169e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9497558526694775,
      "num_tokens": 140327146.0,
      "step": 18940
    },
    {
      "entropy": 1.9672926016151906,
      "epoch": 0.2835422961502026,
      "grad_norm": 0.36126407980918884,
      "learning_rate": 4.172107926118519e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9462910059839487,
      "num_tokens": 140401937.0,
      "step": 18950
    },
    {
      "entropy": 1.979753314703703,
      "epoch": 0.2836919226917067,
      "grad_norm": 0.3449506461620331,
      "learning_rate": 4.171215139024864e-05,
      "loss": 0.2241,
      "mean_token_accuracy": 0.9444792300462723,
      "num_tokens": 140477295.0,
      "step": 18960
    },
    {
      "entropy": 1.9834757447242737,
      "epoch": 0.28384154923321075,
      "grad_norm": 0.41755038499832153,
      "learning_rate": 4.1703219666726293e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.944455549493432,
      "num_tokens": 140549920.0,
      "step": 18970
    },
    {
      "entropy": 1.9919896595180036,
      "epoch": 0.2839911757747148,
      "grad_norm": 0.38521695137023926,
      "learning_rate": 4.1694284092683356e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.9443584073334932,
      "num_tokens": 140623353.0,
      "step": 18980
    },
    {
      "entropy": 2.015780097991228,
      "epoch": 0.28414080231621885,
      "grad_norm": 0.34652242064476013,
      "learning_rate": 4.1685344670185884e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9450482990592718,
      "num_tokens": 140696112.0,
      "step": 18990
    },
    {
      "entropy": 2.0005426786839964,
      "epoch": 0.28429042885772293,
      "grad_norm": 0.2863911986351013,
      "learning_rate": 4.167640140130082e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9479904975742102,
      "num_tokens": 140768181.0,
      "step": 19000
    },
    {
      "epoch": 0.28429042885772293,
      "eval_entropy": 2.02031144156456,
      "eval_loss": 0.24498778581619263,
      "eval_mean_token_accuracy": 0.9452503816246987,
      "eval_num_tokens": 140768181.0,
      "eval_runtime": 555.3498,
      "eval_samples_per_second": 36.013,
      "eval_steps_per_second": 9.003,
      "step": 19000
    },
    {
      "entropy": 2.027205886691809,
      "epoch": 0.284440055399227,
      "grad_norm": 0.2785838842391968,
      "learning_rate": 4.166745428809601e-05,
      "loss": 0.2167,
      "mean_token_accuracy": 0.9448889102786779,
      "num_tokens": 140842749.0,
      "step": 19010
    },
    {
      "entropy": 2.0289974197745324,
      "epoch": 0.28458968194073103,
      "grad_norm": 0.3426930606365204,
      "learning_rate": 4.16585033326402e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.941736264154315,
      "num_tokens": 140914934.0,
      "step": 19020
    },
    {
      "entropy": 1.997819370776415,
      "epoch": 0.2847393084822351,
      "grad_norm": 0.5024489164352417,
      "learning_rate": 4.1649548537003005e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9462288957089185,
      "num_tokens": 140989704.0,
      "step": 19030
    },
    {
      "entropy": 1.9965325437486172,
      "epoch": 0.2848889350237392,
      "grad_norm": 0.3459457457065582,
      "learning_rate": 4.164058990325493e-05,
      "loss": 0.2339,
      "mean_token_accuracy": 0.940864809602499,
      "num_tokens": 141065113.0,
      "step": 19040
    },
    {
      "entropy": 1.998188631236553,
      "epoch": 0.28503856156524326,
      "grad_norm": 0.32374563813209534,
      "learning_rate": 4.163162743346735e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9461703088134528,
      "num_tokens": 141137665.0,
      "step": 19050
    },
    {
      "entropy": 2.0321890391409396,
      "epoch": 0.28518818810674734,
      "grad_norm": 0.2236821949481964,
      "learning_rate": 4.1622661129712576e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9446377016603946,
      "num_tokens": 141215825.0,
      "step": 19060
    },
    {
      "entropy": 2.031581325083971,
      "epoch": 0.28533781464825136,
      "grad_norm": 0.37125664949417114,
      "learning_rate": 4.161369099406376e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9467072561383247,
      "num_tokens": 141291748.0,
      "step": 19070
    },
    {
      "entropy": 1.9868048392236233,
      "epoch": 0.28548744118975544,
      "grad_norm": 0.43177497386932373,
      "learning_rate": 4.160471702859498e-05,
      "loss": 0.2097,
      "mean_token_accuracy": 0.9499659389257431,
      "num_tokens": 141368464.0,
      "step": 19080
    },
    {
      "entropy": 2.0178396977484225,
      "epoch": 0.2856370677312595,
      "grad_norm": 0.38260743021965027,
      "learning_rate": 4.159573923538115e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9464736625552177,
      "num_tokens": 141440329.0,
      "step": 19090
    },
    {
      "entropy": 2.003493504971266,
      "epoch": 0.2857866942727636,
      "grad_norm": 0.38395771384239197,
      "learning_rate": 4.158675761649812e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.9464287050068378,
      "num_tokens": 141516321.0,
      "step": 19100
    },
    {
      "entropy": 2.045782498270273,
      "epoch": 0.2859363208142676,
      "grad_norm": 0.39461347460746765,
      "learning_rate": 4.1577772174022564e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9423138335347175,
      "num_tokens": 141592988.0,
      "step": 19110
    },
    {
      "entropy": 1.9975715763866901,
      "epoch": 0.2860859473557717,
      "grad_norm": 0.3696184754371643,
      "learning_rate": 4.156878291003211e-05,
      "loss": 0.2244,
      "mean_token_accuracy": 0.9454155080020428,
      "num_tokens": 141668219.0,
      "step": 19120
    },
    {
      "entropy": 1.9911875426769257,
      "epoch": 0.2862355738972758,
      "grad_norm": 0.3416394591331482,
      "learning_rate": 4.155978982660522e-05,
      "loss": 0.2092,
      "mean_token_accuracy": 0.947189799323678,
      "num_tokens": 141741721.0,
      "step": 19130
    },
    {
      "entropy": 2.0163810029625893,
      "epoch": 0.28638520043877985,
      "grad_norm": 0.3842177391052246,
      "learning_rate": 4.155079292582126e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9443196214735508,
      "num_tokens": 141815139.0,
      "step": 19140
    },
    {
      "entropy": 2.0187873542308807,
      "epoch": 0.2865348269802839,
      "grad_norm": 0.4396173059940338,
      "learning_rate": 4.154179220976048e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.943511925637722,
      "num_tokens": 141889709.0,
      "step": 19150
    },
    {
      "entropy": 2.0159444510936737,
      "epoch": 0.28668445352178795,
      "grad_norm": 0.30550047755241394,
      "learning_rate": 4.153278768050399e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9404990077018738,
      "num_tokens": 141964485.0,
      "step": 19160
    },
    {
      "entropy": 1.9971608579158784,
      "epoch": 0.28683408006329203,
      "grad_norm": 0.363066166639328,
      "learning_rate": 4.1523779340133805e-05,
      "loss": 0.2359,
      "mean_token_accuracy": 0.9444653902202844,
      "num_tokens": 142038962.0,
      "step": 19170
    },
    {
      "entropy": 1.9893890514969825,
      "epoch": 0.2869837066047961,
      "grad_norm": 0.3109620213508606,
      "learning_rate": 4.151476719073282e-05,
      "loss": 0.2084,
      "mean_token_accuracy": 0.9501341089606286,
      "num_tokens": 142114536.0,
      "step": 19180
    },
    {
      "entropy": 2.00932981967926,
      "epoch": 0.28713333314630013,
      "grad_norm": 0.3176131546497345,
      "learning_rate": 4.150575123438479e-05,
      "loss": 0.2107,
      "mean_token_accuracy": 0.945628322660923,
      "num_tokens": 142192272.0,
      "step": 19190
    },
    {
      "entropy": 2.0379151582717894,
      "epoch": 0.2872829596878042,
      "grad_norm": 0.3827197849750519,
      "learning_rate": 4.149673147317437e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9455595865845681,
      "num_tokens": 142268136.0,
      "step": 19200
    },
    {
      "entropy": 2.003543919324875,
      "epoch": 0.2874325862293083,
      "grad_norm": 0.39703768491744995,
      "learning_rate": 4.14877079091871e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9420764662325383,
      "num_tokens": 142342846.0,
      "step": 19210
    },
    {
      "entropy": 2.002555912733078,
      "epoch": 0.28758221277081236,
      "grad_norm": 0.3634537160396576,
      "learning_rate": 4.1478680544509375e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9474540788680315,
      "num_tokens": 142419379.0,
      "step": 19220
    },
    {
      "entropy": 1.9828537493944167,
      "epoch": 0.28773183931231644,
      "grad_norm": 0.3495555818080902,
      "learning_rate": 4.146964938122849e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.946489142999053,
      "num_tokens": 142493843.0,
      "step": 19230
    },
    {
      "entropy": 1.9926489010453223,
      "epoch": 0.28788146585382046,
      "grad_norm": 0.3554561734199524,
      "learning_rate": 4.146061442143261e-05,
      "loss": 0.2413,
      "mean_token_accuracy": 0.9429801497608423,
      "num_tokens": 142568613.0,
      "step": 19240
    },
    {
      "entropy": 2.076005844771862,
      "epoch": 0.28803109239532454,
      "grad_norm": 0.4142340421676636,
      "learning_rate": 4.145157566721077e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9413616731762886,
      "num_tokens": 142639524.0,
      "step": 19250
    },
    {
      "entropy": 2.0507871650159357,
      "epoch": 0.2881807189368286,
      "grad_norm": 0.40420281887054443,
      "learning_rate": 4.14425331206529e-05,
      "loss": 0.2396,
      "mean_token_accuracy": 0.941462367027998,
      "num_tokens": 142714303.0,
      "step": 19260
    },
    {
      "entropy": 1.9908677011728286,
      "epoch": 0.2883303454783327,
      "grad_norm": 0.33306199312210083,
      "learning_rate": 4.1433486783849796e-05,
      "loss": 0.2081,
      "mean_token_accuracy": 0.9449459314346313,
      "num_tokens": 142791991.0,
      "step": 19270
    },
    {
      "entropy": 2.0439110986888407,
      "epoch": 0.2884799720198367,
      "grad_norm": 0.39535558223724365,
      "learning_rate": 4.1424436658893134e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9424667019397021,
      "num_tokens": 142867304.0,
      "step": 19280
    },
    {
      "entropy": 2.0362544685602186,
      "epoch": 0.2886295985613408,
      "grad_norm": 0.3420041799545288,
      "learning_rate": 4.1415382747875456e-05,
      "loss": 0.2299,
      "mean_token_accuracy": 0.9427588600665331,
      "num_tokens": 142937144.0,
      "step": 19290
    },
    {
      "entropy": 1.9993120409548282,
      "epoch": 0.28877922510284487,
      "grad_norm": 0.35424086451530457,
      "learning_rate": 4.140632505289019e-05,
      "loss": 0.2165,
      "mean_token_accuracy": 0.9457357801496983,
      "num_tokens": 143010041.0,
      "step": 19300
    },
    {
      "entropy": 2.028030426055193,
      "epoch": 0.28892885164434895,
      "grad_norm": 0.421956866979599,
      "learning_rate": 4.1397263576031644e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9474210020154714,
      "num_tokens": 143084543.0,
      "step": 19310
    },
    {
      "entropy": 2.025577488541603,
      "epoch": 0.28907847818585297,
      "grad_norm": 0.377094566822052,
      "learning_rate": 4.138819831939498e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9474115096032619,
      "num_tokens": 143163019.0,
      "step": 19320
    },
    {
      "entropy": 2.0393564343452453,
      "epoch": 0.28922810472735705,
      "grad_norm": 0.3564658761024475,
      "learning_rate": 4.137912928507626e-05,
      "loss": 0.2216,
      "mean_token_accuracy": 0.9469887781888247,
      "num_tokens": 143235748.0,
      "step": 19330
    },
    {
      "entropy": 2.0214167945086956,
      "epoch": 0.2893777312688611,
      "grad_norm": 0.3326159119606018,
      "learning_rate": 4.13700564751724e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9463445458561182,
      "num_tokens": 143310300.0,
      "step": 19340
    },
    {
      "entropy": 2.038857350498438,
      "epoch": 0.2895273578103652,
      "grad_norm": 0.3593785762786865,
      "learning_rate": 4.1360979891781185e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9464337833225727,
      "num_tokens": 143386209.0,
      "step": 19350
    },
    {
      "entropy": 2.010245291888714,
      "epoch": 0.2896769843518693,
      "grad_norm": 0.41534116864204407,
      "learning_rate": 4.1351899537001294e-05,
      "loss": 0.2027,
      "mean_token_accuracy": 0.9472188122570515,
      "num_tokens": 143462973.0,
      "step": 19360
    },
    {
      "entropy": 2.071219979971647,
      "epoch": 0.2898266108933733,
      "grad_norm": 0.3706682324409485,
      "learning_rate": 4.1342815412932276e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9374992590397596,
      "num_tokens": 143534722.0,
      "step": 19370
    },
    {
      "entropy": 2.043961289525032,
      "epoch": 0.2899762374348774,
      "grad_norm": 0.3208138942718506,
      "learning_rate": 4.133372752167452e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9436949111521244,
      "num_tokens": 143608727.0,
      "step": 19380
    },
    {
      "entropy": 1.982392144203186,
      "epoch": 0.29012586397638146,
      "grad_norm": 0.2898438274860382,
      "learning_rate": 4.132463586532932e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9445894751697779,
      "num_tokens": 143683061.0,
      "step": 19390
    },
    {
      "entropy": 1.9748047687113286,
      "epoch": 0.29027549051788554,
      "grad_norm": 0.3796473443508148,
      "learning_rate": 4.131554044599883e-05,
      "loss": 0.2343,
      "mean_token_accuracy": 0.944151297584176,
      "num_tokens": 143758219.0,
      "step": 19400
    },
    {
      "entropy": 1.9767847508192062,
      "epoch": 0.29042511705938956,
      "grad_norm": 0.3115903437137604,
      "learning_rate": 4.130644126578607e-05,
      "loss": 0.1996,
      "mean_token_accuracy": 0.9499203372746706,
      "num_tokens": 143830655.0,
      "step": 19410
    },
    {
      "entropy": 1.9845007054507733,
      "epoch": 0.29057474360089364,
      "grad_norm": 0.3089817762374878,
      "learning_rate": 4.129733832679493e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9446139160543681,
      "num_tokens": 143906544.0,
      "step": 19420
    },
    {
      "entropy": 2.001465954631567,
      "epoch": 0.2907243701423977,
      "grad_norm": 0.26999640464782715,
      "learning_rate": 4.128823163113018e-05,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9457458753138781,
      "num_tokens": 143982199.0,
      "step": 19430
    },
    {
      "entropy": 2.011623577773571,
      "epoch": 0.2908739966839018,
      "grad_norm": 0.3828546106815338,
      "learning_rate": 4.127912118089743e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9462559968233109,
      "num_tokens": 144056359.0,
      "step": 19440
    },
    {
      "entropy": 2.006866566836834,
      "epoch": 0.2910236232254058,
      "grad_norm": 0.35948505997657776,
      "learning_rate": 4.12700069782032e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9453226558864116,
      "num_tokens": 144130014.0,
      "step": 19450
    },
    {
      "entropy": 1.9932134814560414,
      "epoch": 0.2911732497669099,
      "grad_norm": 0.37443414330482483,
      "learning_rate": 4.126088902515485e-05,
      "loss": 0.2465,
      "mean_token_accuracy": 0.9424038253724575,
      "num_tokens": 144205413.0,
      "step": 19460
    },
    {
      "entropy": 2.0161292761564256,
      "epoch": 0.29132287630841397,
      "grad_norm": 0.3447151482105255,
      "learning_rate": 4.1251767323860615e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9430126335471869,
      "num_tokens": 144278970.0,
      "step": 19470
    },
    {
      "entropy": 2.0269536308944227,
      "epoch": 0.29147250284991805,
      "grad_norm": 0.2640238404273987,
      "learning_rate": 4.124264187642958e-05,
      "loss": 0.211,
      "mean_token_accuracy": 0.9454924125224352,
      "num_tokens": 144355082.0,
      "step": 19480
    },
    {
      "entropy": 2.0321346059441567,
      "epoch": 0.2916221293914221,
      "grad_norm": 0.35263651609420776,
      "learning_rate": 4.123351268497172e-05,
      "loss": 0.2028,
      "mean_token_accuracy": 0.9497700709849596,
      "num_tokens": 144430360.0,
      "step": 19490
    },
    {
      "entropy": 2.048047775030136,
      "epoch": 0.29177175593292615,
      "grad_norm": 0.34211254119873047,
      "learning_rate": 4.122437975159787e-05,
      "loss": 0.2337,
      "mean_token_accuracy": 0.9445621252059937,
      "num_tokens": 144504841.0,
      "step": 19500
    },
    {
      "entropy": 2.0127721816301345,
      "epoch": 0.2919213824744302,
      "grad_norm": 0.37980666756629944,
      "learning_rate": 4.121524307841972e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9427575219422579,
      "num_tokens": 144579270.0,
      "step": 19510
    },
    {
      "entropy": 2.02224944755435,
      "epoch": 0.2920710090159343,
      "grad_norm": 0.3055545389652252,
      "learning_rate": 4.120610266754985e-05,
      "loss": 0.2331,
      "mean_token_accuracy": 0.946198432892561,
      "num_tokens": 144650032.0,
      "step": 19520
    },
    {
      "entropy": 1.9993158817291259,
      "epoch": 0.2922206355574384,
      "grad_norm": 0.3966371715068817,
      "learning_rate": 4.1196958521101666e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9443116266280412,
      "num_tokens": 144723709.0,
      "step": 19530
    },
    {
      "entropy": 2.0156676970422267,
      "epoch": 0.2923702620989424,
      "grad_norm": 0.2709941565990448,
      "learning_rate": 4.118781064118945e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9463618010282516,
      "num_tokens": 144795612.0,
      "step": 19540
    },
    {
      "entropy": 2.0284367904067038,
      "epoch": 0.2925198886404465,
      "grad_norm": 0.3392142951488495,
      "learning_rate": 4.1178659029928385e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9485639542341232,
      "num_tokens": 144870682.0,
      "step": 19550
    },
    {
      "entropy": 2.03234349116683,
      "epoch": 0.29266951518195056,
      "grad_norm": 0.3494682013988495,
      "learning_rate": 4.116950368943446e-05,
      "loss": 0.246,
      "mean_token_accuracy": 0.939114723354578,
      "num_tokens": 144943383.0,
      "step": 19560
    },
    {
      "entropy": 2.0234033286571504,
      "epoch": 0.29281914172345463,
      "grad_norm": 0.3731505274772644,
      "learning_rate": 4.116034462182456e-05,
      "loss": 0.208,
      "mean_token_accuracy": 0.9488781578838825,
      "num_tokens": 145016448.0,
      "step": 19570
    },
    {
      "entropy": 2.0094316944479944,
      "epoch": 0.29296876826495866,
      "grad_norm": 0.29979071021080017,
      "learning_rate": 4.115118182921642e-05,
      "loss": 0.1929,
      "mean_token_accuracy": 0.9486490063369274,
      "num_tokens": 145090066.0,
      "step": 19580
    },
    {
      "entropy": 2.0141825422644617,
      "epoch": 0.29311839480646273,
      "grad_norm": 0.3503182530403137,
      "learning_rate": 4.114201531372865e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9429158348590135,
      "num_tokens": 145166099.0,
      "step": 19590
    },
    {
      "entropy": 1.994695892930031,
      "epoch": 0.2932680213479668,
      "grad_norm": 0.42686930298805237,
      "learning_rate": 4.113284507748071e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.9456496275961399,
      "num_tokens": 145238062.0,
      "step": 19600
    },
    {
      "entropy": 2.0545146852731704,
      "epoch": 0.2934176478894709,
      "grad_norm": 0.4200618863105774,
      "learning_rate": 4.112367112259291e-05,
      "loss": 0.2411,
      "mean_token_accuracy": 0.9377747546881438,
      "num_tokens": 145309086.0,
      "step": 19610
    },
    {
      "entropy": 2.0282765552401543,
      "epoch": 0.2935672744309749,
      "grad_norm": 0.4016728401184082,
      "learning_rate": 4.1114493451186446e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9442884162068367,
      "num_tokens": 145385771.0,
      "step": 19620
    },
    {
      "entropy": 2.0473625123500825,
      "epoch": 0.293716900972479,
      "grad_norm": 0.3272794783115387,
      "learning_rate": 4.1105312065383355e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.945939016342163,
      "num_tokens": 145458202.0,
      "step": 19630
    },
    {
      "entropy": 2.017625433951616,
      "epoch": 0.29386652751398307,
      "grad_norm": 0.37465739250183105,
      "learning_rate": 4.1096126967306536e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.9393696300685406,
      "num_tokens": 145529976.0,
      "step": 19640
    },
    {
      "entropy": 2.007725605368614,
      "epoch": 0.29401615405548714,
      "grad_norm": 0.35948389768600464,
      "learning_rate": 4.108693815907974e-05,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9491938173770904,
      "num_tokens": 145601327.0,
      "step": 19650
    },
    {
      "entropy": 2.0509208396077154,
      "epoch": 0.2941657805969912,
      "grad_norm": 0.3008934259414673,
      "learning_rate": 4.10777456428276e-05,
      "loss": 0.2589,
      "mean_token_accuracy": 0.934301046282053,
      "num_tokens": 145671581.0,
      "step": 19660
    },
    {
      "entropy": 2.0360881075263024,
      "epoch": 0.29431540713849524,
      "grad_norm": 0.42072632908821106,
      "learning_rate": 4.1068549420675575e-05,
      "loss": 0.1943,
      "mean_token_accuracy": 0.9494276024401188,
      "num_tokens": 145745750.0,
      "step": 19670
    },
    {
      "entropy": 2.045329646021128,
      "epoch": 0.2944650336799993,
      "grad_norm": 0.46567341685295105,
      "learning_rate": 4.105934949474999e-05,
      "loss": 0.2405,
      "mean_token_accuracy": 0.9418666385114193,
      "num_tokens": 145819587.0,
      "step": 19680
    },
    {
      "entropy": 2.023375128954649,
      "epoch": 0.2946146602215034,
      "grad_norm": 0.38378089666366577,
      "learning_rate": 4.105014586717804e-05,
      "loss": 0.2243,
      "mean_token_accuracy": 0.9466793645173311,
      "num_tokens": 145894487.0,
      "step": 19690
    },
    {
      "entropy": 2.0113506980240343,
      "epoch": 0.2947642867630075,
      "grad_norm": 0.3974805176258087,
      "learning_rate": 4.1040938540087776e-05,
      "loss": 0.2093,
      "mean_token_accuracy": 0.9468415066599846,
      "num_tokens": 145969077.0,
      "step": 19700
    },
    {
      "entropy": 1.9913309440016747,
      "epoch": 0.2949139133045115,
      "grad_norm": 0.31204694509506226,
      "learning_rate": 4.1031727515608074e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9460851542651654,
      "num_tokens": 146044307.0,
      "step": 19710
    },
    {
      "entropy": 2.005850520730019,
      "epoch": 0.2950635398460156,
      "grad_norm": 0.4471956193447113,
      "learning_rate": 4.102251279586872e-05,
      "loss": 0.2269,
      "mean_token_accuracy": 0.9421543184667825,
      "num_tokens": 146116940.0,
      "step": 19720
    },
    {
      "entropy": 1.9804161600768566,
      "epoch": 0.29521316638751965,
      "grad_norm": 0.34054404497146606,
      "learning_rate": 4.1013294383000285e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9456283111125231,
      "num_tokens": 146192863.0,
      "step": 19730
    },
    {
      "entropy": 1.9990719571709632,
      "epoch": 0.29536279292902373,
      "grad_norm": 0.3339153528213501,
      "learning_rate": 4.100407227913425e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9446085713803768,
      "num_tokens": 146264276.0,
      "step": 19740
    },
    {
      "entropy": 1.985842365771532,
      "epoch": 0.29551241947052775,
      "grad_norm": 0.30111441016197205,
      "learning_rate": 4.099484648640292e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.945923775061965,
      "num_tokens": 146338451.0,
      "step": 19750
    },
    {
      "entropy": 2.016427589952946,
      "epoch": 0.29566204601203183,
      "grad_norm": 0.30611446499824524,
      "learning_rate": 4.098561700693947e-05,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9428063247352838,
      "num_tokens": 146412263.0,
      "step": 19760
    },
    {
      "entropy": 2.0081195876002313,
      "epoch": 0.2958116725535359,
      "grad_norm": 0.31625494360923767,
      "learning_rate": 4.097638384287792e-05,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9454154785722494,
      "num_tokens": 146483612.0,
      "step": 19770
    },
    {
      "entropy": 2.0008365988731383,
      "epoch": 0.29596129909504,
      "grad_norm": 0.3585444986820221,
      "learning_rate": 4.096714699635315e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.9458553910255432,
      "num_tokens": 146558585.0,
      "step": 19780
    },
    {
      "entropy": 1.993656063079834,
      "epoch": 0.29611092563654406,
      "grad_norm": 0.35468655824661255,
      "learning_rate": 4.095790646950086e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9443667333573103,
      "num_tokens": 146632495.0,
      "step": 19790
    },
    {
      "entropy": 1.9758825704455376,
      "epoch": 0.2962605521780481,
      "grad_norm": 0.43393224477767944,
      "learning_rate": 4.094866226445763e-05,
      "loss": 0.2032,
      "mean_token_accuracy": 0.9497439593076706,
      "num_tokens": 146709693.0,
      "step": 19800
    },
    {
      "entropy": 1.9865000322461128,
      "epoch": 0.29641017871955216,
      "grad_norm": 0.37553897500038147,
      "learning_rate": 4.09394143833609e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9455925840884447,
      "num_tokens": 146783679.0,
      "step": 19810
    },
    {
      "entropy": 2.001179201900959,
      "epoch": 0.29655980526105624,
      "grad_norm": 0.3062836229801178,
      "learning_rate": 4.0930162828348945e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9439009569585324,
      "num_tokens": 146856949.0,
      "step": 19820
    },
    {
      "entropy": 2.013218576461077,
      "epoch": 0.2967094318025603,
      "grad_norm": 0.34968891739845276,
      "learning_rate": 4.092090760156087e-05,
      "loss": 0.2118,
      "mean_token_accuracy": 0.9468849711120129,
      "num_tokens": 146931126.0,
      "step": 19830
    },
    {
      "entropy": 1.9840328246355057,
      "epoch": 0.29685905834406434,
      "grad_norm": 0.39939936995506287,
      "learning_rate": 4.091164870513666e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9458829414099454,
      "num_tokens": 147006849.0,
      "step": 19840
    },
    {
      "entropy": 2.0314989760518074,
      "epoch": 0.2970086848855684,
      "grad_norm": 0.3434779942035675,
      "learning_rate": 4.0902386141217135e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.940334552153945,
      "num_tokens": 147081819.0,
      "step": 19850
    },
    {
      "entropy": 2.007120017707348,
      "epoch": 0.2971583114270725,
      "grad_norm": 0.2926064729690552,
      "learning_rate": 4.0893119911943964e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9461166225373745,
      "num_tokens": 147156103.0,
      "step": 19860
    },
    {
      "entropy": 2.0122553169727326,
      "epoch": 0.2973079379685766,
      "grad_norm": 0.3425265848636627,
      "learning_rate": 4.088385001945966e-05,
      "loss": 0.2132,
      "mean_token_accuracy": 0.9451687458902598,
      "num_tokens": 147228567.0,
      "step": 19870
    },
    {
      "entropy": 1.9995957918465137,
      "epoch": 0.2974575645100806,
      "grad_norm": 0.3475918173789978,
      "learning_rate": 4.08745764659076e-05,
      "loss": 0.2094,
      "mean_token_accuracy": 0.9486463360488415,
      "num_tokens": 147303495.0,
      "step": 19880
    },
    {
      "entropy": 2.0184695474803447,
      "epoch": 0.2976071910515847,
      "grad_norm": 0.4519127309322357,
      "learning_rate": 4.0865299253431974e-05,
      "loss": 0.2312,
      "mean_token_accuracy": 0.9418253682553768,
      "num_tokens": 147376487.0,
      "step": 19890
    },
    {
      "entropy": 2.0048968225717543,
      "epoch": 0.29775681759308875,
      "grad_norm": 0.34926164150238037,
      "learning_rate": 4.085601838417786e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9442270547151566,
      "num_tokens": 147452696.0,
      "step": 19900
    },
    {
      "entropy": 2.0098508350551127,
      "epoch": 0.29790644413459283,
      "grad_norm": 0.36559855937957764,
      "learning_rate": 4.084673386029115e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9443294763565063,
      "num_tokens": 147526882.0,
      "step": 19910
    },
    {
      "entropy": 1.9947369545698166,
      "epoch": 0.2980560706760969,
      "grad_norm": 0.44540220499038696,
      "learning_rate": 4.0837445683918586e-05,
      "loss": 0.2198,
      "mean_token_accuracy": 0.9443842884153127,
      "num_tokens": 147602093.0,
      "step": 19920
    },
    {
      "entropy": 2.023253387957811,
      "epoch": 0.29820569721760093,
      "grad_norm": 0.39571183919906616,
      "learning_rate": 4.0828153857207766e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.94560980014503,
      "num_tokens": 147675439.0,
      "step": 19930
    },
    {
      "entropy": 2.011840672791004,
      "epoch": 0.298355323759105,
      "grad_norm": 0.35391393303871155,
      "learning_rate": 4.081885838230712e-05,
      "loss": 0.2169,
      "mean_token_accuracy": 0.9453230824321508,
      "num_tokens": 147749723.0,
      "step": 19940
    },
    {
      "entropy": 2.011302823573351,
      "epoch": 0.2985049503006091,
      "grad_norm": 0.3616487979888916,
      "learning_rate": 4.0809559261365945e-05,
      "loss": 0.2356,
      "mean_token_accuracy": 0.9423373688012362,
      "num_tokens": 147824188.0,
      "step": 19950
    },
    {
      "entropy": 2.0500920340418816,
      "epoch": 0.29865457684211316,
      "grad_norm": 0.36686018109321594,
      "learning_rate": 4.080025649653434e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9410277891904115,
      "num_tokens": 147897400.0,
      "step": 19960
    },
    {
      "entropy": 2.0559603571891785,
      "epoch": 0.2988042033836172,
      "grad_norm": 0.33165568113327026,
      "learning_rate": 4.079095008996328e-05,
      "loss": 0.2285,
      "mean_token_accuracy": 0.9429252095520496,
      "num_tokens": 147971127.0,
      "step": 19970
    },
    {
      "entropy": 2.0634577453136442,
      "epoch": 0.29895382992512126,
      "grad_norm": 0.3879674971103668,
      "learning_rate": 4.078164004380458e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9419488787651062,
      "num_tokens": 148044261.0,
      "step": 19980
    },
    {
      "entropy": 2.082177019119263,
      "epoch": 0.29910345646662534,
      "grad_norm": 0.3713471293449402,
      "learning_rate": 4.0772326360210865e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9412368107587099,
      "num_tokens": 148117880.0,
      "step": 19990
    },
    {
      "entropy": 2.051739877462387,
      "epoch": 0.2992530830081294,
      "grad_norm": 0.4093238115310669,
      "learning_rate": 4.0763009041335656e-05,
      "loss": 0.2151,
      "mean_token_accuracy": 0.9473922595381736,
      "num_tokens": 148193531.0,
      "step": 20000
    },
    {
      "epoch": 0.2992530830081294,
      "eval_entropy": 2.0708330021142958,
      "eval_loss": 0.24427860975265503,
      "eval_mean_token_accuracy": 0.9453043068051338,
      "eval_num_tokens": 148193531.0,
      "eval_runtime": 552.4355,
      "eval_samples_per_second": 36.203,
      "eval_steps_per_second": 9.051,
      "step": 20000
    },
    {
      "entropy": 2.0759102158248424,
      "epoch": 0.29940270954963344,
      "grad_norm": 0.32652974128723145,
      "learning_rate": 4.0753688089333254e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9436774171888829,
      "num_tokens": 148266107.0,
      "step": 20010
    },
    {
      "entropy": 2.071843759715557,
      "epoch": 0.2995523360911375,
      "grad_norm": 0.3483104109764099,
      "learning_rate": 4.074436350635884e-05,
      "loss": 0.2304,
      "mean_token_accuracy": 0.942254251614213,
      "num_tokens": 148339488.0,
      "step": 20020
    },
    {
      "entropy": 2.0224280312657354,
      "epoch": 0.2997019626326416,
      "grad_norm": 0.3100298047065735,
      "learning_rate": 4.073503529456843e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9447232559323311,
      "num_tokens": 148412362.0,
      "step": 20030
    },
    {
      "entropy": 2.005318670719862,
      "epoch": 0.29985158917414567,
      "grad_norm": 0.3879857659339905,
      "learning_rate": 4.0725703456118864e-05,
      "loss": 0.2039,
      "mean_token_accuracy": 0.9510106228291988,
      "num_tokens": 148486270.0,
      "step": 20040
    },
    {
      "entropy": 1.9853283494710923,
      "epoch": 0.3000012157156497,
      "grad_norm": 0.379280686378479,
      "learning_rate": 4.071636799316783e-05,
      "loss": 0.2092,
      "mean_token_accuracy": 0.9501786291599273,
      "num_tokens": 148562525.0,
      "step": 20050
    },
    {
      "entropy": 2.0553229197859766,
      "epoch": 0.30015084225715377,
      "grad_norm": 0.42344072461128235,
      "learning_rate": 4.070702890787386e-05,
      "loss": 0.2452,
      "mean_token_accuracy": 0.9360755395144225,
      "num_tokens": 148633811.0,
      "step": 20060
    },
    {
      "entropy": 2.0442332707345487,
      "epoch": 0.30030046879865785,
      "grad_norm": 0.3089575171470642,
      "learning_rate": 4.0697686202396304e-05,
      "loss": 0.2058,
      "mean_token_accuracy": 0.9498985353857279,
      "num_tokens": 148709521.0,
      "step": 20070
    },
    {
      "entropy": 2.020933372527361,
      "epoch": 0.3004500953401619,
      "grad_norm": 0.3511551320552826,
      "learning_rate": 4.0688339878895374e-05,
      "loss": 0.1957,
      "mean_token_accuracy": 0.9477142490446567,
      "num_tokens": 148785250.0,
      "step": 20080
    },
    {
      "entropy": 2.031305019557476,
      "epoch": 0.300599721881666,
      "grad_norm": 0.3700384795665741,
      "learning_rate": 4.0678989939532094e-05,
      "loss": 0.2206,
      "mean_token_accuracy": 0.9467990510165691,
      "num_tokens": 148857675.0,
      "step": 20090
    },
    {
      "entropy": 2.014890494942665,
      "epoch": 0.30074934842317,
      "grad_norm": 0.3483615815639496,
      "learning_rate": 4.066963638646835e-05,
      "loss": 0.2246,
      "mean_token_accuracy": 0.945467672124505,
      "num_tokens": 148932164.0,
      "step": 20100
    },
    {
      "entropy": 2.0515431985259056,
      "epoch": 0.3008989749646741,
      "grad_norm": 0.40009334683418274,
      "learning_rate": 4.066027922186685e-05,
      "loss": 0.2006,
      "mean_token_accuracy": 0.9473388653248549,
      "num_tokens": 149007738.0,
      "step": 20110
    },
    {
      "entropy": 2.025045321881771,
      "epoch": 0.3010486015061782,
      "grad_norm": 0.30723336338996887,
      "learning_rate": 4.0650918447891124e-05,
      "loss": 0.1945,
      "mean_token_accuracy": 0.9517740681767464,
      "num_tokens": 149081891.0,
      "step": 20120
    },
    {
      "entropy": 2.048409187048674,
      "epoch": 0.30119822804768226,
      "grad_norm": 0.5110505223274231,
      "learning_rate": 4.0641554066705554e-05,
      "loss": 0.2276,
      "mean_token_accuracy": 0.9421701367944479,
      "num_tokens": 149154826.0,
      "step": 20130
    },
    {
      "entropy": 2.0346573255956173,
      "epoch": 0.3013478545891863,
      "grad_norm": 0.3433528244495392,
      "learning_rate": 4.0632186080475356e-05,
      "loss": 0.218,
      "mean_token_accuracy": 0.9457694068551064,
      "num_tokens": 149232107.0,
      "step": 20140
    },
    {
      "entropy": 2.10154742076993,
      "epoch": 0.30149748113069036,
      "grad_norm": 0.34369465708732605,
      "learning_rate": 4.062281449136657e-05,
      "loss": 0.2172,
      "mean_token_accuracy": 0.94480160176754,
      "num_tokens": 149305035.0,
      "step": 20150
    },
    {
      "entropy": 2.0551410019397736,
      "epoch": 0.30164710767219444,
      "grad_norm": 0.33466172218322754,
      "learning_rate": 4.0613439301546074e-05,
      "loss": 0.2121,
      "mean_token_accuracy": 0.9461729116737843,
      "num_tokens": 149379035.0,
      "step": 20160
    },
    {
      "entropy": 2.03267692103982,
      "epoch": 0.3017967342136985,
      "grad_norm": 0.3798556625843048,
      "learning_rate": 4.060406051318158e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9483131438493728,
      "num_tokens": 149454733.0,
      "step": 20170
    },
    {
      "entropy": 2.092066039144993,
      "epoch": 0.30194636075520254,
      "grad_norm": 0.35490837693214417,
      "learning_rate": 4.0594678128441636e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9424069199711085,
      "num_tokens": 149528065.0,
      "step": 20180
    },
    {
      "entropy": 2.0757917180657386,
      "epoch": 0.3020959872967066,
      "grad_norm": 0.3507733643054962,
      "learning_rate": 4.0585292149495614e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.9483824774622918,
      "num_tokens": 149602476.0,
      "step": 20190
    },
    {
      "entropy": 2.085632944852114,
      "epoch": 0.3022456138382107,
      "grad_norm": 0.37306100130081177,
      "learning_rate": 4.0575902578513716e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9442973997443914,
      "num_tokens": 149675515.0,
      "step": 20200
    },
    {
      "entropy": 2.065065673738718,
      "epoch": 0.30239524037971477,
      "grad_norm": 0.3211262822151184,
      "learning_rate": 4.056650941766698e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9449353717267513,
      "num_tokens": 149749815.0,
      "step": 20210
    },
    {
      "entropy": 2.0836403131484986,
      "epoch": 0.30254486692121885,
      "grad_norm": 0.3854275345802307,
      "learning_rate": 4.055711266912727e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9469335421919822,
      "num_tokens": 149823995.0,
      "step": 20220
    },
    {
      "entropy": 2.0359827056527138,
      "epoch": 0.30269449346272287,
      "grad_norm": 0.3149131238460541,
      "learning_rate": 4.054771233506728e-05,
      "loss": 0.2153,
      "mean_token_accuracy": 0.9444231435656547,
      "num_tokens": 149898263.0,
      "step": 20230
    },
    {
      "entropy": 2.0499155953526498,
      "epoch": 0.30284412000422695,
      "grad_norm": 0.29920250177383423,
      "learning_rate": 4.053830841766054e-05,
      "loss": 0.2296,
      "mean_token_accuracy": 0.9426768034696579,
      "num_tokens": 149969676.0,
      "step": 20240
    },
    {
      "entropy": 2.0736030772328378,
      "epoch": 0.302993746545731,
      "grad_norm": 0.3825022578239441,
      "learning_rate": 4.05289009190814e-05,
      "loss": 0.2466,
      "mean_token_accuracy": 0.9367811609059572,
      "num_tokens": 150042404.0,
      "step": 20250
    },
    {
      "entropy": 2.07252961024642,
      "epoch": 0.3031433730872351,
      "grad_norm": 0.28683727979660034,
      "learning_rate": 4.0519489841505035e-05,
      "loss": 0.2232,
      "mean_token_accuracy": 0.9401893369853497,
      "num_tokens": 150114646.0,
      "step": 20260
    },
    {
      "entropy": 2.0394650235772134,
      "epoch": 0.3032929996287391,
      "grad_norm": 0.33626025915145874,
      "learning_rate": 4.051007518710746e-05,
      "loss": 0.2076,
      "mean_token_accuracy": 0.9450729813426733,
      "num_tokens": 150189369.0,
      "step": 20270
    },
    {
      "entropy": 2.0168922170996666,
      "epoch": 0.3034426261702432,
      "grad_norm": 0.313623309135437,
      "learning_rate": 4.050065695806551e-05,
      "loss": 0.2073,
      "mean_token_accuracy": 0.9457651685923338,
      "num_tokens": 150263076.0,
      "step": 20280
    },
    {
      "entropy": 2.0252253018319606,
      "epoch": 0.3035922527117473,
      "grad_norm": 0.39041003584861755,
      "learning_rate": 4.049123515655685e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9477199658751487,
      "num_tokens": 150340176.0,
      "step": 20290
    },
    {
      "entropy": 2.0609344869852064,
      "epoch": 0.30374187925325136,
      "grad_norm": 0.337556928396225,
      "learning_rate": 4.048180978475997e-05,
      "loss": 0.226,
      "mean_token_accuracy": 0.9455450732260943,
      "num_tokens": 150414013.0,
      "step": 20300
    },
    {
      "entropy": 2.0581371992826463,
      "epoch": 0.3038915057947554,
      "grad_norm": 0.3807758092880249,
      "learning_rate": 4.047238084485418e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9460260260850191,
      "num_tokens": 150484684.0,
      "step": 20310
    },
    {
      "entropy": 2.0739156499505045,
      "epoch": 0.30404113233625946,
      "grad_norm": 0.41557514667510986,
      "learning_rate": 4.0462948339019604e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9429688390344382,
      "num_tokens": 150558641.0,
      "step": 20320
    },
    {
      "entropy": 2.0311658129096033,
      "epoch": 0.30419075887776353,
      "grad_norm": 0.3977046608924866,
      "learning_rate": 4.0453512269437224e-05,
      "loss": 0.2094,
      "mean_token_accuracy": 0.9484169911593199,
      "num_tokens": 150633445.0,
      "step": 20330
    },
    {
      "entropy": 2.034919075667858,
      "epoch": 0.3043403854192676,
      "grad_norm": 0.3646228611469269,
      "learning_rate": 4.0444072638288824e-05,
      "loss": 0.2128,
      "mean_token_accuracy": 0.9471397165209055,
      "num_tokens": 150708125.0,
      "step": 20340
    },
    {
      "entropy": 2.0485635086894036,
      "epoch": 0.30449001196077163,
      "grad_norm": 0.3745535612106323,
      "learning_rate": 4.0434629447757005e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.9450364284217357,
      "num_tokens": 150781804.0,
      "step": 20350
    },
    {
      "entropy": 2.025724655389786,
      "epoch": 0.3046396385022757,
      "grad_norm": 0.30733734369277954,
      "learning_rate": 4.0425182700025214e-05,
      "loss": 0.2227,
      "mean_token_accuracy": 0.9454188141971827,
      "num_tokens": 150858959.0,
      "step": 20360
    },
    {
      "entropy": 2.011425943672657,
      "epoch": 0.3047892650437798,
      "grad_norm": 0.32149365544319153,
      "learning_rate": 4.04157323972777e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9473905976861715,
      "num_tokens": 150932182.0,
      "step": 20370
    },
    {
      "entropy": 1.9974001124501228,
      "epoch": 0.30493889158528387,
      "grad_norm": 0.4118684232234955,
      "learning_rate": 4.040627854169955e-05,
      "loss": 0.2271,
      "mean_token_accuracy": 0.942804392799735,
      "num_tokens": 151005940.0,
      "step": 20380
    },
    {
      "entropy": 2.021271604299545,
      "epoch": 0.30508851812678794,
      "grad_norm": 0.42837023735046387,
      "learning_rate": 4.039682113547665e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9418392848223448,
      "num_tokens": 151082519.0,
      "step": 20390
    },
    {
      "entropy": 2.006600972265005,
      "epoch": 0.30523814466829197,
      "grad_norm": 0.48619765043258667,
      "learning_rate": 4.0387360180795716e-05,
      "loss": 0.2168,
      "mean_token_accuracy": 0.9463105499744415,
      "num_tokens": 151154859.0,
      "step": 20400
    },
    {
      "entropy": 2.032611910253763,
      "epoch": 0.30538777120979604,
      "grad_norm": 0.38922959566116333,
      "learning_rate": 4.037789567984431e-05,
      "loss": 0.2367,
      "mean_token_accuracy": 0.9382399510592222,
      "num_tokens": 151226945.0,
      "step": 20410
    },
    {
      "entropy": 2.02854560315609,
      "epoch": 0.3055373977513001,
      "grad_norm": 0.4069543778896332,
      "learning_rate": 4.036842763481078e-05,
      "loss": 0.2356,
      "mean_token_accuracy": 0.9406383708119392,
      "num_tokens": 151300508.0,
      "step": 20420
    },
    {
      "entropy": 2.0425567157566547,
      "epoch": 0.3056870242928042,
      "grad_norm": 0.33164307475090027,
      "learning_rate": 4.035895604788431e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9477099508047104,
      "num_tokens": 151376136.0,
      "step": 20430
    },
    {
      "entropy": 2.0062108054757117,
      "epoch": 0.3058366508343082,
      "grad_norm": 0.39279812574386597,
      "learning_rate": 4.0349480921254896e-05,
      "loss": 0.2286,
      "mean_token_accuracy": 0.9416333962231874,
      "num_tokens": 151449782.0,
      "step": 20440
    },
    {
      "entropy": 1.9681022197008133,
      "epoch": 0.3059862773758123,
      "grad_norm": 0.4712926745414734,
      "learning_rate": 4.034000225711335e-05,
      "loss": 0.2491,
      "mean_token_accuracy": 0.9407241377979517,
      "num_tokens": 151523301.0,
      "step": 20450
    },
    {
      "entropy": 2.005452035367489,
      "epoch": 0.3061359039173164,
      "grad_norm": 0.5050312280654907,
      "learning_rate": 4.033052005765132e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9431620825082063,
      "num_tokens": 151599663.0,
      "step": 20460
    },
    {
      "entropy": 2.000017238408327,
      "epoch": 0.30628553045882045,
      "grad_norm": 0.3637327551841736,
      "learning_rate": 4.032103432506126e-05,
      "loss": 0.2273,
      "mean_token_accuracy": 0.9459776900708675,
      "num_tokens": 151673857.0,
      "step": 20470
    },
    {
      "entropy": 2.019589538872242,
      "epoch": 0.3064351570003245,
      "grad_norm": 0.3020767569541931,
      "learning_rate": 4.0311545061536434e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9451740752905607,
      "num_tokens": 151748579.0,
      "step": 20480
    },
    {
      "entropy": 2.035581620782614,
      "epoch": 0.30658478354182855,
      "grad_norm": 0.3323059380054474,
      "learning_rate": 4.0302052269270924e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9468574225902557,
      "num_tokens": 151827662.0,
      "step": 20490
    },
    {
      "entropy": 2.0439609944820405,
      "epoch": 0.30673441008333263,
      "grad_norm": 0.44183799624443054,
      "learning_rate": 4.0292555950459646e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.942878994718194,
      "num_tokens": 151903455.0,
      "step": 20500
    },
    {
      "entropy": 1.9951113767921924,
      "epoch": 0.3068840366248367,
      "grad_norm": 0.337418794631958,
      "learning_rate": 4.02830561072983e-05,
      "loss": 0.2141,
      "mean_token_accuracy": 0.9458775170147419,
      "num_tokens": 151975342.0,
      "step": 20510
    },
    {
      "entropy": 2.0598106063902377,
      "epoch": 0.3070336631663408,
      "grad_norm": 0.30076271295547485,
      "learning_rate": 4.027355274198343e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9402170769870282,
      "num_tokens": 152047764.0,
      "step": 20520
    },
    {
      "entropy": 2.031250324100256,
      "epoch": 0.3071832897078448,
      "grad_norm": 0.3674869239330292,
      "learning_rate": 4.026404585671238e-05,
      "loss": 0.2024,
      "mean_token_accuracy": 0.9463290981948376,
      "num_tokens": 152124084.0,
      "step": 20530
    },
    {
      "entropy": 2.006423906236887,
      "epoch": 0.3073329162493489,
      "grad_norm": 0.3871377408504486,
      "learning_rate": 4.025453545368333e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.9453656852245331,
      "num_tokens": 152197737.0,
      "step": 20540
    },
    {
      "entropy": 2.0046853736042975,
      "epoch": 0.30748254279085296,
      "grad_norm": 0.3386818468570709,
      "learning_rate": 4.024502153509522e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9441407084465027,
      "num_tokens": 152272605.0,
      "step": 20550
    },
    {
      "entropy": 2.0315895207226276,
      "epoch": 0.30763216933235704,
      "grad_norm": 0.3717646598815918,
      "learning_rate": 4.023550410314784e-05,
      "loss": 0.2251,
      "mean_token_accuracy": 0.9426624353975057,
      "num_tokens": 152346719.0,
      "step": 20560
    },
    {
      "entropy": 2.0220543377101423,
      "epoch": 0.30778179587386106,
      "grad_norm": 0.32115015387535095,
      "learning_rate": 4.022598316004183e-05,
      "loss": 0.2107,
      "mean_token_accuracy": 0.9453809712082147,
      "num_tokens": 152421286.0,
      "step": 20570
    },
    {
      "entropy": 1.997485026717186,
      "epoch": 0.30793142241536514,
      "grad_norm": 0.38315844535827637,
      "learning_rate": 4.021645870797856e-05,
      "loss": 0.2177,
      "mean_token_accuracy": 0.9425423528999091,
      "num_tokens": 152496465.0,
      "step": 20580
    },
    {
      "entropy": 2.0124673798680304,
      "epoch": 0.3080810489568692,
      "grad_norm": 0.32588624954223633,
      "learning_rate": 4.0206930749160274e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9485005788505078,
      "num_tokens": 152570609.0,
      "step": 20590
    },
    {
      "entropy": 2.030595215409994,
      "epoch": 0.3082306754983733,
      "grad_norm": 0.33306217193603516,
      "learning_rate": 4.0197399285789993e-05,
      "loss": 0.2265,
      "mean_token_accuracy": 0.9438888899981975,
      "num_tokens": 152644928.0,
      "step": 20600
    },
    {
      "entropy": 2.0564027316868305,
      "epoch": 0.3083803020398773,
      "grad_norm": 0.3430580198764801,
      "learning_rate": 4.018786432007157e-05,
      "loss": 0.238,
      "mean_token_accuracy": 0.9415122382342815,
      "num_tokens": 152719993.0,
      "step": 20610
    },
    {
      "entropy": 2.0483881630003453,
      "epoch": 0.3085299285813814,
      "grad_norm": 0.3424883782863617,
      "learning_rate": 4.0178325854209664e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.9459471311420202,
      "num_tokens": 152790720.0,
      "step": 20620
    },
    {
      "entropy": 2.0609630599617956,
      "epoch": 0.3086795551228855,
      "grad_norm": 0.3543964922428131,
      "learning_rate": 4.0168783890409715e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9428050830960274,
      "num_tokens": 152863954.0,
      "step": 20630
    },
    {
      "entropy": 2.0334871239960193,
      "epoch": 0.30882918166438955,
      "grad_norm": 0.43066349625587463,
      "learning_rate": 4.0159238430878014e-05,
      "loss": 0.2175,
      "mean_token_accuracy": 0.9461889903992414,
      "num_tokens": 152936194.0,
      "step": 20640
    },
    {
      "entropy": 2.017429203540087,
      "epoch": 0.30897880820589363,
      "grad_norm": 0.37669119238853455,
      "learning_rate": 4.014968947782164e-05,
      "loss": 0.2335,
      "mean_token_accuracy": 0.9417244594544172,
      "num_tokens": 153010233.0,
      "step": 20650
    },
    {
      "entropy": 1.9555924907326698,
      "epoch": 0.30912843474739765,
      "grad_norm": 0.4077781140804291,
      "learning_rate": 4.014013703344848e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9468055468052625,
      "num_tokens": 153084880.0,
      "step": 20660
    },
    {
      "entropy": 2.0257609263062477,
      "epoch": 0.30927806128890173,
      "grad_norm": 0.35262539982795715,
      "learning_rate": 4.0130581099967214e-05,
      "loss": 0.2502,
      "mean_token_accuracy": 0.938075703009963,
      "num_tokens": 153156985.0,
      "step": 20670
    },
    {
      "entropy": 2.0132630944252012,
      "epoch": 0.3094276878304058,
      "grad_norm": 0.3171537220478058,
      "learning_rate": 4.012102167958737e-05,
      "loss": 0.2222,
      "mean_token_accuracy": 0.9471158277243376,
      "num_tokens": 153234396.0,
      "step": 20680
    },
    {
      "entropy": 2.02591550052166,
      "epoch": 0.3095773143719099,
      "grad_norm": 0.4488358199596405,
      "learning_rate": 4.011145877451923e-05,
      "loss": 0.2544,
      "mean_token_accuracy": 0.9364600818604231,
      "num_tokens": 153307037.0,
      "step": 20690
    },
    {
      "entropy": 2.031464084982872,
      "epoch": 0.3097269409134139,
      "grad_norm": 0.30517369508743286,
      "learning_rate": 4.010189238697393e-05,
      "loss": 0.2481,
      "mean_token_accuracy": 0.939165124297142,
      "num_tokens": 153380839.0,
      "step": 20700
    },
    {
      "entropy": 2.0346524991095065,
      "epoch": 0.309876567454918,
      "grad_norm": 0.36129963397979736,
      "learning_rate": 4.009232251916337e-05,
      "loss": 0.2191,
      "mean_token_accuracy": 0.9454477079212665,
      "num_tokens": 153454447.0,
      "step": 20710
    },
    {
      "entropy": 2.034428533911705,
      "epoch": 0.31002619399642206,
      "grad_norm": 0.3525198698043823,
      "learning_rate": 4.008274917330028e-05,
      "loss": 0.2176,
      "mean_token_accuracy": 0.9469003856182099,
      "num_tokens": 153527575.0,
      "step": 20720
    },
    {
      "entropy": 2.0697022952139377,
      "epoch": 0.31017582053792614,
      "grad_norm": 0.38144925236701965,
      "learning_rate": 4.0073172351598204e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9391985733062029,
      "num_tokens": 153605523.0,
      "step": 20730
    },
    {
      "entropy": 2.0354940690100194,
      "epoch": 0.31032544707943016,
      "grad_norm": 0.27484214305877686,
      "learning_rate": 4.0063592056271447e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9489546019583941,
      "num_tokens": 153679281.0,
      "step": 20740
    },
    {
      "entropy": 2.034306912124157,
      "epoch": 0.31047507362093424,
      "grad_norm": 0.3726237714290619,
      "learning_rate": 4.005400828953517e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9473531350493432,
      "num_tokens": 153753445.0,
      "step": 20750
    },
    {
      "entropy": 2.044531650096178,
      "epoch": 0.3106247001624383,
      "grad_norm": 0.4144675135612488,
      "learning_rate": 4.0044421053605276e-05,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9460084095597268,
      "num_tokens": 153831007.0,
      "step": 20760
    },
    {
      "entropy": 2.064971076697111,
      "epoch": 0.3107743267039424,
      "grad_norm": 0.4215957224369049,
      "learning_rate": 4.003483035069854e-05,
      "loss": 0.2294,
      "mean_token_accuracy": 0.9425821494311094,
      "num_tokens": 153904281.0,
      "step": 20770
    },
    {
      "entropy": 2.045664390921593,
      "epoch": 0.3109239532454464,
      "grad_norm": 0.38648131489753723,
      "learning_rate": 4.0025236183032486e-05,
      "loss": 0.2124,
      "mean_token_accuracy": 0.9458922579884529,
      "num_tokens": 153979582.0,
      "step": 20780
    },
    {
      "entropy": 2.0467068500816823,
      "epoch": 0.3110735797869505,
      "grad_norm": 0.40097951889038086,
      "learning_rate": 4.001563855282546e-05,
      "loss": 0.2157,
      "mean_token_accuracy": 0.9470040310174227,
      "num_tokens": 154051821.0,
      "step": 20790
    },
    {
      "entropy": 2.035109144449234,
      "epoch": 0.31122320632845457,
      "grad_norm": 0.34178274869918823,
      "learning_rate": 4.00060374622966e-05,
      "loss": 0.2127,
      "mean_token_accuracy": 0.9481329530477524,
      "num_tokens": 154129068.0,
      "step": 20800
    },
    {
      "entropy": 1.9955035082995891,
      "epoch": 0.31137283286995865,
      "grad_norm": 0.32662278413772583,
      "learning_rate": 3.999643291366585e-05,
      "loss": 0.213,
      "mean_token_accuracy": 0.9457138780504465,
      "num_tokens": 154205071.0,
      "step": 20810
    },
    {
      "entropy": 2.0531374648213387,
      "epoch": 0.3115224594114627,
      "grad_norm": 0.3054129183292389,
      "learning_rate": 3.9986824909153965e-05,
      "loss": 0.2274,
      "mean_token_accuracy": 0.9421583835035563,
      "num_tokens": 154280973.0,
      "step": 20820
    },
    {
      "entropy": 2.0693315140902997,
      "epoch": 0.31167208595296675,
      "grad_norm": 0.5104321837425232,
      "learning_rate": 3.997721345098247e-05,
      "loss": 0.2379,
      "mean_token_accuracy": 0.9406526647508144,
      "num_tokens": 154356846.0,
      "step": 20830
    },
    {
      "entropy": 2.023317018151283,
      "epoch": 0.3118217124944708,
      "grad_norm": 0.28538450598716736,
      "learning_rate": 3.996759854137371e-05,
      "loss": 0.209,
      "mean_token_accuracy": 0.9469563696533442,
      "num_tokens": 154432101.0,
      "step": 20840
    },
    {
      "entropy": 1.97511872574687,
      "epoch": 0.3119713390359749,
      "grad_norm": 0.3441958427429199,
      "learning_rate": 3.9957980182550824e-05,
      "loss": 0.2085,
      "mean_token_accuracy": 0.9458885002881289,
      "num_tokens": 154506730.0,
      "step": 20850
    },
    {
      "entropy": 2.0496716529130934,
      "epoch": 0.312120965577479,
      "grad_norm": 0.3860388994216919,
      "learning_rate": 3.994835837673775e-05,
      "loss": 0.2116,
      "mean_token_accuracy": 0.9434463396668434,
      "num_tokens": 154580108.0,
      "step": 20860
    },
    {
      "entropy": 2.049126545339823,
      "epoch": 0.312270592118983,
      "grad_norm": 0.3473508358001709,
      "learning_rate": 3.9938733126159216e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.9452503927052021,
      "num_tokens": 154651675.0,
      "step": 20870
    },
    {
      "entropy": 2.0800929561257364,
      "epoch": 0.3124202186604871,
      "grad_norm": 0.34802186489105225,
      "learning_rate": 3.992910443304076e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9397499542683363,
      "num_tokens": 154725577.0,
      "step": 20880
    },
    {
      "entropy": 2.003300468623638,
      "epoch": 0.31256984520199116,
      "grad_norm": 0.3554462790489197,
      "learning_rate": 3.991947229960869e-05,
      "loss": 0.2111,
      "mean_token_accuracy": 0.9496726907789708,
      "num_tokens": 154802057.0,
      "step": 20890
    },
    {
      "entropy": 2.036792879551649,
      "epoch": 0.31271947174349524,
      "grad_norm": 0.3895892798900604,
      "learning_rate": 3.9909836728090155e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9414533726871014,
      "num_tokens": 154874260.0,
      "step": 20900
    },
    {
      "entropy": 2.0275000773370264,
      "epoch": 0.31286909828499926,
      "grad_norm": 0.41381338238716125,
      "learning_rate": 3.990019772071304e-05,
      "loss": 0.21,
      "mean_token_accuracy": 0.9475589234381914,
      "num_tokens": 154952318.0,
      "step": 20910
    },
    {
      "entropy": 2.064033020287752,
      "epoch": 0.31301872482650334,
      "grad_norm": 0.42414525151252747,
      "learning_rate": 3.989055527970607e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9455028463155031,
      "num_tokens": 155025251.0,
      "step": 20920
    },
    {
      "entropy": 2.0484140418469905,
      "epoch": 0.3131683513680074,
      "grad_norm": 0.3718685209751129,
      "learning_rate": 3.988090940729875e-05,
      "loss": 0.2161,
      "mean_token_accuracy": 0.9460485558956861,
      "num_tokens": 155096279.0,
      "step": 20930
    },
    {
      "entropy": 2.0486058235168456,
      "epoch": 0.3133179779095115,
      "grad_norm": 0.3463824391365051,
      "learning_rate": 3.987126010572137e-05,
      "loss": 0.2266,
      "mean_token_accuracy": 0.9441498599946498,
      "num_tokens": 155170232.0,
      "step": 20940
    },
    {
      "entropy": 2.0465155705809592,
      "epoch": 0.31346760445101557,
      "grad_norm": 0.3537174165248871,
      "learning_rate": 3.986160737720503e-05,
      "loss": 0.2092,
      "mean_token_accuracy": 0.9469104565680027,
      "num_tokens": 155244098.0,
      "step": 20950
    },
    {
      "entropy": 2.06198565736413,
      "epoch": 0.3136172309925196,
      "grad_norm": 0.37638577818870544,
      "learning_rate": 3.98519512239816e-05,
      "loss": 0.2041,
      "mean_token_accuracy": 0.9478725403547287,
      "num_tokens": 155319329.0,
      "step": 20960
    },
    {
      "entropy": 1.996738040447235,
      "epoch": 0.31376685753402367,
      "grad_norm": 0.3048970103263855,
      "learning_rate": 3.9842291648283766e-05,
      "loss": 0.2026,
      "mean_token_accuracy": 0.9497688509523868,
      "num_tokens": 155392002.0,
      "step": 20970
    },
    {
      "entropy": 2.0046715155243873,
      "epoch": 0.31391648407552775,
      "grad_norm": 0.37890535593032837,
      "learning_rate": 3.983262865234499e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9495746187865735,
      "num_tokens": 155466638.0,
      "step": 20980
    },
    {
      "entropy": 2.021643092483282,
      "epoch": 0.3140661106170318,
      "grad_norm": 0.2615390717983246,
      "learning_rate": 3.982296223839953e-05,
      "loss": 0.2057,
      "mean_token_accuracy": 0.9475334361195564,
      "num_tokens": 155542727.0,
      "step": 20990
    },
    {
      "entropy": 2.0476947478950023,
      "epoch": 0.31421573715853585,
      "grad_norm": 0.33889931440353394,
      "learning_rate": 3.9813292408682425e-05,
      "loss": 0.2341,
      "mean_token_accuracy": 0.9452584467828273,
      "num_tokens": 155616047.0,
      "step": 21000
    },
    {
      "epoch": 0.31421573715853585,
      "eval_entropy": 2.049203820157051,
      "eval_loss": 0.24272094666957855,
      "eval_mean_token_accuracy": 0.9455852692008019,
      "eval_num_tokens": 155616047.0,
      "eval_runtime": 551.3391,
      "eval_samples_per_second": 36.275,
      "eval_steps_per_second": 9.069,
      "step": 21000
    },
    {
      "entropy": 2.067938378453255,
      "epoch": 0.3143653637000399,
      "grad_norm": 0.3299214839935303,
      "learning_rate": 3.980361916542952e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9460555110126734,
      "num_tokens": 155686270.0,
      "step": 21010
    },
    {
      "entropy": 2.040417101979256,
      "epoch": 0.314514990241544,
      "grad_norm": 0.31013473868370056,
      "learning_rate": 3.979394251087743e-05,
      "loss": 0.211,
      "mean_token_accuracy": 0.9493416756391525,
      "num_tokens": 155761211.0,
      "step": 21020
    },
    {
      "entropy": 2.086323032528162,
      "epoch": 0.3146646167830481,
      "grad_norm": 0.3625803589820862,
      "learning_rate": 3.9784262447263585e-05,
      "loss": 0.2361,
      "mean_token_accuracy": 0.943095950037241,
      "num_tokens": 155835527.0,
      "step": 21030
    },
    {
      "entropy": 2.0910055972635746,
      "epoch": 0.3148142433245521,
      "grad_norm": 0.3228873312473297,
      "learning_rate": 3.977457897682618e-05,
      "loss": 0.2059,
      "mean_token_accuracy": 0.9497723091393709,
      "num_tokens": 155911674.0,
      "step": 21040
    },
    {
      "entropy": 2.054501846432686,
      "epoch": 0.3149638698660562,
      "grad_norm": 0.3460301160812378,
      "learning_rate": 3.9764892101804206e-05,
      "loss": 0.2086,
      "mean_token_accuracy": 0.9459057450294495,
      "num_tokens": 155986718.0,
      "step": 21050
    },
    {
      "entropy": 2.05505652949214,
      "epoch": 0.31511349640756026,
      "grad_norm": 0.3822898268699646,
      "learning_rate": 3.975520182443744e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9430285263806581,
      "num_tokens": 156059885.0,
      "step": 21060
    },
    {
      "entropy": 2.033085487037897,
      "epoch": 0.31526312294906433,
      "grad_norm": 0.31928619742393494,
      "learning_rate": 3.9745508146966447e-05,
      "loss": 0.2118,
      "mean_token_accuracy": 0.943522747978568,
      "num_tokens": 156133793.0,
      "step": 21070
    },
    {
      "entropy": 1.9927950337529183,
      "epoch": 0.3154127494905684,
      "grad_norm": 0.4609970450401306,
      "learning_rate": 3.9735811071632584e-05,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9451512970030308,
      "num_tokens": 156210434.0,
      "step": 21080
    },
    {
      "entropy": 2.0285411842167376,
      "epoch": 0.31556237603207243,
      "grad_norm": 0.3782942295074463,
      "learning_rate": 3.9726110600677975e-05,
      "loss": 0.2261,
      "mean_token_accuracy": 0.9455464221537113,
      "num_tokens": 156283713.0,
      "step": 21090
    },
    {
      "entropy": 2.0534337289631366,
      "epoch": 0.3157120025735765,
      "grad_norm": 0.2939006984233856,
      "learning_rate": 3.971640673634555e-05,
      "loss": 0.2327,
      "mean_token_accuracy": 0.941748633980751,
      "num_tokens": 156357174.0,
      "step": 21100
    },
    {
      "entropy": 2.072802795469761,
      "epoch": 0.3158616291150806,
      "grad_norm": 0.2700656056404114,
      "learning_rate": 3.970669948087901e-05,
      "loss": 0.2231,
      "mean_token_accuracy": 0.9408281002193689,
      "num_tokens": 156429886.0,
      "step": 21110
    },
    {
      "entropy": 2.030059576779604,
      "epoch": 0.31601125565658467,
      "grad_norm": 0.3797900378704071,
      "learning_rate": 3.9696988836522844e-05,
      "loss": 0.2119,
      "mean_token_accuracy": 0.9481119718402624,
      "num_tokens": 156503894.0,
      "step": 21120
    },
    {
      "entropy": 2.026000624895096,
      "epoch": 0.3161608821980887,
      "grad_norm": 0.3602733910083771,
      "learning_rate": 3.968727480552234e-05,
      "loss": 0.2291,
      "mean_token_accuracy": 0.9416314449161292,
      "num_tokens": 156578105.0,
      "step": 21130
    },
    {
      "entropy": 2.020570170879364,
      "epoch": 0.31631050873959277,
      "grad_norm": 0.3036384880542755,
      "learning_rate": 3.9677557390123545e-05,
      "loss": 0.2147,
      "mean_token_accuracy": 0.944324952736497,
      "num_tokens": 156652887.0,
      "step": 21140
    },
    {
      "entropy": 2.026704975217581,
      "epoch": 0.31646013528109684,
      "grad_norm": 0.31078073382377625,
      "learning_rate": 3.966783659257329e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9464764360338449,
      "num_tokens": 156726684.0,
      "step": 21150
    },
    {
      "entropy": 2.0353177413344383,
      "epoch": 0.3166097618226009,
      "grad_norm": 0.38851556181907654,
      "learning_rate": 3.965811241511921e-05,
      "loss": 0.2181,
      "mean_token_accuracy": 0.9444101713597775,
      "num_tokens": 156798958.0,
      "step": 21160
    },
    {
      "entropy": 2.0476369075477123,
      "epoch": 0.31675938836410494,
      "grad_norm": 0.334753155708313,
      "learning_rate": 3.964838486000969e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9446969959884882,
      "num_tokens": 156871548.0,
      "step": 21170
    },
    {
      "entropy": 2.038842688500881,
      "epoch": 0.316909014905609,
      "grad_norm": 0.36391958594322205,
      "learning_rate": 3.963865392949394e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9459203172475099,
      "num_tokens": 156945769.0,
      "step": 21180
    },
    {
      "entropy": 2.002093390375376,
      "epoch": 0.3170586414471131,
      "grad_norm": 0.42484068870544434,
      "learning_rate": 3.9628919625821905e-05,
      "loss": 0.191,
      "mean_token_accuracy": 0.9506699938327074,
      "num_tokens": 157021140.0,
      "step": 21190
    },
    {
      "entropy": 1.9908698186278344,
      "epoch": 0.3172082679886172,
      "grad_norm": 0.36248889565467834,
      "learning_rate": 3.9619181951244335e-05,
      "loss": 0.2099,
      "mean_token_accuracy": 0.9491431560367346,
      "num_tokens": 157093904.0,
      "step": 21200
    },
    {
      "entropy": 1.9868118532001973,
      "epoch": 0.3173578945301212,
      "grad_norm": 0.33576852083206177,
      "learning_rate": 3.960944090801274e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9462547600269318,
      "num_tokens": 157167987.0,
      "step": 21210
    },
    {
      "entropy": 2.0365115188062193,
      "epoch": 0.3175075210716253,
      "grad_norm": 0.35511136054992676,
      "learning_rate": 3.959969649837945e-05,
      "loss": 0.219,
      "mean_token_accuracy": 0.9461350623518229,
      "num_tokens": 157242746.0,
      "step": 21220
    },
    {
      "entropy": 2.0402232773602007,
      "epoch": 0.31765714761312935,
      "grad_norm": 0.36837494373321533,
      "learning_rate": 3.958994872459752e-05,
      "loss": 0.2203,
      "mean_token_accuracy": 0.9428602252155542,
      "num_tokens": 157317331.0,
      "step": 21230
    },
    {
      "entropy": 1.9857642568647862,
      "epoch": 0.31780677415463343,
      "grad_norm": 0.3189406096935272,
      "learning_rate": 3.958019758892083e-05,
      "loss": 0.2179,
      "mean_token_accuracy": 0.9464772686362266,
      "num_tokens": 157393123.0,
      "step": 21240
    },
    {
      "entropy": 1.992910234630108,
      "epoch": 0.3179564006961375,
      "grad_norm": 0.3239775598049164,
      "learning_rate": 3.9570443093604e-05,
      "loss": 0.215,
      "mean_token_accuracy": 0.9451756544411183,
      "num_tokens": 157465751.0,
      "step": 21250
    },
    {
      "entropy": 2.0002670891582968,
      "epoch": 0.31810602723764153,
      "grad_norm": 0.3075005114078522,
      "learning_rate": 3.956068524090245e-05,
      "loss": 0.2208,
      "mean_token_accuracy": 0.9445587109774352,
      "num_tokens": 157544351.0,
      "step": 21260
    },
    {
      "entropy": 1.9972836568951606,
      "epoch": 0.3182556537791456,
      "grad_norm": 0.3258792757987976,
      "learning_rate": 3.955092403307237e-05,
      "loss": 0.2091,
      "mean_token_accuracy": 0.9458262383937835,
      "num_tokens": 157620844.0,
      "step": 21270
    },
    {
      "entropy": 2.0234507739543917,
      "epoch": 0.3184052803206497,
      "grad_norm": 0.3724513053894043,
      "learning_rate": 3.954115947237072e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9418605446815491,
      "num_tokens": 157695951.0,
      "step": 21280
    },
    {
      "entropy": 2.0016843006014824,
      "epoch": 0.31855490686215376,
      "grad_norm": 0.4644080102443695,
      "learning_rate": 3.9531391561055245e-05,
      "loss": 0.2288,
      "mean_token_accuracy": 0.9438267134130001,
      "num_tokens": 157773275.0,
      "step": 21290
    },
    {
      "entropy": 2.0175721898674963,
      "epoch": 0.3187045334036578,
      "grad_norm": 0.32737523317337036,
      "learning_rate": 3.952162030138445e-05,
      "loss": 0.2082,
      "mean_token_accuracy": 0.9469501547515392,
      "num_tokens": 157848161.0,
      "step": 21300
    },
    {
      "entropy": 1.990839173644781,
      "epoch": 0.31885415994516186,
      "grad_norm": 0.3597821891307831,
      "learning_rate": 3.9511845695617636e-05,
      "loss": 0.2233,
      "mean_token_accuracy": 0.9433918241411448,
      "num_tokens": 157923182.0,
      "step": 21310
    },
    {
      "entropy": 2.010980723053217,
      "epoch": 0.31900378648666594,
      "grad_norm": 0.39587196707725525,
      "learning_rate": 3.9502067746014855e-05,
      "loss": 0.2429,
      "mean_token_accuracy": 0.9394553314894438,
      "num_tokens": 157995680.0,
      "step": 21320
    },
    {
      "entropy": 2.0357247851788998,
      "epoch": 0.31915341302817,
      "grad_norm": 0.32736778259277344,
      "learning_rate": 3.9492286454836956e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9440926939249039,
      "num_tokens": 158070465.0,
      "step": 21330
    },
    {
      "entropy": 1.97601200863719,
      "epoch": 0.31930303956967404,
      "grad_norm": 0.28568172454833984,
      "learning_rate": 3.948250182434554e-05,
      "loss": 0.2114,
      "mean_token_accuracy": 0.9444621555507183,
      "num_tokens": 158145174.0,
      "step": 21340
    },
    {
      "entropy": 1.9470690689980983,
      "epoch": 0.3194526661111781,
      "grad_norm": 0.3590262532234192,
      "learning_rate": 3.947271385680299e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9426858507096767,
      "num_tokens": 158217250.0,
      "step": 21350
    },
    {
      "entropy": 1.9611158460378646,
      "epoch": 0.3196022926526822,
      "grad_norm": 0.3179415166378021,
      "learning_rate": 3.946292255447245e-05,
      "loss": 0.2153,
      "mean_token_accuracy": 0.9469668809324503,
      "num_tokens": 158292197.0,
      "step": 21360
    },
    {
      "entropy": 1.9819176979362965,
      "epoch": 0.3197519191941863,
      "grad_norm": 0.3512411415576935,
      "learning_rate": 3.9453127919617854e-05,
      "loss": 0.2047,
      "mean_token_accuracy": 0.9462457746267319,
      "num_tokens": 158368081.0,
      "step": 21370
    },
    {
      "entropy": 2.0229830175638197,
      "epoch": 0.31990154573569035,
      "grad_norm": 0.38249602913856506,
      "learning_rate": 3.9443329954503886e-05,
      "loss": 0.2342,
      "mean_token_accuracy": 0.9430902604013681,
      "num_tokens": 158442502.0,
      "step": 21380
    },
    {
      "entropy": 2.032430177181959,
      "epoch": 0.3200511722771944,
      "grad_norm": 0.3795258104801178,
      "learning_rate": 3.943352866139602e-05,
      "loss": 0.2289,
      "mean_token_accuracy": 0.9447651542723179,
      "num_tokens": 158514440.0,
      "step": 21390
    },
    {
      "entropy": 2.023717015981674,
      "epoch": 0.32020079881869845,
      "grad_norm": 0.30938658118247986,
      "learning_rate": 3.9423724042560484e-05,
      "loss": 0.2156,
      "mean_token_accuracy": 0.9452714815735817,
      "num_tokens": 158588338.0,
      "step": 21400
    },
    {
      "entropy": 2.0178687386214733,
      "epoch": 0.32035042536020253,
      "grad_norm": 0.42074257135391235,
      "learning_rate": 3.9413916100264275e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9438644349575043,
      "num_tokens": 158660937.0,
      "step": 21410
    },
    {
      "entropy": 1.9600103966891766,
      "epoch": 0.3205000519017066,
      "grad_norm": 0.4060099422931671,
      "learning_rate": 3.940410483677518e-05,
      "loss": 0.2169,
      "mean_token_accuracy": 0.9456734035164118,
      "num_tokens": 158735732.0,
      "step": 21420
    },
    {
      "entropy": 1.9779482908546924,
      "epoch": 0.32064967844321063,
      "grad_norm": 0.3536320626735687,
      "learning_rate": 3.939429025436172e-05,
      "loss": 0.2032,
      "mean_token_accuracy": 0.9473308224231005,
      "num_tokens": 158811649.0,
      "step": 21430
    },
    {
      "entropy": 1.9725464567542077,
      "epoch": 0.3207993049847147,
      "grad_norm": 0.28914687037467957,
      "learning_rate": 3.9384472355293207e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9494631376117468,
      "num_tokens": 158887313.0,
      "step": 21440
    },
    {
      "entropy": 1.982947215437889,
      "epoch": 0.3209489315262188,
      "grad_norm": 0.4032208323478699,
      "learning_rate": 3.937465114183972e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9434507060796022,
      "num_tokens": 158962765.0,
      "step": 21450
    },
    {
      "entropy": 1.987280560284853,
      "epoch": 0.32109855806772286,
      "grad_norm": 0.3512982130050659,
      "learning_rate": 3.9364826616272086e-05,
      "loss": 0.2204,
      "mean_token_accuracy": 0.9463640470057726,
      "num_tokens": 159034200.0,
      "step": 21460
    },
    {
      "entropy": 1.9863001890480518,
      "epoch": 0.3212481846092269,
      "grad_norm": 0.46394428610801697,
      "learning_rate": 3.935499878086192e-05,
      "loss": 0.2015,
      "mean_token_accuracy": 0.9495463334023952,
      "num_tokens": 159109359.0,
      "step": 21470
    },
    {
      "entropy": 2.0119037181138992,
      "epoch": 0.32139781115073096,
      "grad_norm": 0.30585619807243347,
      "learning_rate": 3.9345167637881594e-05,
      "loss": 0.2197,
      "mean_token_accuracy": 0.9402464888989925,
      "num_tokens": 159185335.0,
      "step": 21480
    },
    {
      "entropy": 2.0103465363383295,
      "epoch": 0.32154743769223504,
      "grad_norm": 0.4745115637779236,
      "learning_rate": 3.933533318960423e-05,
      "loss": 0.2042,
      "mean_token_accuracy": 0.948231453448534,
      "num_tokens": 159261565.0,
      "step": 21490
    },
    {
      "entropy": 2.0037326201796533,
      "epoch": 0.3216970642337391,
      "grad_norm": 0.32413962483406067,
      "learning_rate": 3.9325495438303746e-05,
      "loss": 0.2279,
      "mean_token_accuracy": 0.940834591165185,
      "num_tokens": 159335105.0,
      "step": 21500
    },
    {
      "entropy": 1.9720140799880028,
      "epoch": 0.3218466907752432,
      "grad_norm": 0.3732118010520935,
      "learning_rate": 3.931565438625479e-05,
      "loss": 0.205,
      "mean_token_accuracy": 0.9474518183618784,
      "num_tokens": 159407885.0,
      "step": 21510
    },
    {
      "entropy": 1.9738122932612896,
      "epoch": 0.3219963173167472,
      "grad_norm": 0.3450663387775421,
      "learning_rate": 3.9305810035732796e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9433425858616828,
      "num_tokens": 159481590.0,
      "step": 21520
    },
    {
      "entropy": 1.9817461341619491,
      "epoch": 0.3221459438582513,
      "grad_norm": 0.33707523345947266,
      "learning_rate": 3.929596238901395e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.945607977733016,
      "num_tokens": 159552873.0,
      "step": 21530
    },
    {
      "entropy": 1.9840748824179173,
      "epoch": 0.32229557039975537,
      "grad_norm": 0.3030901551246643,
      "learning_rate": 3.9286111448375196e-05,
      "loss": 0.2362,
      "mean_token_accuracy": 0.9407924819737673,
      "num_tokens": 159628268.0,
      "step": 21540
    },
    {
      "entropy": 2.003737914562225,
      "epoch": 0.32244519694125945,
      "grad_norm": 0.36346423625946045,
      "learning_rate": 3.9276257216094264e-05,
      "loss": 0.2254,
      "mean_token_accuracy": 0.9422803707420826,
      "num_tokens": 159702337.0,
      "step": 21550
    },
    {
      "entropy": 1.9975286461412907,
      "epoch": 0.32259482348276347,
      "grad_norm": 0.37526774406433105,
      "learning_rate": 3.9266399694449616e-05,
      "loss": 0.2117,
      "mean_token_accuracy": 0.9459686815738678,
      "num_tokens": 159776289.0,
      "step": 21560
    },
    {
      "entropy": 2.0102029852569103,
      "epoch": 0.32274445002426755,
      "grad_norm": 0.3834324777126312,
      "learning_rate": 3.925653888572048e-05,
      "loss": 0.205,
      "mean_token_accuracy": 0.9490348465740681,
      "num_tokens": 159846946.0,
      "step": 21570
    },
    {
      "entropy": 1.9851542390882968,
      "epoch": 0.3228940765657716,
      "grad_norm": 0.3518499433994293,
      "learning_rate": 3.924667479218687e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9484173983335495,
      "num_tokens": 159920715.0,
      "step": 21580
    },
    {
      "entropy": 1.9847568929195405,
      "epoch": 0.3230437031072757,
      "grad_norm": 0.39872080087661743,
      "learning_rate": 3.923680741612952e-05,
      "loss": 0.2275,
      "mean_token_accuracy": 0.944013575464487,
      "num_tokens": 159995436.0,
      "step": 21590
    },
    {
      "entropy": 2.007246517390013,
      "epoch": 0.3231933296487797,
      "grad_norm": 0.2699209153652191,
      "learning_rate": 3.922693675982996e-05,
      "loss": 0.2245,
      "mean_token_accuracy": 0.9413744397461414,
      "num_tokens": 160072182.0,
      "step": 21600
    },
    {
      "entropy": 2.0074453584849836,
      "epoch": 0.3233429561902838,
      "grad_norm": 0.3509252667427063,
      "learning_rate": 3.921706282557045e-05,
      "loss": 0.2171,
      "mean_token_accuracy": 0.9467690706253051,
      "num_tokens": 160146983.0,
      "step": 21610
    },
    {
      "entropy": 2.077109782397747,
      "epoch": 0.3234925827317879,
      "grad_norm": 0.4871334731578827,
      "learning_rate": 3.9207185615634014e-05,
      "loss": 0.23,
      "mean_token_accuracy": 0.9448326826095581,
      "num_tokens": 160219243.0,
      "step": 21620
    },
    {
      "entropy": 2.0690233670175076,
      "epoch": 0.32364220927329196,
      "grad_norm": 0.39562880992889404,
      "learning_rate": 3.919730513230445e-05,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9388488363474607,
      "num_tokens": 160291335.0,
      "step": 21630
    },
    {
      "entropy": 1.9964601308107377,
      "epoch": 0.323791835814796,
      "grad_norm": 0.35720333456993103,
      "learning_rate": 3.9187421377866305e-05,
      "loss": 0.2031,
      "mean_token_accuracy": 0.9470710754394531,
      "num_tokens": 160363896.0,
      "step": 21640
    },
    {
      "entropy": 1.9931552797555923,
      "epoch": 0.32394146235630006,
      "grad_norm": 0.35890018939971924,
      "learning_rate": 3.917753435460487e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9434368804097175,
      "num_tokens": 160438144.0,
      "step": 21650
    },
    {
      "entropy": 2.0072352454066276,
      "epoch": 0.32409108889780414,
      "grad_norm": 0.3876381516456604,
      "learning_rate": 3.9167644064806194e-05,
      "loss": 0.2118,
      "mean_token_accuracy": 0.9457936495542526,
      "num_tokens": 160514775.0,
      "step": 21660
    },
    {
      "entropy": 2.011367876827717,
      "epoch": 0.3242407154393082,
      "grad_norm": 0.33291056752204895,
      "learning_rate": 3.915775051075711e-05,
      "loss": 0.2072,
      "mean_token_accuracy": 0.9475776456296444,
      "num_tokens": 160587499.0,
      "step": 21670
    },
    {
      "entropy": 1.987958462536335,
      "epoch": 0.3243903419808123,
      "grad_norm": 0.3264543116092682,
      "learning_rate": 3.9147853694745154e-05,
      "loss": 0.2368,
      "mean_token_accuracy": 0.9422347642481327,
      "num_tokens": 160658756.0,
      "step": 21680
    },
    {
      "entropy": 1.96802050024271,
      "epoch": 0.3245399685223163,
      "grad_norm": 0.4058969020843506,
      "learning_rate": 3.913795361905867e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9491988968104124,
      "num_tokens": 160736808.0,
      "step": 21690
    },
    {
      "entropy": 1.9849737845361233,
      "epoch": 0.3246895950638204,
      "grad_norm": 0.30971574783325195,
      "learning_rate": 3.912805028598673e-05,
      "loss": 0.2178,
      "mean_token_accuracy": 0.9447855114936828,
      "num_tokens": 160811990.0,
      "step": 21700
    },
    {
      "entropy": 1.975720302760601,
      "epoch": 0.32483922160532447,
      "grad_norm": 0.3507390022277832,
      "learning_rate": 3.911814369781914e-05,
      "loss": 0.2039,
      "mean_token_accuracy": 0.9463858529925346,
      "num_tokens": 160887607.0,
      "step": 21710
    },
    {
      "entropy": 2.0132723189890385,
      "epoch": 0.32498884814682855,
      "grad_norm": 0.3478752374649048,
      "learning_rate": 3.910823385684651e-05,
      "loss": 0.2334,
      "mean_token_accuracy": 0.9430777702480555,
      "num_tokens": 160963013.0,
      "step": 21720
    },
    {
      "entropy": 2.0074999004602434,
      "epoch": 0.32513847468833257,
      "grad_norm": 0.4104067385196686,
      "learning_rate": 3.909832076536015e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9465911034494638,
      "num_tokens": 161036093.0,
      "step": 21730
    },
    {
      "entropy": 1.9822812102735043,
      "epoch": 0.32528810122983665,
      "grad_norm": 0.5129075646400452,
      "learning_rate": 3.908840442565214e-05,
      "loss": 0.2392,
      "mean_token_accuracy": 0.9407647550106049,
      "num_tokens": 161108188.0,
      "step": 21740
    },
    {
      "entropy": 1.976319255679846,
      "epoch": 0.3254377277713407,
      "grad_norm": 0.3484298884868622,
      "learning_rate": 3.907848484001533e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9454923130571842,
      "num_tokens": 161182163.0,
      "step": 21750
    },
    {
      "entropy": 1.936514262855053,
      "epoch": 0.3255873543128448,
      "grad_norm": 0.3309216797351837,
      "learning_rate": 3.9068562010743295e-05,
      "loss": 0.2194,
      "mean_token_accuracy": 0.9472241543233395,
      "num_tokens": 161256575.0,
      "step": 21760
    },
    {
      "entropy": 1.963012869656086,
      "epoch": 0.3257369808543488,
      "grad_norm": 0.344516396522522,
      "learning_rate": 3.9058635940130374e-05,
      "loss": 0.2077,
      "mean_token_accuracy": 0.9462473016232252,
      "num_tokens": 161332014.0,
      "step": 21770
    },
    {
      "entropy": 1.9798145152628421,
      "epoch": 0.3258866073958529,
      "grad_norm": 0.41650813817977905,
      "learning_rate": 3.904870663047165e-05,
      "loss": 0.2152,
      "mean_token_accuracy": 0.9455726474523545,
      "num_tokens": 161406741.0,
      "step": 21780
    },
    {
      "entropy": 1.9715054392814637,
      "epoch": 0.326036233937357,
      "grad_norm": 0.3356781303882599,
      "learning_rate": 3.903877408406295e-05,
      "loss": 0.22,
      "mean_token_accuracy": 0.947141955792904,
      "num_tokens": 161480601.0,
      "step": 21790
    },
    {
      "entropy": 2.0067955769598482,
      "epoch": 0.32618586047886106,
      "grad_norm": 0.3800460994243622,
      "learning_rate": 3.902883830320086e-05,
      "loss": 0.2204,
      "mean_token_accuracy": 0.9450761683285236,
      "num_tokens": 161552332.0,
      "step": 21800
    },
    {
      "entropy": 1.9599581368267536,
      "epoch": 0.32633548702036513,
      "grad_norm": 0.33722949028015137,
      "learning_rate": 3.901889929018272e-05,
      "loss": 0.2123,
      "mean_token_accuracy": 0.9493962861597538,
      "num_tokens": 161624610.0,
      "step": 21810
    },
    {
      "entropy": 1.95061514377594,
      "epoch": 0.32648511356186916,
      "grad_norm": 0.2997848689556122,
      "learning_rate": 3.900895704730658e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9451621443033218,
      "num_tokens": 161698215.0,
      "step": 21820
    },
    {
      "entropy": 1.9601900659501552,
      "epoch": 0.32663474010337323,
      "grad_norm": 0.3582102656364441,
      "learning_rate": 3.8999011576871295e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.9441618472337723,
      "num_tokens": 161771814.0,
      "step": 21830
    },
    {
      "entropy": 1.9685291223227979,
      "epoch": 0.3267843666448773,
      "grad_norm": 0.3601633608341217,
      "learning_rate": 3.898906288117642e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9425498012453317,
      "num_tokens": 161847655.0,
      "step": 21840
    },
    {
      "entropy": 2.0029328264296056,
      "epoch": 0.3269339931863814,
      "grad_norm": 0.4203227758407593,
      "learning_rate": 3.8979110962522246e-05,
      "loss": 0.2182,
      "mean_token_accuracy": 0.9481376122683287,
      "num_tokens": 161920869.0,
      "step": 21850
    },
    {
      "entropy": 1.9949205078184604,
      "epoch": 0.3270836197278854,
      "grad_norm": 0.40930211544036865,
      "learning_rate": 3.896915582320988e-05,
      "loss": 0.232,
      "mean_token_accuracy": 0.9432474978268146,
      "num_tokens": 161990111.0,
      "step": 21860
    },
    {
      "entropy": 2.007215920090675,
      "epoch": 0.3272332462693895,
      "grad_norm": 0.3967520594596863,
      "learning_rate": 3.8959197465541095e-05,
      "loss": 0.2105,
      "mean_token_accuracy": 0.9459788091480732,
      "num_tokens": 162061571.0,
      "step": 21870
    },
    {
      "entropy": 1.9882694467902184,
      "epoch": 0.32738287281089357,
      "grad_norm": 0.4378608167171478,
      "learning_rate": 3.8949235891818444e-05,
      "loss": 0.245,
      "mean_token_accuracy": 0.9427026618272066,
      "num_tokens": 162135916.0,
      "step": 21880
    },
    {
      "entropy": 2.0088863745331764,
      "epoch": 0.32753249935239764,
      "grad_norm": 0.49103930592536926,
      "learning_rate": 3.893927110434524e-05,
      "loss": 0.259,
      "mean_token_accuracy": 0.9365622412413359,
      "num_tokens": 162207405.0,
      "step": 21890
    },
    {
      "entropy": 2.032796887308359,
      "epoch": 0.32768212589390167,
      "grad_norm": 0.3637041747570038,
      "learning_rate": 3.89293031054255e-05,
      "loss": 0.2256,
      "mean_token_accuracy": 0.9419725600630044,
      "num_tokens": 162280512.0,
      "step": 21900
    },
    {
      "entropy": 2.0212688580155374,
      "epoch": 0.32783175243540574,
      "grad_norm": 0.4264254570007324,
      "learning_rate": 3.8919331897364e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9406461082398891,
      "num_tokens": 162354151.0,
      "step": 21910
    },
    {
      "entropy": 1.993683560192585,
      "epoch": 0.3279813789769098,
      "grad_norm": 0.3256816267967224,
      "learning_rate": 3.890935748246628e-05,
      "loss": 0.2187,
      "mean_token_accuracy": 0.94453533180058,
      "num_tokens": 162427681.0,
      "step": 21920
    },
    {
      "entropy": 1.9639060840010643,
      "epoch": 0.3281310055184139,
      "grad_norm": 0.29746904969215393,
      "learning_rate": 3.889937986303858e-05,
      "loss": 0.2104,
      "mean_token_accuracy": 0.9487059436738491,
      "num_tokens": 162504018.0,
      "step": 21930
    },
    {
      "entropy": 2.0068055264651776,
      "epoch": 0.328280632059918,
      "grad_norm": 0.35065099596977234,
      "learning_rate": 3.888939904138792e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9442271359264851,
      "num_tokens": 162578850.0,
      "step": 21940
    },
    {
      "entropy": 1.9904586605727672,
      "epoch": 0.328430258601422,
      "grad_norm": 0.37796810269355774,
      "learning_rate": 3.8879415019822035e-05,
      "loss": 0.2258,
      "mean_token_accuracy": 0.9440772298723459,
      "num_tokens": 162650736.0,
      "step": 21950
    },
    {
      "entropy": 2.0260983295738697,
      "epoch": 0.3285798851429261,
      "grad_norm": 0.35720184445381165,
      "learning_rate": 3.8869427800649414e-05,
      "loss": 0.2262,
      "mean_token_accuracy": 0.9437341153621673,
      "num_tokens": 162722914.0,
      "step": 21960
    },
    {
      "entropy": 2.0135708704590796,
      "epoch": 0.32872951168443015,
      "grad_norm": 0.29358145594596863,
      "learning_rate": 3.885943738617927e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9464553758502007,
      "num_tokens": 162798839.0,
      "step": 21970
    },
    {
      "entropy": 1.982262659817934,
      "epoch": 0.32887913822593423,
      "grad_norm": 0.4329977035522461,
      "learning_rate": 3.884944377872158e-05,
      "loss": 0.207,
      "mean_token_accuracy": 0.94520303606987,
      "num_tokens": 162872994.0,
      "step": 21980
    },
    {
      "entropy": 2.0072369672358037,
      "epoch": 0.32902876476743825,
      "grad_norm": 0.2786324620246887,
      "learning_rate": 3.883944698058702e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9437058694660664,
      "num_tokens": 162947911.0,
      "step": 21990
    },
    {
      "entropy": 1.9870320603251457,
      "epoch": 0.32917839130894233,
      "grad_norm": 0.2841031551361084,
      "learning_rate": 3.882944699408706e-05,
      "loss": 0.2184,
      "mean_token_accuracy": 0.9439545970410108,
      "num_tokens": 163022854.0,
      "step": 22000
    },
    {
      "epoch": 0.32917839130894233,
      "eval_entropy": 1.984092717027664,
      "eval_loss": 0.24280153214931488,
      "eval_mean_token_accuracy": 0.9454244793534279,
      "eval_num_tokens": 163022854.0,
      "eval_runtime": 560.2949,
      "eval_samples_per_second": 35.695,
      "eval_steps_per_second": 8.924,
      "step": 22000
    },
    {
      "entropy": 1.9947519056499003,
      "epoch": 0.3293280178504464,
      "grad_norm": 0.3883761465549469,
      "learning_rate": 3.881944382153385e-05,
      "loss": 0.2287,
      "mean_token_accuracy": 0.9467568263411522,
      "num_tokens": 163097124.0,
      "step": 22010
    },
    {
      "entropy": 1.9650073267519474,
      "epoch": 0.3294776443919505,
      "grad_norm": 0.3093666732311249,
      "learning_rate": 3.88094374652403e-05,
      "loss": 0.2011,
      "mean_token_accuracy": 0.9489787604659796,
      "num_tokens": 163172170.0,
      "step": 22020
    },
    {
      "entropy": 1.9942278049886226,
      "epoch": 0.3296272709334545,
      "grad_norm": 0.33886194229125977,
      "learning_rate": 3.879942792752007e-05,
      "loss": 0.2199,
      "mean_token_accuracy": 0.9466099195182324,
      "num_tokens": 163247442.0,
      "step": 22030
    },
    {
      "entropy": 1.9741010949015618,
      "epoch": 0.3297768974749586,
      "grad_norm": 0.32196399569511414,
      "learning_rate": 3.878941521068754e-05,
      "loss": 0.2148,
      "mean_token_accuracy": 0.9476872123777866,
      "num_tokens": 163322716.0,
      "step": 22040
    },
    {
      "entropy": 2.0231268793344497,
      "epoch": 0.32992652401646266,
      "grad_norm": 0.38922715187072754,
      "learning_rate": 3.877939931705783e-05,
      "loss": 0.2329,
      "mean_token_accuracy": 0.9390785004943609,
      "num_tokens": 163397244.0,
      "step": 22050
    },
    {
      "entropy": 2.0104342699050903,
      "epoch": 0.33007615055796674,
      "grad_norm": 0.39372381567955017,
      "learning_rate": 3.876938024894679e-05,
      "loss": 0.2388,
      "mean_token_accuracy": 0.9394273795187473,
      "num_tokens": 163470546.0,
      "step": 22060
    },
    {
      "entropy": 1.9760512001812458,
      "epoch": 0.33022577709947076,
      "grad_norm": 0.35849446058273315,
      "learning_rate": 3.875935800867101e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9438125740736723,
      "num_tokens": 163542740.0,
      "step": 22070
    },
    {
      "entropy": 1.9456640169024468,
      "epoch": 0.33037540364097484,
      "grad_norm": 0.29396793246269226,
      "learning_rate": 3.8749332598547795e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9456471778452397,
      "num_tokens": 163616598.0,
      "step": 22080
    },
    {
      "entropy": 1.9560938701033592,
      "epoch": 0.3305250301824789,
      "grad_norm": 0.33489927649497986,
      "learning_rate": 3.873930402089522e-05,
      "loss": 0.2069,
      "mean_token_accuracy": 0.9448212962597609,
      "num_tokens": 163691681.0,
      "step": 22090
    },
    {
      "entropy": 1.947559378296137,
      "epoch": 0.330674656723983,
      "grad_norm": 0.3570326566696167,
      "learning_rate": 3.8729272278032045e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9434581238776445,
      "num_tokens": 163763139.0,
      "step": 22100
    },
    {
      "entropy": 1.949570919573307,
      "epoch": 0.3308242832654871,
      "grad_norm": 0.29445546865463257,
      "learning_rate": 3.8719237372277826e-05,
      "loss": 0.2012,
      "mean_token_accuracy": 0.9492198441177606,
      "num_tokens": 163837047.0,
      "step": 22110
    },
    {
      "entropy": 1.9799493305385112,
      "epoch": 0.3309739098069911,
      "grad_norm": 0.4709635376930237,
      "learning_rate": 3.8709199305952776e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.945633015781641,
      "num_tokens": 163908952.0,
      "step": 22120
    },
    {
      "entropy": 1.9920643419027328,
      "epoch": 0.3311235363484952,
      "grad_norm": 0.3318102955818176,
      "learning_rate": 3.8699158081377886e-05,
      "loss": 0.221,
      "mean_token_accuracy": 0.946069548651576,
      "num_tokens": 163983292.0,
      "step": 22130
    },
    {
      "entropy": 1.9889032244682312,
      "epoch": 0.33127316288999925,
      "grad_norm": 0.2715649902820587,
      "learning_rate": 3.8689113700874875e-05,
      "loss": 0.2155,
      "mean_token_accuracy": 0.9487286407500506,
      "num_tokens": 164056774.0,
      "step": 22140
    },
    {
      "entropy": 2.0127250850200653,
      "epoch": 0.33142278943150333,
      "grad_norm": 0.291069895029068,
      "learning_rate": 3.8679066166766165e-05,
      "loss": 0.2138,
      "mean_token_accuracy": 0.9456589709967375,
      "num_tokens": 164132069.0,
      "step": 22150
    },
    {
      "entropy": 2.015047777444124,
      "epoch": 0.33157241597300735,
      "grad_norm": 0.4519747197628021,
      "learning_rate": 3.8669015481374936e-05,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9462128173559904,
      "num_tokens": 164206344.0,
      "step": 22160
    },
    {
      "entropy": 1.9632880724966526,
      "epoch": 0.33172204251451143,
      "grad_norm": 0.3658117949962616,
      "learning_rate": 3.865896164702509e-05,
      "loss": 0.2076,
      "mean_token_accuracy": 0.9492143057286739,
      "num_tokens": 164278842.0,
      "step": 22170
    },
    {
      "entropy": 1.9718986608088016,
      "epoch": 0.3318716690560155,
      "grad_norm": 0.4263376295566559,
      "learning_rate": 3.864890466604124e-05,
      "loss": 0.2277,
      "mean_token_accuracy": 0.9421748846769333,
      "num_tokens": 164354290.0,
      "step": 22180
    },
    {
      "entropy": 1.9770046323537827,
      "epoch": 0.3320212955975196,
      "grad_norm": 0.3859645128250122,
      "learning_rate": 3.863884454074874e-05,
      "loss": 0.2046,
      "mean_token_accuracy": 0.9482002187520265,
      "num_tokens": 164427628.0,
      "step": 22190
    },
    {
      "entropy": 1.9907614447176456,
      "epoch": 0.3321709221390236,
      "grad_norm": 0.38428595662117004,
      "learning_rate": 3.862878127347368e-05,
      "loss": 0.2253,
      "mean_token_accuracy": 0.9448198597878218,
      "num_tokens": 164498394.0,
      "step": 22200
    },
    {
      "entropy": 2.013235003501177,
      "epoch": 0.3323205486805277,
      "grad_norm": 0.3255077004432678,
      "learning_rate": 3.861871486654286e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9478859852999448,
      "num_tokens": 164573016.0,
      "step": 22210
    },
    {
      "entropy": 2.02057753354311,
      "epoch": 0.33247017522203176,
      "grad_norm": 0.34081074595451355,
      "learning_rate": 3.860864532228382e-05,
      "loss": 0.2291,
      "mean_token_accuracy": 0.9440531175583601,
      "num_tokens": 164648557.0,
      "step": 22220
    },
    {
      "entropy": 1.964355606585741,
      "epoch": 0.33261980176353584,
      "grad_norm": 0.3173394501209259,
      "learning_rate": 3.8598572643024794e-05,
      "loss": 0.1986,
      "mean_token_accuracy": 0.9502705112099648,
      "num_tokens": 164723907.0,
      "step": 22230
    },
    {
      "entropy": 1.9810654371976852,
      "epoch": 0.3327694283050399,
      "grad_norm": 0.333822101354599,
      "learning_rate": 3.8588496831094797e-05,
      "loss": 0.2428,
      "mean_token_accuracy": 0.9424056190997362,
      "num_tokens": 164797241.0,
      "step": 22240
    },
    {
      "entropy": 1.985083206743002,
      "epoch": 0.33291905484654394,
      "grad_norm": 0.33599618077278137,
      "learning_rate": 3.857841788882351e-05,
      "loss": 0.2318,
      "mean_token_accuracy": 0.9448844894766808,
      "num_tokens": 164870567.0,
      "step": 22250
    },
    {
      "entropy": 1.9875101156532764,
      "epoch": 0.333068681388048,
      "grad_norm": 0.3883637487888336,
      "learning_rate": 3.8568335818541385e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9459547851234674,
      "num_tokens": 164947524.0,
      "step": 22260
    },
    {
      "entropy": 1.9986580431461334,
      "epoch": 0.3332183079295521,
      "grad_norm": 0.3494413495063782,
      "learning_rate": 3.855825062257955e-05,
      "loss": 0.2268,
      "mean_token_accuracy": 0.9417286153882742,
      "num_tokens": 165020527.0,
      "step": 22270
    },
    {
      "entropy": 2.0103697285056112,
      "epoch": 0.33336793447105617,
      "grad_norm": 0.3924461007118225,
      "learning_rate": 3.8548162303269915e-05,
      "loss": 0.246,
      "mean_token_accuracy": 0.9407783579081297,
      "num_tokens": 165092675.0,
      "step": 22280
    },
    {
      "entropy": 2.0127294093370436,
      "epoch": 0.3335175610125602,
      "grad_norm": 0.4026166796684265,
      "learning_rate": 3.8538070862945045e-05,
      "loss": 0.2113,
      "mean_token_accuracy": 0.9490899715572596,
      "num_tokens": 165167796.0,
      "step": 22290
    },
    {
      "entropy": 2.027315577864647,
      "epoch": 0.33366718755406427,
      "grad_norm": 0.408602774143219,
      "learning_rate": 3.852797630393828e-05,
      "loss": 0.2217,
      "mean_token_accuracy": 0.9447690561413765,
      "num_tokens": 165246609.0,
      "step": 22300
    },
    {
      "entropy": 2.0412926405668257,
      "epoch": 0.33381681409556835,
      "grad_norm": 0.39664673805236816,
      "learning_rate": 3.851787862858367e-05,
      "loss": 0.229,
      "mean_token_accuracy": 0.9428691904991865,
      "num_tokens": 165316268.0,
      "step": 22310
    },
    {
      "entropy": 1.9958080969750882,
      "epoch": 0.3339664406370724,
      "grad_norm": 0.2754150331020355,
      "learning_rate": 3.8507777839215946e-05,
      "loss": 0.2067,
      "mean_token_accuracy": 0.9483322449028492,
      "num_tokens": 165392222.0,
      "step": 22320
    },
    {
      "entropy": 1.989423394203186,
      "epoch": 0.33411606717857645,
      "grad_norm": 0.38209885358810425,
      "learning_rate": 3.849767393817061e-05,
      "loss": 0.2272,
      "mean_token_accuracy": 0.9427867665886879,
      "num_tokens": 165464842.0,
      "step": 22330
    },
    {
      "entropy": 1.9861025840044022,
      "epoch": 0.3342656937200805,
      "grad_norm": 0.3378436863422394,
      "learning_rate": 3.848756692778388e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9442888684570789,
      "num_tokens": 165537609.0,
      "step": 22340
    },
    {
      "entropy": 1.9796036012470721,
      "epoch": 0.3344153202615846,
      "grad_norm": 0.42756444215774536,
      "learning_rate": 3.8477456810392645e-05,
      "loss": 0.2332,
      "mean_token_accuracy": 0.9443303264677525,
      "num_tokens": 165613384.0,
      "step": 22350
    },
    {
      "entropy": 1.9721853196620942,
      "epoch": 0.3345649468030887,
      "grad_norm": 0.3878195583820343,
      "learning_rate": 3.8467343588334565e-05,
      "loss": 0.2189,
      "mean_token_accuracy": 0.9427981745451689,
      "num_tokens": 165688650.0,
      "step": 22360
    },
    {
      "entropy": 1.9629192598164082,
      "epoch": 0.3347145733445927,
      "grad_norm": 0.3269806206226349,
      "learning_rate": 3.845722726394799e-05,
      "loss": 0.2101,
      "mean_token_accuracy": 0.9468015395104885,
      "num_tokens": 165764963.0,
      "step": 22370
    },
    {
      "entropy": 1.9656088538467884,
      "epoch": 0.3348641998860968,
      "grad_norm": 0.3811371922492981,
      "learning_rate": 3.844710783957199e-05,
      "loss": 0.2115,
      "mean_token_accuracy": 0.9466866906732321,
      "num_tokens": 165838988.0,
      "step": 22380
    },
    {
      "entropy": 1.9964581154286862,
      "epoch": 0.33501382642760086,
      "grad_norm": 0.38405466079711914,
      "learning_rate": 3.843698531754637e-05,
      "loss": 0.2305,
      "mean_token_accuracy": 0.9400617584586144,
      "num_tokens": 165913235.0,
      "step": 22390
    },
    {
      "entropy": 1.9889946311712265,
      "epoch": 0.33516345296910494,
      "grad_norm": 0.28389832377433777,
      "learning_rate": 3.842685970021161e-05,
      "loss": 0.2235,
      "mean_token_accuracy": 0.9433514337986708,
      "num_tokens": 165988131.0,
      "step": 22400
    },
    {
      "entropy": 1.967279040068388,
      "epoch": 0.335313079510609,
      "grad_norm": 0.33313578367233276,
      "learning_rate": 3.8416730989908964e-05,
      "loss": 0.2054,
      "mean_token_accuracy": 0.9457957621663808,
      "num_tokens": 166064206.0,
      "step": 22410
    },
    {
      "entropy": 1.9666185714304447,
      "epoch": 0.33546270605211304,
      "grad_norm": 0.3528568148612976,
      "learning_rate": 3.8406599188980345e-05,
      "loss": 0.2202,
      "mean_token_accuracy": 0.9473166991025209,
      "num_tokens": 166138974.0,
      "step": 22420
    },
    {
      "entropy": 1.9873205877840519,
      "epoch": 0.3356123325936171,
      "grad_norm": 0.3034018874168396,
      "learning_rate": 3.839646429976843e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9425133876502514,
      "num_tokens": 166213737.0,
      "step": 22430
    },
    {
      "entropy": 1.9689572058618068,
      "epoch": 0.3357619591351212,
      "grad_norm": 0.342240571975708,
      "learning_rate": 3.838632632461656e-05,
      "loss": 0.2135,
      "mean_token_accuracy": 0.9487418826669455,
      "num_tokens": 166286471.0,
      "step": 22440
    },
    {
      "entropy": 1.9795271143317223,
      "epoch": 0.33591158567662527,
      "grad_norm": 0.349244087934494,
      "learning_rate": 3.8376185265868835e-05,
      "loss": 0.2221,
      "mean_token_accuracy": 0.9435827914625406,
      "num_tokens": 166359597.0,
      "step": 22450
    },
    {
      "entropy": 1.9884016081690787,
      "epoch": 0.3360612122181293,
      "grad_norm": 0.31827735900878906,
      "learning_rate": 3.8366041125870035e-05,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9439912039786578,
      "num_tokens": 166433703.0,
      "step": 22460
    },
    {
      "entropy": 1.9906642042100429,
      "epoch": 0.33621083875963337,
      "grad_norm": 0.37262120842933655,
      "learning_rate": 3.8355893906965676e-05,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9450167387723922,
      "num_tokens": 166506633.0,
      "step": 22470
    },
    {
      "entropy": 1.9850214943289757,
      "epoch": 0.33636046530113745,
      "grad_norm": 0.36646273732185364,
      "learning_rate": 3.834574361150197e-05,
      "loss": 0.227,
      "mean_token_accuracy": 0.9426586218178272,
      "num_tokens": 166579717.0,
      "step": 22480
    },
    {
      "entropy": 1.9831901043653488,
      "epoch": 0.3365100918426415,
      "grad_norm": 0.46940261125564575,
      "learning_rate": 3.833559024182584e-05,
      "loss": 0.2252,
      "mean_token_accuracy": 0.9434010289609432,
      "num_tokens": 166656778.0,
      "step": 22490
    },
    {
      "entropy": 1.9995835065841674,
      "epoch": 0.33665971838414555,
      "grad_norm": 0.3591897189617157,
      "learning_rate": 3.832543380028495e-05,
      "loss": 0.2068,
      "mean_token_accuracy": 0.9477128904312849,
      "num_tokens": 166731960.0,
      "step": 22500
    },
    {
      "entropy": 1.9835202097892761,
      "epoch": 0.3368093449256496,
      "grad_norm": 0.28715255856513977,
      "learning_rate": 3.831527428922762e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.9474842969328165,
      "num_tokens": 166808581.0,
      "step": 22510
    },
    {
      "entropy": 1.9983441658318042,
      "epoch": 0.3369589714671537,
      "grad_norm": 0.40073856711387634,
      "learning_rate": 3.8305111711002916e-05,
      "loss": 0.2293,
      "mean_token_accuracy": 0.9438222028315068,
      "num_tokens": 166883693.0,
      "step": 22520
    },
    {
      "entropy": 2.0246086359024047,
      "epoch": 0.3371085980086578,
      "grad_norm": 0.4178582727909088,
      "learning_rate": 3.829494606796062e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.944524110853672,
      "num_tokens": 166957017.0,
      "step": 22530
    },
    {
      "entropy": 1.9891228027641774,
      "epoch": 0.33725822455016186,
      "grad_norm": 0.28491654992103577,
      "learning_rate": 3.82847773624512e-05,
      "loss": 0.223,
      "mean_token_accuracy": 0.9438227467238903,
      "num_tokens": 167031159.0,
      "step": 22540
    },
    {
      "entropy": 2.0074817262589932,
      "epoch": 0.3374078510916659,
      "grad_norm": 0.35180261731147766,
      "learning_rate": 3.827460559682585e-05,
      "loss": 0.2163,
      "mean_token_accuracy": 0.9468038789927959,
      "num_tokens": 167107109.0,
      "step": 22550
    },
    {
      "entropy": 2.0127397678792476,
      "epoch": 0.33755747763316996,
      "grad_norm": 0.40493640303611755,
      "learning_rate": 3.826443077343645e-05,
      "loss": 0.2436,
      "mean_token_accuracy": 0.94097690731287,
      "num_tokens": 167178330.0,
      "step": 22560
    },
    {
      "entropy": 2.0012123823165893,
      "epoch": 0.33770710417467403,
      "grad_norm": 0.35406494140625,
      "learning_rate": 3.825425289463561e-05,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9445201061666012,
      "num_tokens": 167253525.0,
      "step": 22570
    },
    {
      "entropy": 1.982150959223509,
      "epoch": 0.3378567307161781,
      "grad_norm": 0.3615398705005646,
      "learning_rate": 3.824407196277663e-05,
      "loss": 0.2129,
      "mean_token_accuracy": 0.9488407384604216,
      "num_tokens": 167328276.0,
      "step": 22580
    },
    {
      "entropy": 1.974751351773739,
      "epoch": 0.33800635725768213,
      "grad_norm": 0.32915040850639343,
      "learning_rate": 3.823388798021352e-05,
      "loss": 0.2211,
      "mean_token_accuracy": 0.9452420361340046,
      "num_tokens": 167402691.0,
      "step": 22590
    },
    {
      "entropy": 1.9997020199894906,
      "epoch": 0.3381559837991862,
      "grad_norm": 0.28292548656463623,
      "learning_rate": 3.822370094930102e-05,
      "loss": 0.2209,
      "mean_token_accuracy": 0.944814781844616,
      "num_tokens": 167477560.0,
      "step": 22600
    },
    {
      "entropy": 2.0099263824522495,
      "epoch": 0.3383056103406903,
      "grad_norm": 0.4152122139930725,
      "learning_rate": 3.8213510872394527e-05,
      "loss": 0.2126,
      "mean_token_accuracy": 0.9452424012124538,
      "num_tokens": 167550774.0,
      "step": 22610
    },
    {
      "entropy": 1.99155660495162,
      "epoch": 0.33845523688219437,
      "grad_norm": 0.41171714663505554,
      "learning_rate": 3.820331775185018e-05,
      "loss": 0.2247,
      "mean_token_accuracy": 0.9456832867115736,
      "num_tokens": 167623134.0,
      "step": 22620
    },
    {
      "entropy": 2.006849817931652,
      "epoch": 0.3386048634236984,
      "grad_norm": 0.31894227862358093,
      "learning_rate": 3.819312159002481e-05,
      "loss": 0.2094,
      "mean_token_accuracy": 0.9501057080924511,
      "num_tokens": 167698149.0,
      "step": 22630
    },
    {
      "entropy": 2.0163633465766906,
      "epoch": 0.33875448996520247,
      "grad_norm": 0.36198076605796814,
      "learning_rate": 3.818292238927595e-05,
      "loss": 0.214,
      "mean_token_accuracy": 0.9464807070791721,
      "num_tokens": 167771687.0,
      "step": 22640
    },
    {
      "entropy": 2.032291793823242,
      "epoch": 0.33890411650670654,
      "grad_norm": 0.3464234471321106,
      "learning_rate": 3.817272015196184e-05,
      "loss": 0.2225,
      "mean_token_accuracy": 0.9428748242557049,
      "num_tokens": 167843821.0,
      "step": 22650
    },
    {
      "entropy": 2.032166563719511,
      "epoch": 0.3390537430482106,
      "grad_norm": 0.3723832368850708,
      "learning_rate": 3.8162514880441395e-05,
      "loss": 0.2142,
      "mean_token_accuracy": 0.9415746599435806,
      "num_tokens": 167917269.0,
      "step": 22660
    },
    {
      "entropy": 2.0131359875202177,
      "epoch": 0.3392033695897147,
      "grad_norm": 0.32584264874458313,
      "learning_rate": 3.8152306577074284e-05,
      "loss": 0.2255,
      "mean_token_accuracy": 0.9457547824829817,
      "num_tokens": 167990250.0,
      "step": 22670
    },
    {
      "entropy": 1.978609497100115,
      "epoch": 0.3393529961312187,
      "grad_norm": 0.40151363611221313,
      "learning_rate": 3.8142095244220844e-05,
      "loss": 0.2103,
      "mean_token_accuracy": 0.9466174811124801,
      "num_tokens": 168065061.0,
      "step": 22680
    },
    {
      "entropy": 1.967474202811718,
      "epoch": 0.3395026226727228,
      "grad_norm": 0.34766316413879395,
      "learning_rate": 3.8131880884242114e-05,
      "loss": 0.1996,
      "mean_token_accuracy": 0.9491662163287401,
      "num_tokens": 168140405.0,
      "step": 22690
    },
    {
      "entropy": 1.9728846274316312,
      "epoch": 0.3396522492142269,
      "grad_norm": 0.39087334275245667,
      "learning_rate": 3.812166349949982e-05,
      "loss": 0.2212,
      "mean_token_accuracy": 0.945611210912466,
      "num_tokens": 168213220.0,
      "step": 22700
    },
    {
      "entropy": 1.9953075408935548,
      "epoch": 0.33980187575573095,
      "grad_norm": 0.4397093951702118,
      "learning_rate": 3.811144309235643e-05,
      "loss": 0.2354,
      "mean_token_accuracy": 0.9421325724571943,
      "num_tokens": 168284072.0,
      "step": 22710
    },
    {
      "entropy": 2.006800177693367,
      "epoch": 0.339951502297235,
      "grad_norm": 0.2698858678340912,
      "learning_rate": 3.810121966517506e-05,
      "loss": 0.2054,
      "mean_token_accuracy": 0.9460940405726432,
      "num_tokens": 168361882.0,
      "step": 22720
    },
    {
      "entropy": 2.0176710568368437,
      "epoch": 0.34010112883873905,
      "grad_norm": 0.32487982511520386,
      "learning_rate": 3.8090993220319564e-05,
      "loss": 0.2146,
      "mean_token_accuracy": 0.9448831107467413,
      "num_tokens": 168433910.0,
      "step": 22730
    },
    {
      "entropy": 2.014442837238312,
      "epoch": 0.34025075538024313,
      "grad_norm": 0.32062992453575134,
      "learning_rate": 3.808076376015447e-05,
      "loss": 0.2365,
      "mean_token_accuracy": 0.9427271839231253,
      "num_tokens": 168506596.0,
      "step": 22740
    },
    {
      "entropy": 2.001173871010542,
      "epoch": 0.3404003819217472,
      "grad_norm": 0.3502158522605896,
      "learning_rate": 3.807053128704501e-05,
      "loss": 0.2219,
      "mean_token_accuracy": 0.9434507202357054,
      "num_tokens": 168579666.0,
      "step": 22750
    },
    {
      "entropy": 1.9657455727458,
      "epoch": 0.34055000846325123,
      "grad_norm": 0.4495643377304077,
      "learning_rate": 3.806029580335712e-05,
      "loss": 0.233,
      "mean_token_accuracy": 0.9418537691235542,
      "num_tokens": 168651765.0,
      "step": 22760
    },
    {
      "entropy": 1.9818620055913925,
      "epoch": 0.3406996350047553,
      "grad_norm": 0.32124900817871094,
      "learning_rate": 3.805005731145743e-05,
      "loss": 0.197,
      "mean_token_accuracy": 0.9500485964119434,
      "num_tokens": 168724225.0,
      "step": 22770
    },
    {
      "entropy": 1.9738273717463017,
      "epoch": 0.3408492615462594,
      "grad_norm": 0.5319764614105225,
      "learning_rate": 3.803981581371325e-05,
      "loss": 0.2157,
      "mean_token_accuracy": 0.9453102320432663,
      "num_tokens": 168798274.0,
      "step": 22780
    },
    {
      "entropy": 1.9728849284350871,
      "epoch": 0.34099888808776346,
      "grad_norm": 0.4935317635536194,
      "learning_rate": 3.802957131249259e-05,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9478708177804946,
      "num_tokens": 168872597.0,
      "step": 22790
    },
    {
      "entropy": 2.0031559698283674,
      "epoch": 0.3411485146292675,
      "grad_norm": 0.33401283621788025,
      "learning_rate": 3.8019323810164185e-05,
      "loss": 0.2224,
      "mean_token_accuracy": 0.9448565445840359,
      "num_tokens": 168949215.0,
      "step": 22800
    },
    {
      "entropy": 2.0192716658115386,
      "epoch": 0.34129814117077156,
      "grad_norm": 0.42012497782707214,
      "learning_rate": 3.8009073309097435e-05,
      "loss": 0.2299,
      "mean_token_accuracy": 0.9414010502398014,
      "num_tokens": 169025774.0,
      "step": 22810
    },
    {
      "entropy": 1.994770310819149,
      "epoch": 0.34144776771227564,
      "grad_norm": 0.3326359689235687,
      "learning_rate": 3.7998819811662415e-05,
      "loss": 0.2204,
      "mean_token_accuracy": 0.9447783924639225,
      "num_tokens": 169097540.0,
      "step": 22820
    },
    {
      "entropy": 1.993638276308775,
      "epoch": 0.3415973942537797,
      "grad_norm": 0.428134024143219,
      "learning_rate": 3.798856332022992e-05,
      "loss": 0.2164,
      "mean_token_accuracy": 0.9461646623909473,
      "num_tokens": 169173635.0,
      "step": 22830
    },
    {
      "entropy": 1.9706661030650139,
      "epoch": 0.3417470207952838,
      "grad_norm": 0.3939938545227051,
      "learning_rate": 3.797830383717146e-05,
      "loss": 0.2319,
      "mean_token_accuracy": 0.9438707262277604,
      "num_tokens": 169249926.0,
      "step": 22840
    },
    {
      "entropy": 1.9754851102828979,
      "epoch": 0.3418966473367878,
      "grad_norm": 0.4234197735786438,
      "learning_rate": 3.7968041364859197e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9411979436874389,
      "num_tokens": 169325409.0,
      "step": 22850
    },
    {
      "entropy": 2.0119520142674445,
      "epoch": 0.3420462738782919,
      "grad_norm": 0.5001817941665649,
      "learning_rate": 3.7957775905665994e-05,
      "loss": 0.2297,
      "mean_token_accuracy": 0.9439186301082373,
      "num_tokens": 169400930.0,
      "step": 22860
    },
    {
      "entropy": 1.9859436497092247,
      "epoch": 0.342195900419796,
      "grad_norm": 0.31200987100601196,
      "learning_rate": 3.79475074619654e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9457449648529291,
      "num_tokens": 169476342.0,
      "step": 22870
    },
    {
      "entropy": 2.0226414375007153,
      "epoch": 0.34234552696130005,
      "grad_norm": 0.3898511230945587,
      "learning_rate": 3.793723603613168e-05,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9461279354989529,
      "num_tokens": 169551401.0,
      "step": 22880
    },
    {
      "entropy": 2.0437833279371262,
      "epoch": 0.3424951535028041,
      "grad_norm": 0.323141872882843,
      "learning_rate": 3.7926961630539753e-05,
      "loss": 0.2267,
      "mean_token_accuracy": 0.9444992698729038,
      "num_tokens": 169624185.0,
      "step": 22890
    },
    {
      "entropy": 2.015289480239153,
      "epoch": 0.34264478004430815,
      "grad_norm": 0.36977940797805786,
      "learning_rate": 3.791668424756525e-05,
      "loss": 0.2302,
      "mean_token_accuracy": 0.942643329501152,
      "num_tokens": 169698352.0,
      "step": 22900
    },
    {
      "entropy": 2.017540106922388,
      "epoch": 0.34279440658581223,
      "grad_norm": 0.3740963041782379,
      "learning_rate": 3.790640388958449e-05,
      "loss": 0.1969,
      "mean_token_accuracy": 0.9530144531279803,
      "num_tokens": 169770764.0,
      "step": 22910
    },
    {
      "entropy": 2.003644347935915,
      "epoch": 0.3429440331273163,
      "grad_norm": 0.2741229236125946,
      "learning_rate": 3.789612055897447e-05,
      "loss": 0.2214,
      "mean_token_accuracy": 0.9436117220669985,
      "num_tokens": 169842991.0,
      "step": 22920
    },
    {
      "entropy": 1.9877101913094521,
      "epoch": 0.3430936596688203,
      "grad_norm": 0.28630098700523376,
      "learning_rate": 3.7885834258112876e-05,
      "loss": 0.2058,
      "mean_token_accuracy": 0.9502271637320518,
      "num_tokens": 169922802.0,
      "step": 22930
    },
    {
      "entropy": 1.9902968160808086,
      "epoch": 0.3432432862103244,
      "grad_norm": 0.39478009939193726,
      "learning_rate": 3.7875544989378095e-05,
      "loss": 0.1956,
      "mean_token_accuracy": 0.9487415678799153,
      "num_tokens": 169998269.0,
      "step": 22940
    },
    {
      "entropy": 1.9926106356084348,
      "epoch": 0.3433929127518285,
      "grad_norm": 0.39118868112564087,
      "learning_rate": 3.786525275514918e-05,
      "loss": 0.2084,
      "mean_token_accuracy": 0.9488235156983137,
      "num_tokens": 170073208.0,
      "step": 22950
    },
    {
      "entropy": 1.9972306117415428,
      "epoch": 0.34354253929333256,
      "grad_norm": 0.3511309325695038,
      "learning_rate": 3.785495755780587e-05,
      "loss": 0.2042,
      "mean_token_accuracy": 0.9496672347187995,
      "num_tokens": 170146234.0,
      "step": 22960
    },
    {
      "entropy": 1.9996977671980858,
      "epoch": 0.34369216583483664,
      "grad_norm": 0.37187445163726807,
      "learning_rate": 3.7844659399728626e-05,
      "loss": 0.2263,
      "mean_token_accuracy": 0.9438729170709849,
      "num_tokens": 170219538.0,
      "step": 22970
    },
    {
      "entropy": 2.0048478439450266,
      "epoch": 0.34384179237634066,
      "grad_norm": 0.3165706992149353,
      "learning_rate": 3.7834358283298534e-05,
      "loss": 0.1842,
      "mean_token_accuracy": 0.9544384509325028,
      "num_tokens": 170293173.0,
      "step": 22980
    },
    {
      "entropy": 2.047573671489954,
      "epoch": 0.34399141891784474,
      "grad_norm": 0.4817623794078827,
      "learning_rate": 3.782405421089741e-05,
      "loss": 0.2401,
      "mean_token_accuracy": 0.9407077964395285,
      "num_tokens": 170363769.0,
      "step": 22990
    },
    {
      "entropy": 2.0440934896469116,
      "epoch": 0.3441410454593488,
      "grad_norm": 0.3303025960922241,
      "learning_rate": 3.781374718490774e-05,
      "loss": 0.2096,
      "mean_token_accuracy": 0.9483917437493801,
      "num_tokens": 170437591.0,
      "step": 23000
    },
    {
      "epoch": 0.3441410454593488,
      "eval_entropy": 2.0569298552274704,
      "eval_loss": 0.24171218276023865,
      "eval_mean_token_accuracy": 0.9456899745345115,
      "eval_num_tokens": 170437591.0,
      "eval_runtime": 558.9131,
      "eval_samples_per_second": 35.784,
      "eval_steps_per_second": 8.946,
      "step": 23000
    }
  ],
  "logging_steps": 10,
  "max_steps": 66834,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7075226725891178e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}